BMJ · 2018-04-30 · Malaz Boustani, School of Medicine, University of Indiana, Indiana United States Professor of Medicine Louise Robinson, Newcastle University Institute for Ageing,

Confidential: For Review Only

The association between anticholinergic medication and

dementia incidence varies with pattern of exposure and medication class: A nested case-control study using the

Clinical Practice Research Datalink

Journal: BMJ

Manuscript ID BMJ.2017.042315

Article Type: Research

BMJ Journal: BMJ

Date Submitted by the Author: 10-Nov-2017

Complete List of Authors: Richardson, Kathryn; University of East Anglia, ; Fox, Chris; Norwich Medical School, Department of Psychological Sciences Maidment, Ian; University of Aston, Steel, Nicholas; University of East Anglia, Norwich Medical School Loke, Yoon; University of East Anglia, School of Medicine Arthur, Antony; University of East Anglia, School of Health Sciences University of East Anglia, Norwich Research Park

myint, phyo; University of Aberdeen, ; University of East Anglia, Norwich Medical School Grossi, Carlota; University of East Anglia, School of Health Sciences Mattishent, Katharina; University of East Anglia, Norwich Medical School Bennett, Kathleen; Royal College of Surgeons in Ireland (RCSI), Population Health Sciences Campbell, Noll; Purdue University College of Pharmacy, Department of Pharmacy Practice Boustani, Malaz; Indiana University, Center for Aging Research Robinson, Louise; Institute for Health and Society, Newcastle University Brayne, Carol; University of Cambridge, Institute of Public Health Matthews, Fiona; Newcastle University, Institute for Health and Society,

Biomedical Research Building, Campus for Ageing and Vitality Savva, George; Quadram Institute Bioscience, Analytical Sciences Unit; University of East Anglia, School of Health Sciences

Keywords: dementia, pharmacoepidemiology, anticholinergic, Primary care

https://mc.manuscriptcentral.com/bmj

BMJ

Confidential: For Review OnlyThe association between anticholinergic medication and dementia incidence varies with

pattern of exposure and medication class: A nested case-control study using the Clinical

Practice Research Datalink

Kathryn Richardson, Chris Fox, Ian Maidment, Nicholas Steel, Yoon K Loke, Antony Arthur, Phyo K

Myint, Carlota M Grossi, Katharina Mattishent, Kathleen Bennett, Noll Campbell, Malaz Boustani,

Louise Robinson, Carol Brayne, Fiona E Matthews, George M Savva

Kathryn Richardson, School of Health Sciences, University of East Anglia, Norwich, NR4 7TJ, UK

Research Fellow in Statistics

Chris Fox, Norwich Medical School, University of East Anglia, Norwich, NR4 7TJ,

Clinical Professor

Ian Maidment, School of Life and Health Sciences, Aston University, Birmingham, B4 7ET, UK

Senior Lecturer in Clinical Pharmacy

Nicholas Steel, Norwich Medical School, University of East Anglia, Norwich, NR4 7TJ

Clinical Professor in Public Health

Yoon K Loke, Norwich Medical School, University of East Anglia, Norwich, NR4 7TJ

Professor of Medicine and Pharmacology

Antony Arthur, School of Health Sciences, University of East Anglia, Norwich, NR4 7TJ, UK

Professor of Nursing Science,

Phyo K Myint, School of Medicine and Dentistry, University of Aberdeen, Aberdeen, AB25 2ZD, UK

Clinical Chair in Medicine of Old Age,

Carlota M Grossi, School of Health Sciences, University of East Anglia, Norwich, NR4 7TJ, UK

Senior Research Associate,

Katharina Mattishent, Norwich Medical School, University of East Anglia, Norwich, NR4 7TJ

Clinical Research Fellow,

Kathleen Bennett, Division of Population Health Sciences, Royal College of Surgeons in Ireland,

Dublin, Ireland Associate Professor in Pharmacoepidemiology

Noll Campbell, Department of Pharmacy Practice, College of Pharmacy, Purdue University, US

Assistant Professor

Malaz Boustani, School of Medicine, University of Indiana, Indiana United States

Professor of Medicine

Louise Robinson, Newcastle University Institute for Ageing, Newcastle University, NE4 5PL, UK

Professor of Primary Care and Ageing

Carol Brayne, Professor of Public Health Medicine, Cambrdge Institute of Public Health, University of

Cambridge, CB2 0SR, UK

Fiona E Matthews, Institute for Health and Society, Newcastle University, NE4 5PL, UK

Professor of Epidemiology

George M Savva, School of Health Sciences, University of East Anglia, NR4 7TJ, UK

Senior Lecturer in Applied Statistics

Correspondence to: K Richardson [email protected]

Word count: 4999

Page 1 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Key words: dementia, Alzheimer’s disease, pharmacoepidemiology, case-control study,

anticholinergic, medication use, antimuscarinic

"What this paper adds"

Section 1: What is already known on this subject

• Use of medications with strong anticholinergic activity is associated with impaired

cognition in the short term.

• Published observational studies have found associations between anticholinergic use

and future cognitive decline and dementia incidence, but whether this is causal and

is directly attributable to anticholinergic activity is still unknown.

Section 2: What this study adds

• Our study confirms the previously observed association, but only for specific classes

of definite anticholinergics. Anticholinergic antidepressants, antiparkinsonians and

urologicals are linked to future dementia incidence, but antihistamines and

antispasmodics are not.

• Use of anticholinergic antidepressants and urologicals are positively associated with

dementia diagnosis 15-20 years after the exposure.

• Clinical practice and future research to reduce potentially harmful medication should

focus on risks associated with specific classes of anticholinergic medication.

Page 2 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review OnlyABSTRACT

OBJECTIVES: To estimate how duration and level of exposure to different classes of

anticholinergic medication is associated with subsequent incident dementia.

DESIGN: Case-control study

SETTING: General practices in the United Kingdom contributing to the Clinical Practice

Research Datalink.

PARTICIPANTS: 40,770 patients aged 65-99 years and diagnosed with dementia between

April 2006 and July 2015, and 283,933 controls without dementia matched 7:1 on date, sex,

age, deprivation, and years of data history.

INTERVENTIONS: Daily defined doses (DDD) of anticholinergic medications coded according

to the Anticholinergic Cognitive Burden (ACB) scale; in total and grouped by subclass;

prescribed between 4-20 years prior to dementia diagnosis.

MAIN OUTCOME MEASURES: Odds ratios for incident dementia, adjusted for a wide range

of demographics and health-related covariates.

RESULTS: 14,453 (35%) cases and 86,403 (30%) controls were prescribed at least one potent

(ACB score 3, ACB3) anticholinergic during the exposure period. The adjusted odds ratio

(aOR) for ‘any ACB3’ was 1.11 (95% confidence interval 1.08 to 1.14), and an increasing odds

of dementia was significantly associated with an increasing average ACB score. When

considered by class, ACB3 antihistamines or antispasmodics were not significantly linked to

dementia. Within ACB3 antidepressants, urologicals, and antiparkinsonians, risks increased

with exposure with this increase observed for exposure 15-20 years before diagnosis.

CONCLUSIONS: In the largest study to date, a robust association between some classes of

anticholinergic medication use and future dementia incidence was observed. This could be

caused by a class-specific effect, or by medications being used for very early symptoms of

dementia. Future pharmacological and clinical research must examine subgroups of

anticholinergics as opposed to anticholinergic effects per se or summing scales for

anticholinergic exposure.

TRIAL REGISTRATION: EUPAS8705 (ENCePP e-register of studies)

Page 3 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review OnlyIntroduction

Dementia is a leading cause of disability and death worldwide (1), and its prevention is a

global public health priority. Dementia is caused by a number of different

neurodegenerative processes that contribute to irreversible cognitive decline and

associated symptoms, progressive loss of independence and daily functioning. Mixed

dementias are more prevalent than is often recognised, with symptoms often more closely

linked to overall pathological burden as opposed to any specific disease process (2,3). No

disease modifying treatments for dementia exist, however, age-specific dementia incidence

across populations is declining, suggesting that changing lifestyles or environment may lead

to a meaningful change in dementia prevalence (4). Hence identifying and reducing

exposure to risk factors that can affect any aspect of long term brain health is important for

dementia prevention and cognitive health in the population (5).

Multiple medication use is increasing in middle aged and older populations (6,7), but the

potential harms of long term medication use are not well understood. Medications with

anticholinergic activity (henceforth anticholinergics) block the neurotransmitter

acetylcholine in the central or peripheral nervous system, and have diverse actions

depending on site. Anticholinergics are successfully used in the treatment of many

conditions including urinary incontinence, Parkinson’s disease, depression, epilepsy,

gastrointestinal disorders and to manage allergies. It is well known that anticholinergics

affect cognition (8), and guidelines suggest they are to be avoided among frail older people

(9). Use of anticholinergics among people with dementia is recognised as inappropriate by

both Beers (10) and STOPP (11) criteria for potentially inappropriate prescribing. Over the

past decade, prolonged exposure to anticholinergics has been linked to long term cognitive

decline or dementia incidence among community living cohorts and nursing home residents

Page 4 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Only(12–17). However, these studies have been limited in their ability to determine if the

increased risk is specific to anticholinergic action itself, and whether or not the association is

due to the medications themselves or the underlying conditions for which they were

prescribed.

Here we present a nested case-control study using the UK’s Clinical Practice Research

Datalink (CPRD; 18), in which we select patients with a new dementia diagnosis, and

compare their prescriptions of anticholinergic medication between 4 and 20 years prior to

dementia diagnosis with that of a matched group of control patients without dementia. Our

objectives were (a) to estimate the association between chronic anticholinergic medication

use with future dementia incidence while controlling for potential confounders, (b) to

explore whether any observed effect was specific to particular subclasses of medication,

and (c) to test how the association varied with time to dementia incidence and amount of

exposure within each class.

Methods

Study design

We performed a nested case-control study using data from CPRD, which includes

anonymised diagnosis, referral and prescribing records for more than 11.3 million patients

from 674 primary care practices across the UK, and is broadly representative of the UK

population in terms of age, sex and ethnicity (18). All coded information in the primary

care record for each selected patient is available to researchers. This includes demographic

detail, lifestyle information, and any diagnoses, symptoms recorded by the general

practitioner, referrals to other healthcare services and subsequent findings, and treatments

Page 5 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Onlyinitiated or continued in primary care. In the UK a patient’s registered primary care practice

is the coordinator of the vast majority of their healthcare, and acts as the main gateway for

access to secondary care. Hence most of a patient’s health care information, except the

detail of secondary care episodes or privately obtained healthcare, where they have not

been communicated to primary care, is held in their primary care record. An ‘up to

standard’ (UTS) date is recorded for each CPRD practice, which is the date at which the data

recorded by the practice is considered to be of an acceptable research standard where

diagnoses are assumed to have been recorded correctly. Recording of therapies and

diagnoses made prior to UTS date are available, but may be incomplete.

Selection of cases and controls

Patients with a recorded diagnosis of dementia made between April 2006 and July 2015,

and aged 65-99 years were eligible to be selected as cases. We also required at least six

year of UTS data prior to diagnosis to allow for a 4 year lag before dementia diagnosis, as

well as at least one year drug exposure period (DEP), and one year before the DEP for

ascertainment of covariates. Dementia diagnosis was defined as the presence in the record

of any Read codes for dementia as a diagnosis, symptom or referral, or prescription of a

cognitive enhancer (memantine, donepezil, rivastigmine, galantamine or tacrine) if that was

also followed by a dementia diagnosis code within 12 months. Cases were excluded if they

had been diagnosed with motor neurone disease, HIV/AIDS, multiple sclerosis, Down

syndrome or alcohol abuse prior to dementia diagnosis.

Page 6 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review OnlyUsing the dementia diagnosis date as the index date, each case was matched to up to seven

controls who had not been diagnosed with dementia diagnosis up to the index date. Cases

and controls were matched on sex, year of birth (within three years), years of UTS data

history, and practice-level index of multiple deprivation (IMD) quintile. We used incidence

density sampling to select controls, hence cases were eligible to be selected as controls for

other cases with earlier index dates (19).

Patients were excluded if their date of dementia diagnosis was ambiguous. That is, if

“dementia annual review’, “H/O: dementia”, “Assessment of psychotic and behavioural

symptoms of dementia”, or “Antipsychotic drug therapy for dementia” was recorded before

the index date.

Anticholinergic medication exposure

A drug exposure period (DEP) was defined for each case-control group, starting one year

after UTS data recording and ending four years before the index date. We excluded

exposures in the four years prior to index date to avoid protopathic bias, whereby the

medication is given for a sign or symptom of dementia prior to dementia diagnosis (22).

Estimating the anticholinergic effect of individual medications on the human brain is

difficult, and serum anticholinergic activity as measured via receptor bioassay do not

correlate well with effects on cognition (16,20), and so the anticholinergic effects of

medications are usually classified using scales developed by expert consensus aided by

literature review (21). In the current study all medication prescribed to each patient during

the DEP was classified according to the 2012 update of the Anticholinergic Cognitive Burden

Page 7 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Only(ACB) scale (23). Medications with serum anticholinergic activity or in vitro affinity to

muscarinic receptors but with no known clinically relevant negative cognitive effects are

scored 1, while drugs with established and clinically relevant anti-cholinergic effects are

scored 2 based on blood-brain penetration and 3 if also have reported associations with

delirium. All other drugs are scored 0. For the medications available in the UK in the last 30

years not rated on the ACB scale, we made the following assumptions. We classified all

thiazide diuretics, loop diuretics and antihistamines that had not been scored as ACB1, all

tricyclic antidepressants not scored as ACB3, and all creams, eye and ear drops as not having

anticholinergic activity.

Our primary exposures were the number of defined daily doses (DDDs) prescribed within

each category of ACB score during the DEP. The DDD is defined as the assumed average

maintenance dose per day for a drug based on its main indication in adults, using the DDD

values assigned by the World Health Organisation’s (WHO) Collaborating Centre for Drug

Statistics Methodology. We classified exposure as 0, 1-13, 14-89, 90-364, 365-1459 or

more than 1460 DDDs during the DEP.

We further categorised the DDDs within the class of each medication with anticholinergic

properties based as either analgesic (WHO Anatomical Therapeutic Chemical [ATC] N02*),

antidepressant (ATC N06A*), antipsychotic (ATC N05A*), cardiovascular (ATC C*), gastro-

intestinal (ATC A*), Parkinsonian (ATC N04*), respiratory (ATC R*), urological (ATC G04*), or

other. These categories were pre-specified based on common classes of anticholinergic

medications.

Finally, an average ACB score was calculated as the average across the DEP of the sum of

ACB score each medication being used at any given time. This was then categorised into

Page 8 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Onlygroups reflecting approximate quintiles of the sample, but with two additional groups

representing those with particularly high (3-5 and 5 or more) average anticholinergic scores.

Covariates

We considered potential confounders as any variable suspected to be linked to dementia

incidence or an indication for any of the medications examined. For a full list of covariates

and their definition see the Appendix, but in short we included diagnoses of cardiovascular

disease, other dementia risk factors and correlates, other indications for anticholinergics,

other medication use, sociodemographic variables and records of health-related lifestyle

where available.

Exposures could occur at any time during an individual’s DEP, and so as with many case-

control studies (24) it is not clear at what point we should determine the presence or

absence of confounding factors. Our primary analysis measured confounders recorded up to

the end of the DEP to best capture the medication indications. However, as some variables

could be consequences of medication exposure we also coded the status of each patient

with respect to each covariate up to the start of the DEP.

Statistical analysis

Patterns of exposures and covariates were described for case and control groups separately.

Primary analysis compared the number of DDDs of medications with ACB score of 1, 2 and 3

prescribed to cases and controls during the DEP, controlling for covariates recorded at the

end of each DEP. We used conditional logistic regression to estimate the association

Page 9 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Onlybetween anticholinergic prescriptions and dementia diagnosis, adjusting for covariates

described above. Adjusted odds ratios (OR) are reported with 95% confidence intervals (CI),

however p<0.01 was pre-specified as a threshold for statistical significance owing to large

number of subgroups being examined. All analysis was conducted using Stata version 14.

Secondary pre-specified analyses included (i) disaggregation of exposure by drug class and

(ii) testing the effect of exposure 4-10, 10-15 and 15-20 years prior to the index date in

those with 6 or more, 11 or more and 16 or more years of UTS data history respectively and

(iii) testing the effect of average ACB score over the DEP.

Sensitivity analyses included repeating our primary analysis (i) adjusted for covariates

measured up to the start rather than the end of the DEP, and the following post-hoc

analyses of (ii) emulating a new-user design by excluding patients prescribed medications

with an ACB score of 2 and 3 in the 12 months prior to the DEP (25), and (iii) recoding binary

exposure variables to correspond to 90 or more DDDs prescribed instead of any prescription

during the DEP, to represent longer term rather than one-off exposure..

To test the likely impact of missing lifestyle covariate data, we compared findings with and

without adjustment for these lifestyle variables in (i) complete data and (ii) in the full

dataset with use of a missing category for each variable.

For a final pre-specified sensitivity analysis, we recoded medications according to the

Anticholinergic Drug Scale (ADS) instead of the ACB scale. ADS classifies the degree of

anticholinergic activity of each medication on a scale from 0 (no anticholinergic effect) to 3

(marked anticholinergic effect) according to literature review and expert consensus (26).

Sample size was determined by the maximum available data within CPRD.

Page 10 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Protocol registration

The study protocol was approved by the Independent Scientific Advisory Committee (ISAC)

for CPRD research (protocol number 15_056R) prior to any data being released to the

research team and was made available to journal reviewers. The study was registered on

the ENCePP e-register of pharmacoepidemiology studies (register number EUPAS9705).

Patient involvement

Four Alzheimer’s Society Research Network Volunteers acted as study monitors and service

user representatives on our study steering committee. These individuals have all acted as

carers of people with dementia. The monitors contributed to our protocol development by

sharing their experiences of psychoactive medication use and their view of the balance

between the benefits of medication use and potential cognitive decline. Monitors met the

study team to discuss study progress every six months. They are assisting us with the

dissemination of our study, making sure that lay summaries of results are accessible and

avoid possible misinterpretation.

Results

The source population consisted of 66,136 patients diagnosed with dementia between April

2006 and July 2015. After applying the exclusion criteria, our analysis included 40,770 cases

and 283,933 controls, with the vast majority of cases being matched to seven controls

Page 11 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Only(efigure 1). Most patients (n=254,083, 78%) were from general practices in England, with

also 30,817 (9%) from Scotland, 29,575 (9%) from Wales, and 10,228 (3%) from Northern

Ireland. The median (IQR) age of patients at index date was 83 (78-87) years and 63% were

female. The median DEP duration was 7.1 (IQR: 4.0-11.3; range: 1-16) years. Diagnoses of

conditions related to dementia and indications for anticholinergics increased during the

DEP, for example the proportion of dementia cases diagnosed with depression before the

start of the DEP was 12% (n=5,071), rising to 20% (n=8,030) during the DEP (table 1).

Information on smoking status, harmful alcohol use, and BMI becomes more complete by

the end of the DEP. As expected, the average BMI decreases during the DEP for dementia

cases, but increases for the control group.

Frequency of anticholinergic medication use

14,453 (35%) cases and 86,403 (30%) controls were prescribed at least one anticholinergic

classified as ACB3 during their DEP. A total of 1,793,505 ACB3 prescriptions were written

during the DEP, with the five most common being amitriptyline (29% of ACB3 prescriptions)

dosulepin (16%), paroxetine (8%), oxybutynin (7%) and tolterodine (7%). Only 1429 (3.5%)

cases and 7909 (2.8%) controls were prescribed an ACB2 medication, with carbamazepine

accounting for 87% of ACB2 prescriptions. The vast majority of patients received at least

one ACB1 prescription during their DEP (89% of cases and 87% of controls) with

cardiovascular medication accounting for 63% of these prescriptions. The number of

prescriptions within each ACB category is further described in etable 1.

Page 12 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review OnlyPrimary analysis

There was a positive and significant association between the prescription of any ACB1, ACB2

or ACB3 medication and dementia with corresponding odds ratios of 1.10 (95% confidence

interval 1.06 to 1.15), 1.10 (1.03 to 1.16), and 1.11 (1.08 to 1.14), adjusted for covariates

measured at the end of the DEP (table 2). A dose-response effect was evident for ACB2 and

ACB3 doses prescribed, supporting a causal interpretation However there was a consistent

OR of approximately 1.1 for those prescribed ACB1 medications, regardless of how many

were prescribed (table 2), and including those with less than 14 DDD exposure, suggesting

residual confounding rather than a causal effect of ACB1 on dementia incidence. The

associations are attenuated from the crude rates when adjusting for medication use prior to

DEP, and are further reduced only slightly when adjusting for covariates at the end of the

DEP.

There was a dose-response effect linking average anticholinergic load, measured by the

average total ACB score over the DEP, with dementia incidence, although this is only evident

for those with an average score of 3 or more (table 2).

Exposure by medication class

When analysed by class, there was a substantial association between dementia incidence

and any prescription of ACB3 antidepressants, Parkinsonian drugs and urological

medications, but no association with ACB3 antispasmodics, antipsychotics, antihistamines,

or other ACB3s (table 3). ACB2 prescriptions were relatively rare, and so results are

imprecise in this group, but there is some evidence for an association between dementia

Page 13 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Onlyincidence and prescription of ACB2 Parkinsonian drugs. For ACB1, associations with

increased dementia risk are only observed for antidepressants but not with any other ACB1

medication class. Associations between dementia incidence and number of DDDs by class

is shown in appendix table 2 and are consistent with the findings in table 3.

Exposure by time before dementia diagnosis

When examining the effect of exposure in three different time different periods (4-10, 10-

15, and 15-20 years) prior to the index date, associations with any ACB3 prescriptions were

consistent across the DEP while associations for ACB1 and ACB2 were more apparent closer

to the index date (table 4). In particular, the prescription of any ACB3 medication 15-20

years prior to dementia was significantly associated with greater dementia incidence with

OR of 1.17 (1.10 to 1.24) adjusted for covariates at the start of the DEP. Prescriptions at 15-

20 years prior to dementia for ACB3 antidepressants and urologicals remained consistently

significantly associated with dementia incidence with OR of 1.19 (1.10 to 1.29) and 1.27

(1.09 to 1.48). However, for ACB1 antidepressants, the association with dementia increased

for prescriptions given in periods closer to dementia.

Sensitivity analyses

Excluding those with an ACB2 or ACB3 prescription in the 12 months before the DEP, hence

restricting to ‘new’ ACB2/3 users during the DEP, excluded 5,215 cases and 62,161 controls.

This led to small reductions in the association between dementia incidence and any ACB3

with OR of 1.07 (1.04 to 1.10) and any ACB2 with OR of 1.12 (1.04 to 1.21) adjusted for

covariates measured at the end of the DEP (etable 3). No substantial changes were seen in

Page 14 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Onlythe association between dementia incidence and specific classes of medication use when

restricted to new users. In particular, ACB3 antidepressants and urologicals and ACB1

antidepressants remained significantly associated with dementia.

Findings remained similar when also analysed by 90+ DDDs of exposure versus less, rather

than any prescription versus none (results not shown).

Although there were differences in the number of medications rated ACB1 and ACB2 and

ADS1 and ADS2, the proportion of patients with any prescription of ACB3 or ADS3

medications were similar (etable 4). There was good concordance between ADS3 and ACB3

(Cohen’s kappa = 0.91). Prescription of any ADS3 medication had a marginally lower

association with dementia with an odds ratio of 1.08 (1.06 to 1.11) adjusted for covariates

at the end of the DEP, than ACB3 (odds ratio 1.11 [1.08 to 1.14]). However, when analysed

according to the medication class, the findings were very similar with the ADS3

antidepressants, Parkinsonian and urologicals scoring and ADS1 antidepressants scoring

consistently associated with greater dementia incidence (data not shown), but no

association with other ADS classes.

Missing data

There were relatively high proportions with missing data for the BMI, smoking and alcohol

variables. There was very little difference in any results when comparing estimates with and

without adjustment for these variables among the main dataset and those with complete

data (etable 5), suggesting that our findings are not sensitive to missing data in these

covariates.

Page 15 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review OnlyDiscussion

Statement of principal findings

In this case-control study of older adults in the United Kingdom, there was a significant

association between increasing total anticholinergic burden over the previous 4-20 years

and incident dementia diagnosis. However this dose-response effect was not seen for

cumulative use of medications classed 1 on the ACB scale, and was only evident for certain

classes of anticholinergics. ACB3 antidepressants (predominantly Amitriptyline, Dosuplepin

and Paroxetine) and urologicals (predominantly Oxybutynin and Tolterodine) were

consistently associated with incident dementia. These relationships were seen even for

exposures between 15 and 20 years prior to the dementia diagnosis, suggesting that reverse

causation or confounding with early dementia symptoms are less likely explanations for the

effect, although depression in midlife may be a risk factor for later dementia. Prescription

of ACB3 antihistamines and gastrologicals was not associated with dementia in any of our

analysis. Other antidepressants (those coded as ACB1, predominantly SSRIs) were

associated with dementia, but that association was greater for prescriptions close to

dementia incidence, suggesting that reverse causation could be a possible explanation for

this observed association. Other ACB1 medications were not associated with increased

dementia incidence.

Strengths and weaknesses of the study

Our study used a large population representative primary care database that allowed a

detailed analysis of the association between dementia incidence and prescriptions of many

Page 16 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Onlydifferent drugs classes up to 20 years prior to dementia diagnosis. This is the first study to

simultaneously examine classes of anticholinergic medications, hence to have discovered

differences in association with dementia across these classes. However, there are several

limitations to our approach.

Dementia diagnosis codes in CPRD reflect GP diagnoses well (27), however dementia is

known to be under-diagnosed, with up to 50% of people with dementia not receiving a

diagnosis (28). It is possible that those more likely to be prescribed anticholinergic

medication are also more likely to be diagnosed with dementia, leading to surveillance bias,

but this should be apparent for all medication classes and so is unlikely to account for our

findings.

It is difficult to accurately assess the anticholinergic activity of each medication, and

anticholinergic scales differ with respect to how they classify medications (21). However,

there was good agreement between the anticholinergic drug scale (ADS) and the

anticholinergic cognitive burden (ACB) and our findings did not differ when using each scale.

We do not know whether patients were adherent to their prescribed medications.

Medications obtained over-the-counter (OTC) are not recorded in our data, hence we will

have likely underestimated the use of particular antihistamines. DDDs can be difficult to

establish for certain medications yet represent the best available method for comparing the

levels of exposure of different drug classes. Our findings did not change when we instead

analysed the number of prescriptions to quantify exposure (results not shown).

As in any observational study, unmeasured or residual confounding could underlie positive

associations, and many causes of dementia are unknown. In particular, primary care

records do not hold detailed lifestyle or demographic information, but do include detailed

Page 17 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Onlyrecords of diagnoses and symptoms that have allowed good confounding control. With

respect to study design: our estimates generalise to those surviving at least four years after

their initial anticholinergic exposure, and the nested-case control design provides unbiased

estimates of effects when compared to the equivalent cohort study (29).

Strengths and weaknesses in relation to other studies

Many studies have linked anticholinergic medication use with concurrent or short term

cognitive effects (8). Few studies have examined associations of long term anticholinergic

exposure with future cognitive decline or dementia incidence, but these tend to report

positive associations (12,14,15). Our study findings are consistent with these studies. In

particular a US cohort study of 3,434 participants followed over an average of 7 years,

similar effect sizes were found to our study (12). Anticholinergic antidepressants and other

anticholinergics were linked to dementia incidence, consistent with our results, however the

authors were not able to test specific classes of non-antidepressants. A US case-control

study of 141,940 nursing home residents with depression (15), found a greater adjusted

odds ratio of 1.24 (1.20 to 1.28) for anticholinergic medication use in the 90 days before

dementia diagnosis based on administrative data. Our analysis builds on these studies using

a longer history in a larger sample of patients enabling a further disaggregation of the

effects of specific classes in time windows prior to diagnosis. Although set in the UK our

findings are likely to generalizable to other developed countries.

Page 18 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review OnlyMeaning of the study: possible explanations and implications for clinicians and policymakers

Possible explanations for our findings are that other actions of specific groups of

anticholinergics may underlie observed effects, or that the medications are markers of

prodromal symptoms or dementia risk factors. Alternatively the class-specific association

we have observed may reflect a difference in the ability of different groups of

anticholinergics to cross the blood-brain barrier.

Mechanistic evidence for a link between anticholingerics and dementia incidence is limited,

but neuropathological studies in humans and mice do support a role of anticholinergics

affecting neurodegenerative pathology (30,31). Most recently a cross-sectional analysis of

Neuroimaging Initiative (ADNI) and Indiana Memory and Ageing Study (IMAS) data linked

anticholinergic use to reduced glucose metabolism and increased brain atrophy, and to

future mild cognitive impairment or Alzheimer’s disease incidence among cognitively

normal participants (HR=2.47, p=0.01) (32), but did not disaggregate subclasses of

medications. Evidence from anticholinergic cessation trials have not shown improvements

in cognitive function, but these have been underpowered and focussed on short term

outcomes (33).

Anticholinergic urologicals, particularly Oxybutynin, have been consistently associated with

short-term cognitive decline in RCTs (8,34), so a long-term risk of dementia is plausible.

Lower urinary tract symptoms themselves have been linked to future dementia incidence

(35) and may be a symptom of early neurodegeneration (36).

ACB1 antidepressants (mainly SSRIs) were only associated with dementia close to the time

of prescription. Conversely the finding of a constant association between dementia and

Page 19 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review OnlyACB3 antidepressants across the 20 years before dementia incidence, is more likely to

represent a causal link. Although we adjusted for depression diagnosis, severity, symptoms

and duration, residual confounding and incomplete coding cannot be excluded.

Similarly, patients with Parkinson’s disease have an elevated risk of dementia (38), yet

anticholinergic antiparkinsonian drugs have previously been associated with greater

cognitive decline in a cohort of 235 Parkinson’s disease patients (39).

Previous research linking all anticholinergics to dementia incidence may have dissuaded

some patients and clinicians from the use of certain antihistamine or gastrological

medication for which we see no long-term association with dementia. Our results, although

in need of independent confirmation should reassure these patients. Nevertheless, we do

not dispute the possible short-term harms of all anticholinergic medication use among

vulnerable groups, including short-term cognitive impairment, and would advocate current

guidance of vigilant use or avoidance among frail older people.

Methodological considerations for future research

Adjusting for time-varying confounders is difficult in a matched case-control study (24,40).

We measured covariates at the start and end of the DEP. Since there was little difference

between these estimates in most cases we are confident that the effect of time-varying

covariates does not significantly affect our findings. Secondly, to be certain that potential

confounders preceded medication initiation and not vice versa, we further excluded

patients who were ‘prevalent’ users at the start of each DEP, again with little difference in

results. We suggest that authors undertaking case-control studies comparing cumulative

Page 20 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Onlyexposure over a long time period carefully consider how time-varying covariates and

exposure prior to the period in which they are being directly measured might affect study

findings.

Conclusion

Many people use anticholinergic medications at some point in their lives, and many are

prescribed to manage chronic conditions leading to potentially long exposures. There are

clear robust associations between levels of anticholinergic antidepressants and

anticholinergic urologicals and risk of dementia diagnosis up to twenty years after exposure.

Future research in this area must consider mechanisms underlying differential effects of

individual drug classes. Carefully conducted prospective studies in specific patient cohorts

comparing the long-term cognitive effects and neuropathological correlates of specific drug

classes are now needed. Clinicians should continue to be vigilant with respect to the use of

anticholinergic medications, and in particular should consider the risk of long term cognitive

effects, as well as short-term effects, associated with specific classes when weighing their

risks and benefits.

Funding: This research was supported by funding from the Alzheimer’s Society (AS-PG-2013-

017). The funders had no role is the design of the study or the interpretation of the

findings.

Data source: This study is based in part on data from the Clinical Practice Research Datalink

obtained under licence from the UK Medicines and Healthcare products Regulatory Agency

Page 21 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Only(MHRA). However, the interpretation and conclusions contained in this report are those of

the authors alone.

Data sharing: Data from the Clinical Practice Research Datalink cannot be shared by the

authors but is available directly from CPRD. Full code lists corresponding to the outcome

and each of the covariates we included are available on request.

Ethical approval: The study was approved by the Independent Scientific Advisory

Committee for CPRD for Clinical Practice Research Datalink research (protocol No 15_056R).

No further ethical approval was required for the analysis of the data. The CPRD Group has

obtained ethical approval from a multi-centre research ethics committee for all purely

observational research using CPRD data.

Contributorship:

KR, CF, GS, PM and IM conceived and developed the initial study. KR and GS drafted the

statistical analysis plan. KR, NS, IM, CF, and NC developed the code lists. KR conducted the

statistical analysis and is the guarantor. All authors contributed to the study protocol

development and revision, the interpretation of findings and the revision of the manuscript.

Transparency statement:

KR affirms that this manuscript is an honest, accurate, and transparent account of the study

being reported; that no important aspects of the study have been omitted; and that any

discrepancies from the study as planned (and, if relevant, registered) have been explained.

Acknowledgements: We would like to thank Ms Tarita Murray-Thomas (MHRA) for

extracting the CPRD data, Mr Keshav Bajaj (Aston University) for assisting with medication

Page 22 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Onlycoding, Prof Ryan Carnahan for providing the ADS scale medications, and Mr Barry

Plumpton, Mrs Ann McLauchlan, Mrs Barbara di Vita, and Mrs Gloria Swan for providing

much appreciated assistance in interpretation and oversight as Alzheimer’s Society Research

Network Volunteers.

The Corresponding Author has the right to grant on behalf of all authors and does grant on

behalf of all authors, a worldwide licence to the Publishers and its licensees in perpetuity, in

all forms, formats and media (whether known now or created in the future), to i) publish,

reproduce, distribute, display and store the Contribution, ii) translate the Contribution into

other languages, create adaptations, reprints, include within collections and create

summaries, extracts and/or, abstracts of the Contribution, iii) create any other derivative

work(s) based on the Contribution, iv) to exploit all subsidiary rights in the Contribution, v)

the inclusion of electronic links from the Contribution to third party material where-ever it

may be located; and, vi) licence any third party to do any or all of the above.

Conflict of interest.

All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf

and declare: no support from any organisation for the submitted work beyond the Alzheimer's

Society grant listed below; IM has received personal fees for guest lectures and to support travel

from Astellas Pharmaceuticals; YL reports personal fees from Thame Pharmaceuticals, NC and CF

have received grants and personal fees from Astellas Pharmaceuticals; no other relationships or

activities that could appear to have influenced the submitted work.

Page 23 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review OnlyReferences

1. Deaths registered in England and Wales (Series DR) - Office for National Statistics [Internet].

[cited 2017 Feb 12]. Available from:

https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/b

ulletins/deathsregisteredinenglandandwalesseriesdr/2015#dementia-and-alzheimer-disease-

leading-cause-of-death-for-both-men-and-women-aged-80-and-over

2. Savva GM, Wharton SB, Ince PG, Forster G, Matthews FE, Brayne C, et al. Age, Neuropathology,

and Dementia. N Engl J Med. 2009 May 28;360(22):2302–9.

3. White LR, Edland SD, Hemmy LS, Montine KS, Zarow C, Sonnen JA, et al. Neuropathologic

comorbidity and cognitive impairment in the Nun and Honolulu-Asia Aging Studies. Neurology.

2016 Mar 15;86(11):1000–8.

4. Matthews FE, Arthur A, Barnes LE, Bond J, Jagger C, Robinson L, et al. A two-decade comparison

of prevalence of dementia in individuals aged 65 years and older from three geographical areas

of England: results of the Cognitive Function and Ageing Study I and II. The Lancet. 2013

Oct;382(9902):1405–12.

5. Winblad B, Amouyel P, Andrieu S, Ballard C, Brayne C, Brodaty H, et al. Defeating Alzheimer’s

disease and other dementias: a priority for European science and society. Lancet Neurol. 2016

Apr 1;15(5):455–532.

6. Gao L, Maidment I, Matthews FE, Robinson L, Brayne C. Medication usage change in older

people (65+) in England over 20 years: findings from CFAS I and CFAS II. Age Ageing. :1–6.

7. Prescriptions Dispensed in the Community, Statistics for England - 2005-2015 - NHS Digital

[Internet]. [cited 2017 Oct 16]. Available from: http://digital.nhs.uk/catalogue/PUB20664

8. Tannenbaum C, Paquette A, Hilmer S, Holroyd-Leduc J, Carnahan R. A Systematic Review of

Amnestic and Non-Amnestic Mild Cognitive Impairment Induced by Anticholinergic,

Antihistamine, GABAergic and Opioid Drugs. Drugs Aging. 2012;29(8):639–58.

9. National Institute for Health and Care Excellence. Urinary incontinence in women: management.

NICE guideline CG171 [Internet]. 2015 [cited 2017 Feb 10]. Available from:

https://www.nice.org.uk/guidance/cg171?unlid=

10. Fick DM, Cooper JW, Wade WE, Waller JL, Maclean JR, Beers MH. Updating the Beers Criteria for

Potentially Inappropriate Medication Use in Older Adults: Results of a US Consensus Panel of

Experts. Arch Intern Med. 2003 Dec 8;163(22):2716–24.

11. Gallagher P, O’Mahony D. STOPP (Screening Tool of Older Persons’ potentially inappropriate

Prescriptions): application to acutely ill elderly patients and comparison with Beers’ criteria. Age

Ageing. 2008 Nov;37(6):673–9.

12. Gray SL, Anderson ML, Dublin S, Hanlon JT, Hubbard R, Walker R, et al. Cumulative Use of Strong

Anticholinergic Medications and Incident Dementia. JAMA Intern Med. 2015 Mar 1;175(3):401–

7.

13. Carriere I, Fourrier-Reglat A, Dartigues J-F, Rouaud O, Pasquier F, Ritchie K, et al. Drugs With

Anticholinergic Properties, Cognitive Decline, and Dementia in an Elderly General Population:

The 3-City Study. Arch Intern Med. 2009 Jul 27;169(14):1317–24.

Page 24 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Only14. Fox C, Richardson K, Maidment ID, Savva GM, Matthews FE, Smithard D, et al. Anticholinergic

Medication Use and Cognitive Impairment in the Older Population: The Medical Research

Council Cognitive Function and Ageing Study. J Am Geriatr Soc. 2011 Aug;59(8):1477–83.

15. Chatterjee S, Bali V, Carnahan RM, Johnson ML, Chen H, Aparasu RR. Anticholinergic Medication

Use and Risk of Dementia Among Elderly Nursing Home Residents with Depression. Am J Geriatr

Psychiatry Off J Am Assoc Geriatr Psychiatry. 2016 Jun;24(6):485–95.

16. Boustani M, Campbell N, Munger S, Maidment I, Fox C. Impact of anticholinergics on the aging

brain: a review and practical application. Aging Health. 2008 Jun;4(3):311–20.

17. Cai X, Campbell N, Khan B, Callahan C, Boustani M. Long-term anticholinergic use and the aging

brain. Alzheimers Dement [Internet]. 2012 Nov [cited 2013 Jan 10]; Available from:

http://www.alzheimersanddementia.com/article/S1552-5260(12)00081-7/abstract

18. Herrett E, Gallagher AM, Bhaskaran K, Forbes H, Mathur R, van Staa T, et al. Data Resource

Profile: Clinical Practice Research Datalink (CPRD). Int J Epidemiol. 2015 Jun 1;44(3):827–36.

19. Vandenbroucke JP, Pearce N. Case–control studies: basic concepts. Int J Epidemiol. 2012 Oct

1;41(5):1480–9.

20. Salahudeen MS, Chyou T, Nishtala PS. Serum Anticholinergic Activity and Cognitive and

Functional Adverse Outcomes in Older People: A Systematic Review and Meta-Analysis of the

Literature. PLOS ONE. 2016 Mar 21;11(3):e0151084.

21. Salahudeen MS, Hilmer SN, Nishtala PS. Comparison of Anticholinergic Risk Scales and

Associations with Adverse Health Outcomes in Older People. J Am Geriatr Soc. 2015;63(1):85–

90.

22. Horwitz RI, Feinstein AR. The problem of ‘protopathic bias’ in case-control studies. Am J Med.

1980 Feb;68(2):255–8.

23. Campbell NL, Maidment I, Fox C, Khan B, Boustani M. The 2012 Update to the Anticholinergic

Cognitive Burden Scale. J Am Geriatr Soc. 2013;61:S1–S232.

24. Schneeweiss S. A basic study design for expedited safety signal evaluation based on electronic

healthcare data. Pharmacoepidemiol Drug Saf. 2010 Aug 1;19(8):858–68.

25. Ray WA. Evaluating Medication Effects Outside of Clinical Trials: New-User Designs. Am J

Epidemiol. 2003 Nov 1;158(9):915–20.

26. Carnahan RM, Lund BC, Perry PJ, Pollock BG, Culp KR. The Anticholinergic Drug Scale as a

Measure of Drug-Related Anticholinergic Burden: Associations With Serum Anticholinergic

Activity. J Clin Pharmacol. 2006;46(12):1481–6.

27. Imfeld P, Brauchli Pernus YB, Jick SS, Meier CR. Epidemiology, Co-Morbidities, and Medication

Use of Patients with Alzheimer’s Disease or Vascular Dementia in the UK. J Alzheimers Dis. 2013

Jan 1;35(3):565–73.

28. Brown A, Kirichek O, Balkwill A, Reeves G, Beral V, Sudlow C, et al. Comparison of dementia

recorded in routinely collected hospital admission data in England with dementia recorded in

primary care. Emerg Themes Epidemiol. 2016;13:11.

Page 25 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Only29. Langholz B, Richardson D. Are nested case-control studies biased? Epidemiol Camb Mass. 2009

May;20(3):321–9.

30. Yoshiyama Y, Kojima A, Itoh K, Isose S, Koide M, Hori K, et al. Does Anticholinergic Activity Affect

Neuropathology? Implication of Neuroinflammation in Alzheimer’s Disease. Neurodegener Dis.

2015;15(3):140–8.

31. Perry EK, Kilford L, Lees AJ, Burn DJ, Perry RH. Increased Alzheimer pathology in Parkinson’s

disease related to antimuscarinic drugs. Ann Neurol. 2003;54(2):235–238.

32. Risacher SL, McDonald BC, Tallman EF, West JD, Farlow MR, Unverzagt FW, et al. Association

Between Anticholinergic Medication Use and Cognition, Brain Metabolism, and Brain Atrophy in

Cognitively Normal Older Adults. JAMA Neurol. 2016 Jun 1;73(6):721–32.

33. Kersten H, Molden E, Tolo IK, Skovlund E, Engedal K, Wyller TB. Cognitive Effects of Reducing

Anticholinergic Drug Burden in a Frail Elderly Population: A Randomized Controlled Trial. J

Gerontol A Biol Sci Med Sci. 2013 Mar 1;68(3):271–8.

34. Kay GG, Ebinger U. Preserving cognitive function for patients with overactive bladder: evidence

for a differential effect with darifenacin. Int J Clin Pract. 2008 Nov;62(11):1792–800.

35. Chiang C-H, Wu M-P, Ho C-H, Weng S-F, Huang C-C, Hsieh W-T, et al. Lower Urinary Tract

Symptoms Are Associated with Increased Risk of Dementia among the Elderly: A Nationwide

Study. BioMed Res Int. 2015 Jul 28;2015:e187819.

36. Sakakibara R, Panicker J, Fowler CJ, Tateno F, Kishi M, Tsuyusaki Y, et al. Is overactive bladder a

brain disease? The pathophysiological role of cerebral white matter in the elderly. Int J Urol.

2014 Jan 1;21(1):33–8.

37. Bali V, Chatterjee S, Carnahan RM, Chen H, Johnson ML, Aparasu RR. Risk of Dementia Among

Elderly Nursing Home Patients Using Paroxetine and Other Selective Serotonin Reuptake

Inhibitors. Psychiatr Serv. 2015 Aug 3;66(12):1333–40.

38. Aarsland D, Andersen K, Larsen JP, Lolk A, Nielsen H, Kragh–Sørensen P. Risk of dementia in

Parkinson’s disease A community-based, prospective study. Neurology. 2001 Mar 27;56(6):730–

6.

39. Ehrt U, Broich K, Larsen JP, Ballard C, Aarsland D. Use of drugs with anticholinergic effect and

impact on cognition in Parkinson’s disease: a cohort study. J Neurol Neurosurg Psychiatry. 2010

Feb 1;81(2):160–5.

40. Rose S, van der Laan M. A Double Robust Approach to Causal Effects in Case-Control Studies. Am

J Epidemiol. 2014 Mar 15;179(6):663–9.

Page 26 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review OnlyTable 1 – Characteristics of 40,770 case patients with dementia and 283,933 controls from the Clinical Practice

Research Datalink. Values are numbers (percentages) unless stated otherwise

At start of drug exposure period

At end of drug exposure period

Characteristic Cases (n=40,770)

Controls

(n=283,933)

Cases

(n=40,770)

Controls

(n=283,933)

Lifestyle

Smoker - current* 5,058 12.4

34,652 12.2

4,516 11.1

30,976 10.9

Harmful alcohol use* 166 0.4

980 0.3

377 0.9

2,316 0.8

BMI*† 26.4 6.3

26.5 6.9

26.1 5.2

26.8 5.6

Cardio/cerebrovascular and related diagnoses

Diabetes 2,734 6.7

16,362 5.8

5,392 13.2

34,015 12.0

Diabetes complications 418 1.0

2,206 0.8

1,378 3.4

7,678 2.7

Hyperlipidemia 2,740 6.7

16,369 5.8

6,481 15.9

40,796 14.4

Hypertension 13,328 32.7

91,737 32.3

22,104 54.2

155,262 54.7

Stroke/Transient Ischaemic Attack 1,969 4.8

11,734 4.1

4,747 11.6

25,856 9.1

Congestive heart disease 5,279 12.9

33,947 12.0

8,388 20.6

54,615 19.2

Heart failure 819 2.0

5,921 2.1

2,140 5.2

15,961 5.6

Peripheral arterial disease 897 2.2

5,953 2.1

2,110 5.2

13,818 4.9

Atrial fibrillation 1,303 3.2

8,856 3.1

3,744 9.2

24,591 8.7

Angina 3,704 9.1

23,213 8.2

6,006 14.7

38,602 13.6

Myocardial infarction 1,820 4.5

12,051 4.2

2,908 7.1

19,602 6.9

Coronary artery operations 831 2.0

5,261 1.9

1,711 4.2

11,002 3.9

Deep vein thrombosis 552 1.4

4,172 1.5

1,110 2.7

8,101 2.9

Mental health diagnoses

Depression 5,071 12.4

29,676 10.5

8,030 19.7

44,264 15.6

Severe depression 400 1.0

2,175 0.8

804 2.0

4,056 1.4

Depression symptoms 1,393 3.4

7,880 2.8

4,577 11.2

24,220 8.5

Depression dura[on‡, years† 13.8 12.8

14.2 12.8

14.3 13.0

15.2 13.2

Severe mental illness 294 0.7

1,558 0.5

431 1.1

2,078 0.7

Anxiety 3,641 8.9

22,229 7.8

6,190 15.2

35,598 12.5

Anxiety symptoms 1,172 2.9

6,457 2.3

3,763 9.2

20,366 7.2

Other diagnoses

Parkinson's disease 237 0.6

958 0.3

1,448 3.6

4,283 1.5

Epilepsy 542 1.3

2,799 1.0

771 1.9

3,860 1.4

Insomnia 2,940 7.2

18,407 6.5

7,351 18.0

45,227 15.9

Fatigue 2,666 6.5

15,825 5.6

8,561 21.0

50,351 17.7

Other sleep problems 583 1.4

3,501 1.2

2,772 6.8

15,878 5.6

Hemiplegia and paraplegia 131 0.3

816 0.3

235 0.6

1,406 0.5

Drug abuse 297 0.7

1,694 0.6

414 1.0

2,357 0.8

Migraine 1,320 3.2

8,625 3.0

1,839 4.5

11,999 4.2

Headache 2,540 6.2

15,653 5.5

6,102 15.0

37,602 13.2

Back/neck pain 12,225 30.0

79,108 27.9

21,123 51.8

138,554 48.8

Neuropathy 750 1.8

4,766 1.7

2,030 5.0

13,098 4.6

Meniere's disease 340 0.8

2,353 0.8

471 1.2

3,397 1.2

Restless legs syndrome 169 0.4

1,088 0.4

610 1.5

3,955 1.4

Chronic obstructive pulmonary disease 899 2.2

6,327 2.2

2,534 6.2

18,012 6.3

Asthma 3,134 7.7

20,991 7.4

4,375 10.7

29,822 10.5

Rhinitis 2,556 6.3

16,617 5.9

5,155 12.6

32,956 11.6

GERD/Oesophagitis 2,876 7.1

18,277 6.4

6,141 15.1

39,437 13.9

Peptic/gastric ulcer 1,122 2.8

6,779 2.4

2,530 6.2

15,724 5.5

Irritable bowel syndrome 210 0.5

1,316 0.5

2,328 5.7

15,277 5.4

Inflammatory Bowel disease 1,087 2.7

6,692 2.4

2,451 6.0

15,523 5.5

Intestinal surgery/Colostomy/ileostomy 1,438 3.5

9,515 3.4

3,137 7.7

20,215 7.1

Liver disease 93 0.2

650 0.2

207 0.6

1,312 0.5

Osteoarthritis 7,925 19.4

50,644 17.8

14,627 35.9

94,681 33.3

Rheumatoid arthritis 486 1.2

3,700 1.3

806 2.0

5,799 2.0

Page 27 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Dermatitis 6,986 17.1

45,613 16.1 13,364 32.8

88,331 31.1

Eczema 4,015 9.8

26,696 9.4 8,952 22.0

59,677 21.0

Psoriasis 1,059 2.6

6,856 2.4 1,696 4.2

11,200 3.9

Urinary incontinence 1,222 3.0

7,186 2.5

3,225 7.9

18,027 6.3

Renal disease/ chronic kidney disease 457 1.1

3,138 1.1 5,817 14.3

37,981 13.4

Prostatism 1,496 3.7

9,176 3.2 3,490 8.6

21,464 7.6

Cancer 3,133 7.7

21,956 7.7 6,307 15.5

44,567 15.7

History in 12 months before DEP

Any falls 808 2.0

4,391 1.6 808 2.0

4,391 1.6

Any fractures 596 1.5

3,693 1.3 596 1.5

3,693 1.3

Physician consulta[ons† 5.4 6.3 4.8 5.9 5.4 6.3 4.8 5.9

Abbreviations: BMI, body mass index; DEP, drug exposure period; GERD, Gastroesophageal reflux disease

* Missing smoking data for 9,664 cases and 70,877 controls at the start of the DEP, and 1,465 cases and 13,596

controls at the end of the DEP

Missing BMI data for 13,086 cases and 94,734 controls at the start of the DEP, and 4,736 cases and 37,526 controls

at the end of the DEP

Missing alcohol data for 22,940 cases and 159,721 controls at the start of the DEP, and 12,702 cases and 90,129

controls at the end of the DEP

† mean (SD)

‡ years since first depression symptom or depression diagnosis

Page 28 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Page 29 of 34

Table 2. Crude and adjusted odds ratios of dementia by prescription of any, DDDs, and total burden of anticholinergics according to the Anticholinergic

Cognitive Burden score

Exposure Cases (n=40,770) Controls (n=283,933) Unadjusted

Adjusted for covariates

measured at start of DEP


measured at end of DEP

during DEP n % n % OR 95% CI aORa 95% CI aOR

b 95% CI

Any use

No prescriptions 4295 10.5% 36329 12.8% 1.00 1.00

ACB 1 prescription 36437 89.4% 247406 87.1% 1.25* 1.21 to 1.29 1.11* 1.07 to 1.15 1.10* 1.06 to 1.15



Number of DDDs

ACB 1

0 4333 10.6% 36527 12.9% 1.00 1.00 1.00

>0-13 3000 7.4% 20530 7.2% 1.24* 1.18 to 1.30 1.16* 1.11 to 1.22 1.14* 1.08 to 1.20

14-89 2530 6.2% 17088 6.0% 1.25* 1.19 to 1.32 1.15* 1.09 to 1.21 1.13* 1.07 to 1.19

90-364 4253 10.4% 28497 10.0% 1.26* 1.21 to 1.32 1.12* 1.07 to 1.18 1.09* 1.04 to 1.15

365-1459 8549 21.0% 56607 19.9% 1.28* 1.23 to 1.33 1.12* 1.07 to 1.17 1.10* 1.05 to 1.15

1460+ 18105 44.4% 124684 43.9% 1.23* 1.19 to 1.28 1.05 1.01 to 1.10 1.05 1.01 to 1.10

ACB 2

0 39341 96.5% 276024 97.2% 1.00 1.00 1.00

>0-13 704 1.7% 4301 1.5% 1.15* 1.06 to 1.25 1.01 0.93 to 1.10 1.02 0.94 to 1.11

14-89 266 0.7% 1533 0.5% 1.22* 1.07 to 1.39 1.08 0.94 to 1.23 1.07 0.94 to 1.23

90-364 218 0.5% 1035 0.4% 1.48* 1.28 to 1.72 1.23* 1.06 to 1.43 1.20 1.03 to 1.40

365-1459 176 0.4% 826 0.3% 1.50* 1.27 to 1.76 1.23 1.04 to 1.45 1.18 0.99 to 1.39

1460+ 65 0.2% 214 0.1% 2.15* 1.63 to 2.83 1.62* 1.22 to 2.15 1.57* 1.18 to 2.09

ACB 3

0 26317 64.5% 197530 69.6% 1.00 1.00 1.00

>0-13 5663 13.9% 38084 13.4% 1.13* 1.10 to 1.17 1.07* 1.04 to 1.11 1.04 1.01 to 1.08

14-89 2440 6.0% 15154 5.3% 1.23* 1.17 to 1.28 1.13* 1.08 to 1.18 1.07* 1.02 to 1.12

90-364 2786 6.8% 15462 5.4% 1.37* 1.31 to 1.43 1.24* 1.19 to 1.30 1.17* 1.12 to 1.22

365-1459 2512 6.2% 12626 4.4% 1.51* 1.45 to 1.58 1.35* 1.29 to 1.42 1.27* 1.21 to 1.33

1460+ 1052 2.6% 5077 1.8% 1.59* 1.49 to 1.71 1.40* 1.30 to 1.50 1.31* 1.22 to 1.41

Average ACB burden

<0.1 9517 23.3% 74445 26.2% 1.00 1.00 1.00

0.1-<0.5 6084 14.9% 41528 14.6% 1.15* 1.11 to 1.19 1.09* 1.05,1.13 1.07* 1.03 to 1.11

Page 29 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Page 30 of 34

0.5-<1 5781 14.2% 40413 14.2% 1.12* 1.08 to 1.16 1.06* 1.02 to 1.10 1.06* 1.02 to 1.10

1-<2 9030 22.2% 62348 22.0% 1.14* 1.10 to 1.17 1.06* 1.03 to 1.10 1.07* 1.01 to 1.09

2-<3 4914 12.1% 33113 11.7% 1.17* 1.13 to 1.21 1.07* 1.03 to 1.12 1.07* 1.02 to 1.12

3-<5 3874 9.5% 23954 8.4% 1.28* 1.23 to 1.33 1.13* 1.08 to 1.19 1.12* 1.07 to 1.18

5+ 1570 3.9% 8132 2.9% 1.53* 1.44 to 1.62 1.31* 1.23 to 1.40 1.28* 1.19 to 1.36

Abbreviations: DEP, drug exposure period; DDD, defined daily dose; OR, odds ratio, aOR, adjusted odds ratio; CI confidence interval; ACB Anticholinergic

Cognitive Burden

a Adjusted for age, region, any falls, any fractures and number of GP consultations in the 12 months prior to the DEP; number of prescriptions during

the DEP for the following medications not rated as anticholinergic: benzodiazepines, z-drugs, antidepressants, antinausea and antivertigo

preparations, antiepileptics, and anti-Parkinson drugs; and the following variables measured at the start of the DEP: BMI, smoking status, harmful

alcohol use, depression duration (0, 0-2, 2-5, 5-10, 10-20, >20 years), and all diagnoses listed in table 1.

b Adjusted for age, region, any falls, any fractures and number of GP consultations in the 12 months prior to the DEP; number of prescriptions during

the DEP for the following medications not rated as anticholinergic: benzodiazepines, z-drugs, antidepressants, antinausea and antivertigo

preparations, antiepileptics, and anti-Parkinson drugs; and the following variables measured at the end of the DEP: BMI, smoking status, harmful

alcohol use, depression duration (0, 0-5, 5-10, 10-15, 15-20, >20 years), and all diagnoses listed in table 1.

* p-value < 0.01

Page 30 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Page 31 of 34

Table 3. Adjusted odds ratios of dementia by any prescription of an anticholinergic medication by

Anticholinergic Cognitive Burden score and drug class

Drug class

Cases (n=40,770) Controls (n=283,933)


measured at start of

DEP



n % n % aORa 95% CI aOR

b 95% CI

ACB1

Analgesic 23,871 58.6% 158,162 55.7% 1.02 1.00 to 1.05 1.02 0.99 to 1.04

Antidepressant 5,958 14.6% 28,767 10.1% 1.37* 1.32 to 1.42 1.25* 1.20 to 1.30

Antipsychotic 8,051 19.7% 50,079 17.6% 1.05* 1.02 to 1.08 1.04 1.01 to 1.07

Cardiovascular 27,926 68.5% 191,895 67.6% 0.97 0.94 to 0.99 0.98 0.95 to 1.01

Gastro-intestinal 10,845 26.6% 71,814 25.3% 0.97 0.94 to 0.99 0.96* 0.93 to 0.99

Respiratory 9,385 23.0% 62,787 22.1% 0.99 0.97 to 1.02 0.99 0.97 to 1.02

Other 11,521 28.3% 77,345 27.2% 0.95* 0.92 to 0.97 0.95* 0.92 to 0.98

ACB2

Analgesic 385 0.9% 2,337 0.8% 1.03 0.92 to 1.15 1.03 0.92 to 1.16

Antipsychotic 22 0.1% 69 0.0% 1.44 0.87 to 2.36 1.35 0.82 to 2.23

Parkinsonian 57 0.1% 141 0.0% 1.55* 1.12 to 2.14 1.32 0.96 to 1.82

Respiratory 19 0.0% 123 0.0% 0.89 0.55 to 1.45 0.83 0.51 to 1.36

Other 985 2.4% 5,454 1.9% 1.07 1.00 to 1.15 1.09 1.01 to 1.17

ACB3


Antipsychotic 1,036 2.5% 5,140 1.8% 1.09 1.02 to 1.18 1.07 1.00 to 1.16

Gastro-intestinal 1,817 4.5% 12,057 4.2% 0.94 0.89 to 0.99 0.94 0.89 to 0.99

Parkinsonian 270 0.7% 951 0.3% 1.45* 1.25 to 1.68 1.29* 1.11 to 1.50


Urological 3,261 8.0% 16,873 5.9% 1.23* 1.18 to 1.28 1.18* 1.13 to 1.23

Other 284 0.7% 1,741 0.6% 0.99 0.87 to 1.13 0.99 0.87 to 1.13

Abbreviations: DEP, drug exposure period; aOR, adjusted odds ratio; CI confidence interval; ACB

Anticholinergic cognitive burden

Page 31 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Page 32 of 34

a Adjusted for age, region, any falls, any fractures and number of GP consultations in the 12

months prior to the DEP; number of prescriptions during the DEP for the following medications

not rated as anticholinergic: benzodiazepines, z-drugs, antidepressants, antinausea and

antivertigo preparations, antiepileptics, and anti-Parkinson drugs; and the following variables

measured at the start of the DEP: BMI, smoking status, harmful alcohol use, depression duration

(0, 0-2, 2-5, 5-10, 10-20, >20 years), and all diagnoses listed in table 1.

b Adjusted for age, region, any falls, any fractures and number of GP consultations in the 12

months prior to the DEP; number of prescriptions during the DEP for the following medications

not rated as anticholinergic: benzodiazepines, z-drugs, antidepressants, antinausea and

antivertigo preparations, antiepileptics, and anti-Parkinson drugs; and the following variables

measured at the end of the DEP: BMI, smoking status, harmful alcohol use, depression duration

(0, 0-5, 5-10, 10-15, 15-20, >20 years), and all diagnoses listed in table 1.

* p-value < 0.01

Page 32 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Page 33 of 34

Table 4. Adjusted odds ratios of dementia by prescription of anticholinergic medications, by drug class, and by time period before dementia diagnosis

Exposure period

15-20 years prior to index datea 10-15 years prior to index date

b 4-10 years prior to index date

c

Drug class

No. of

cases

No. of

controls

No. of

cases

No. of

controls

No. of

cases

No. of

controls

(N=10,684) (N=74,145) ORd 95% CI

(N=23,959) (N=166,735) ORd 95% CI

(N=40,770) (N=283,933) ORd 95% CI

Any use by ACB score

No prescriptions 3,638 27,905 1.00 5,602 44,790 1.00 4,492 38,579 1.00

ACB 1 prescription 6,789 44,564 1.05 1.00 to 1.10 17,867 118,973 1.06* 1.02 to 1.10 35,722 242,210 1.06* 1.02 to 1.09

ACB 2 prescription 193 1,057 1.07 0.91 to 1.25 493 2,556 1.14* 1.03 to 1.26 1,054 5,734 1.11* 1.03 to 1.18

ACB 3 prescription 1,972 11,321 1.17* 1.10 to 1.24 5,242 30,303 1.15* 1.10 to 1.19 12,338 72,335 1.13* 1.10 to 1.15

Any use by ACB score and class

No prescription 3,638 27,905 1.00

5,602 44,790 1.00

4,492 38,579 1.00

ACB1

Analgesic 3,638 22,914 1.06 1.01 to 1.11 9,771 62,234 1.05* 1.02 to 1.08 21,756 143,993 0.99 0.96 to 1.01

Antidepressant 262 1,429 1.08 0.93 to 1.24 3,813 24,345 1.18* 1.11 to 1.26 5,413 25,566 1.37* 1.32 to 1.42

Antipsychotic 932 6,032 0.99 0.92 to 1.07 2,507 15,429 1.04 0.99 to 1.09 6,392 39,642 1.03 1.00 to 1.07

Cardiovascular 3,672 23,851 1.02 0.97 to 1.08 11,785 79,119 0.99 0.95 to 1.02 27,256 187,816 0.95* 0.92 to 0.97

Gastro-intestinal 1,863 12,023 0.97 0.92 to 1.03

4,447 28,709 0.98 0.94 to 1.02

8,404 55,944 0.95* 0.92 to 0.98

Respiratory 990 6,658 0.96 0.89 to 1.04 2,913 18,767 1.01 0.97 to 1.06 7,784 52,560 0.98 0.95 to 1.01

Other 1,230 8,230 0.94 0.88 to 1.02 1,396 7,091 0.98 0.94 to 1.02 9,792 66,244 0.91* 0.89 to 0.94

ACB2

Analgesic 64 358 1.09 0.83 to 1.43 102 654 0.92 0.75 to 1.14 254 1,518 1.05 0.91 to 1.20

Antipsychotic 8 16 NA

10 28 NA

14 48 NA

Parkinsonian <5 12 NA 17 22 NA 47 116 1.59 1.11 to 2.28

Respiratory <5 26 NA 8 35 NA 12 75 NA

Other 120 675 0.99 0.80 to 1.21 363 1,873 1.12 1.00 to 1.26 745 4,056 1.07 0.99 to 1.17

ACB3

Antidepressant 1,155 6,302 1.19* 1.10 to 1.29

3,237 17,828 1.16* 1.11 to 1.22

7,342 41,826 1.12* 1.08 to 1.15

Antipsychotic 189 952 1.20 1.00 to 1.42 377 2,024 0.98 0.86 to 1.10 781 3,679 1.14* 1.04 to 1.23

Page 33 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Page 34 of 34

Gastro-intestinal 273 1,568 1.11 0.97 to 1.28 522 3,453 0.93 0.84 to 1.02 1,283 8,721 0.91* 0.85 to 0.97

Parkinsonian 29 146 1.29 0.83 to 2.01

104 379 1.54* 1.22 to 1.96

230 786 1.56* 1.33 to 1.83

Respiratory 524 3,247 1.07 0.97 to 1.18 1,228 7,615 1.04 0.97 to 1.10 2,839 17,807 1.03 0.99 to 1.08

Urological 207 995 1.27* 1.09 to 1.48 810 4,139 1.22* 1.13 to 1.32 2,870 14,654 1.23* 1.18 to 1.29

Other 43 281 0.97 0.70 to 1.34 93 569 0.97 0.78 to 1.21 176 1,051 1.02 0.86 to 1.20

Abbreviations: OR, odds ratio, aOR, adjusted odds ratio; CI confidence interval; ACB Anticholinergic cognitive burden

a Including patients with ≥16 years of UTS data history before the index date

b Including patients with ≥11 years of UTS data history before the index date

c Including patients with ≥6 years of UTS data history before the index date

d Adjusted for age, region, number of prescriptions during the exposure period for the following medications not rated as anticholinergic:

benzodiazepines, z-drugs, antidepressants, antinausea and antivertigo preparations, antiepileptics, and anti-Parkinson drugs; and the following

variables measured at the start of the exposure period: any falls, any fractures and number of GP consultations in the 12 months prior to the

exposure period: BMI, smoking status, harmful alcohol use, depression duration (0, 0-2, 2-5, 5-10, 10-20, >20 years), and all diagnoses listed in table

1.

* p-value < 0.01

Page 34 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Supplementary Material

Appendix

CPRD medcodes, corresponding Read codes and Read terms used to define dementia

medcode read_code read_term

1350 E00..12 Senile/presenile dementia

1916 E00..11 Senile dementia

1917 F110.00 Alzheimer's disease

2882 E00z.00 Senile or presenile psychoses NOS

4357 Eu02z14 [X] Senile dementia NOS

4693 Eu02z00 [X] Unspecified dementia

6578 Eu01.00 [X]Vascular dementia

7323 E000.00 Uncomplicated senile dementia

7572 F116.00 Lewy body disease

7664 Eu00.00 [X]Dementia in Alzheimer's disease

8195 Eu00z11 [X]Alzheimer's dementia unspec

8634 E004.11 Multi infarct dementia

8934 Eu01200 [X]Subcortical vascular dementia

9509 Eu02300 [X]Dementia in Parkinson's disease

9565 Eu01.11 [X]Arteriosclerotic dementia

11136 F111.00 Pick's disease

11175 Eu01100 [X]Multi-infarct dementia

11379 Eu00112 [X]Senile dementia,Alzheimer's type

12621 Eu02.00 [X]Dementia in other diseases classified elsewhere

15165 E001.00 Presenile dementia

15249 E00y.00 Other senile and presenile organic psychoses

16797 F110000 Alzheimer's disease with early onset

18386 E002000 Senile dementia with paranoia

19393 Eu01z00 [X]Vascular dementia, unspecified

19477 E004.00 Arteriosclerotic dementia

21887 E002100 Senile dementia with depression

25386 E041.00 Dementia in conditions EC

25704 Eu00011 [X]Presenile dementia,Alzheimer's type

26270 Eu02500 [X]Lewy body dementia

27677 E001300 Presenile dementia with depression

27759 Eu02z16 [X] Senile dementia, depressed or paranoid type

27935 Eu02z15 [X] Senile psychosis NOS

28402 Eu02000 [X]Dementia in Pick's disease

29386 Eu00z00 [X]Dementia in Alzheimer's disease, unspecified

29512 F112.00 Senile degeneration of brain

30032 E001200 Presenile dementia with paranoia

30706 Eu00200 [X]Dementia in Alzheimer's dis, atypical or mixed type

31016 Eu01300 [X]Mixed cortical and subcortical vascular dementia

32057 F110100 Alzheimer's disease with late onset

33707 E00..00 Senile and presenile organic psychotic conditions

34944 Eu02z13 [X] Primary degenerative dementia NOS

Page 35 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


37014 Eu02200 [X]Dementia in Huntington's disease

37015 E003.00 Senile dementia with delirium

38438 E001z00 Presenile dementia NOS

38678 Eu00100 [X]Dementia in Alzheimer's disease with late onset

41089 E002z00 Senile dementia with depressive or paranoid features NOS

42279 E004z00 Arteriosclerotic dementia NOS

42602 E001000 Uncomplicated presenile dementia

43089 E004000 Uncomplicated arteriosclerotic dementia

43292 E004300 Arteriosclerotic dementia with depression

43346 Eu00113 [X]Primary degen dementia of Alzheimer's type, senile onset

44674 E002.00 Senile dementia with depressive or paranoid features

46488 Eu01000 [X]Vascular dementia of acute onset

46762 Eu00111 [X]Alzheimer's disease type 1

47619 Eu02z12 [X] Presenile psychosis NOS

48501 Eu02z11 [X] Presenile dementia NOS

49263 Eu00000 [X]Dementia in Alzheimer's disease with early onset

49513 E001100 Presenile dementia with delirium

51494 E00y.11 Presbyophrenic psychosis

53446 Eu04100 [X]Delirium superimposed on dementia

54106 Eu02100 [X]Dementia in Creutzfeldt-Jakob disease

55313 Eu01y00 [X]Other vascular dementia

55467 E004200 Arteriosclerotic dementia with paranoia

55838 Eu01111 [X]Predominantly cortical dementia

56912 E004100 Arteriosclerotic dementia with delirium

59122 Fyu3000 [X]Other Alzheimer's disease

60059 Eu00012 [X]Primary degen dementia, Alzheimer's type, presenile onset

61528 Eu00013 [X]Alzheimer's disease type 2

62132 E02y100 Drug-induced dementia

64267 Eu02y00 [X]Dementia in other specified diseases classif elsewhere

109288 A411000

Page 36 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Definition of covariates

We extracted and coded the following covariates at the start and end of each patient’s Drug Exposure

Period (DEP), unless stated otherwise, from the patient’s record:

Diagnosis of cardiovascular and related diseases: any previous diagnosis of diabetes, diabetes

complications, hyperlipidemia, hypertension, stroke/ Transient Ischaemic Attack, congestive heart disease,

heart failure, peripheral arterial disease, atrial fibrillation, angina, myocardial infarction, coronary artery

operations, deep vein thrombosis.

Other potential dementia risk factors or correlates: depression (diagnosis, maximum severity, symptoms

and duration), anxiety, anxiety symptoms, Parkinson's disease, severe mental illness, epilepsy, drug abuse,

cancer, insomnia, sleep problems, migraine, headache, pain, neuropathy, Meniere's disease, restless leg

syndrome

Other indications for anticholinergics: Chronic obstructive pulmonary disease, asthma, rhinitis,

Gastroesophageal reflux disease, peptic/gastric ulcer, Irritable bowel syndrome, Inflammatory Bowel

disease, intestinal surgery, liver disease, arthritis, dermatitis, eczema, psoriasis, incontinence, chronic

kidney disease, or prostatism.

Other factors: GP practice region (Scotland, Northern Ireland, Wales and 10 health regions of England),

smoking status (never, former, current or unknown), body mass index (<20, 20-24.9, 25-29.9, 30+ kg/m2,

or missing), and problematic alcohol use (categorised as more than 49 units per week for men and more

than 35 units per week for women, or missing). In the 12 months prior to the DEP: any falls, any fractures,

and number of GP consultations.

Other medication: Number of prescriptions during the DEP of other medications rated without

anticholinergic activity from the following classes: benzodiazepines (ATC N05BA, N05CD, N03AE), z-drugs

(ATC N05CF), antidepressants (ATC N06A), antinausea and antivertigo preparations (ATC N07C, A03F),

antiepileptics (ATC N03), and anti-Parkinson drugs (ATC N04).

Page 37 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


eFigure1 – Selection of 40,770 patients with dementia and 283,933 controls from the Clinical Practice Research

Datalink.

Page 38 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


eTable 1. Frequency* of prescriptions during the drug exposure period according to Anticholinergic Cognitive Burden score and drug class

ACB1 (25,122,902 prescriptions)

ACB2 (132,731 prescriptions)

ACB3 (1,793,505 prescriptions)

Class Drug % Drug % Drug %

Antidepressant Citalopram 1%

Amitriptyline 29%

Dosulepin 16%

Paroxetine 8%

Lofepramine 2%

Clomipramine 2%

Imipramine 2%

Trimipramine 2%

Nortriptyline 1%

Gastrointestinal Ranitidine 2%

Hyoscine 1%

Dicycloverine 1%

Atropine 1%

Respiratory

Chlorphenamine 4%

Hydroxyzine 2%

Promethazine 1%

Antipsychotic Prochlorperazine 1%

Zuclopenthixol 1%

Trifluoperazine 2%

Olanzapine 2%

Chlorpromazine 1%

Thioridazine 1%

Parkinsonian Amantadine 2% Procyclidine 1%

Urological

Tolterodine 7%

Oxybutynin 7%

Solifenacin 1%

Analgesic CODEINE 6%

Nefopam 6%

Dextropropoxyphene 4%

Pethidine 2%

Dihydrocodeine 3%

Tramadol 1%

Cardiovascular Bendroflumethiazide 13%

Atenolol 11%

Furosemide 6%

Ramipril 5%

Lisinopril 4%

Page 39 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Isosorbide 4%

Warfarin 4%

Nifedipine 3%

Doxazosin 3%

Digoxin 2%

Enalapril 2%

Perindopril 2%

Amiloride 2%

Glyceryl Trinitrate 2%

Other Prednisolone 2%

Carbamazepine 87%

Diazepam 1%

* Percentage of total prescriptions in the drug exposure period within each Anticholinergic Cognitive Burden score category with prevalence of 1% or more

ACB, Anticholinergic Cognitive Burden

Page 40 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


eTable 2. Adjusted odds ratios of dementia by number of DDDs of anticholinergic medications by drug class and ACB score

Drug class

Cases (n=40,770) Controls (283,933)


measured at start of DEP




b 95% CI

ACB1

Analgesic

0 16,899 41.4% 125,771 44.3% 1.00 1.00

>0-13 6,581 16.1% 45,258 15.9% 1.03 1.00 to 1.07 1.02 0.99 to 1.05

14-89 8,166 20.0% 55,924 19.7% 1.00 0.97 to 1.03 1.00 0.97 to 1.03

90-364 5,216 12.8% 33,483 11.8% 1.03 0.99 to 1.07 1.04 1.00 to 1.08

365-1459 3,293 8.1% 20,172 7.1% 1.04 1.00 to 1.09 1.06 1.01 to 1.11

1460+ 615 1.5% 3,325 1.2% 1.12 1.02 to 1.23 1.16* 1.05 to 1.27

Antidepressant

0 34,812 85.4% 255,166 89.9% 1.00 1.00

>0-13 114 0.3% 713 0.3% 1.05 0.86 to 1.28 0.99 0.81 to 1.21

14-89 2,103 5.2% 10,967 3.9% 1.29* 1.23 to 1.36 1.19* 1.13 to 1.25

90-364 1,481 3.6% 7,222 2.5% 1.36* 1.28 to 1.44 1.23* 1.16 to 1.31

365-1459 1,552 3.8% 6,883 2.4% 1.47* 1.39 to 1.57 1.34* 1.26 to 1.43

1460+ 708 1.7% 2,982 1.1% 1.55* 1.42 to 1.69 1.47* 1.34 to 1.60

Antipsychotic

0 32,719 80.3% 233,854 82.4% 1.00 1.00

>0-13 6,453 15.8% 41,234 14.5% 1.04 1.01 to 1.07 1.03 1.00 to 1.06

14-89 1,102 2.7% 6,195 2.2% 1.10* 1.02 to 1.17 1.09 1.01 to 1.16

90-364 359 0.9% 1,977 0.7% 1.09 0.97 to 1.23 1.07 0.95 to 1.20

365-1459 108 0.3% 542 0.2% 1.11 0.90 to 1.38 1.09 0.88 to 1.36

1460+ 29 0.1% 131 0.0% 1.04 0.68 to 1.59 1.03 0.67 to 1.57

Cardiovascular

0 12,844 31.5% 92,038 32.4% 1.00 1.00

>0-13 713 1.7% 4,550 1.6% 1.00 0.92 to 1.09 0.99 0.91 to 1.07

14-89 1,863 4.6% 11,912 4.2% 1.03 0.98 to 1.09 1.03 0.97 to 1.09

90-364 2,625 6.4% 17,242 6.1% 1.01 0.96 to 1.06 1.01 0.96 to 1.06

365-1459 6,802 16.7% 46,285 16.3% 0.98 0.94 to 1.01 0.98 0.95 to 1.02

1460+ 15,923 39.1% 111,906 39.4% 0.93* 0.90 to 0.96 0.95* 0.91 to 0.99

Gastro-intestinal

0 29,925 73.4% 212,119 74.7% 1.00

1.00

Page 41 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


>0-13 2,995 7.3% 19,201 6.8% 1.03 0.98 to 1.07 1.01 0.97 to 1.06

14-89 3,592 8.8% 24,251 8.5% 0.95* 0.91 to 0.99 0.94* 0.90 to 0.98

90-364 2,216 5.4% 15,138 5.3% 0.92* 0.88 to 0.97 0.92* 0.87 to 0.96

365-1459 1,607 3.9% 10,409 3.7% 0.96 0.91 to 1.01 0.96 0.90 to 1.01

1460+ 435 1.1% 2,815 1.0% 0.94 0.85 to 1.05 0.95 0.86 to 1.06

Respiratory

0 31,385 77.0% 221,146 77.9% 1.00

1.00

>0-13 1,507 3.7% 10,163 3.6% 1.01 0.95 to 1.07 1.02 0.96 to 1.08

14-89 4,702 11.5% 31,037 10.9% 1.01 0.98 to 1.05 1.00 0.97 to 1.04

90-364 1,515 3.7% 10,357 3.6% 0.95 0.89 to 1.00 0.95 0.89 to 1.00

365-1459 1,095 2.7% 7,363 2.6% 0.97 0.91 to 1.04 0.98 0.92 to 1.05

1460+ 566 1.4% 3,867 1.4% 0.94 0.86 to 1.04 0.96 0.87 to 1.05

Other

0 29,249 71.7% 206,588 72.8% 1.00 1.00

>0-13 2,653 6.5% 17,447 6.1% 0.98 0.93 to 1.03 0.97 0.93 to 1.02

14-89 5,013 12.3% 33,590 11.8% 0.95* 0.92 to 0.99 0.95 0.92 to 0.99

90-364 2,182 5.4% 14,688 5.2% 0.93* 0.88 to 0.98 0.94 0.89 to 0.99

365-1459 1,362 3.3% 9,441 3.3% 0.92* 0.86 to 0.97 0.91* 0.86 to 0.97

1460+ 311 0.8% 2,179 0.8% 0.86 0.76 to 0.97 0.86 0.76 to 0.98

ACB2

Analgesic

0 40,385 99.1% 281,596 99.2% 1.00

1.00

>0-13 116 0.3% 832 0.3% 0.89 0.73 to 1.08 0.89 0.73 to 1.09

14-89 208 0.5% 1,189 0.4% 1.10 0.94 to 1.27 1.10 0.95 to 1.28

90-364 39 0.1% 208 0.1% 1.13 0.79 to 1.59 1.12 0.79 to 1.59

365+ 22 0.1% 108 0.0% 1.21 0.76 to 1.93 1.21 0.76 to 1.93

Antipsychotic

0 40,748 99.9% 283,864 100.0% 1.00 1.00

>0-89 6 0.0% 24 0.0% 1.11 0.45 to 2.75 0.92 0.36 to 2.33

90+ 16 0.0% 45 0.0% 1.62 0.90 to 2.92 1.61 0.89 to 2.91

Parkinsonian

0 40,713 99.9% 283,792 100.0% 1.00

1.00

>0-89 20 0.0% 63 0.0% 1.43 0.85 to 2.40 1.33 0.79 to 2.23

90-364 19 0.0% 37 0.0% 1.80 1.02 to 3.17 1.47 0.84 to 2.59

365+ 18 0.0% 41 0.0% 1.48 0.84 to 2.62 1.18 0.67 to 2.08

Respiratory

Page 42 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


0 40,751 100.0% 283,810 100.0% 1.00 1.00

>0-13 6 0.0% 41 0.0% 0.89 0.37 to 2.11 0.80 0.34 to 1.93

14+ 13 0.0% 82 0.0% 0.89 0.49 to 1.61 0.85 0.47 to 1.53

Other

0 39,785 97.6% 278,479 98.1% 1.00 1.00

>0-13 424 1.0% 2,499 0.9% 1.02 0.92 to 1.14 1.05 0.94 to 1.17

14-89 218 0.5% 1,338 0.5% 0.97 0.84 to 1.12 0.99 0.85 to 1.15

90-364 152 0.4% 773 0.3% 1.17 0.98 to 1.39 1.17 0.97 to 1.40

365-1459 143 0.4% 674 0.2% 1.26 1.05 to 1.52 1.23 1.01 to 1.48

1460+ 48 0.1% 170 0.1% 1.58* 1.14 to 2.20 1.56* 1.12 to 2.17

ACB3

Antidepressant

0 31,947 78.4% 233,116 82.1% 1.00 1.00

>0-13 3,389 8.3% 21,235 7.5% 1.08* 1.04 to 1.13 1.07* 1.02 to 1.11

14-89 1,505 3.7% 8,976 3.2% 1.07 1.01 to 1.14 1.05 0.99 to 1.12

90-364 1,775 4.4% 9,682 3.4% 1.16* 1.10 to 1.23 1.14* 1.08 to 1.21

365-1459 1,498 3.7% 7,732 2.7% 1.24* 1.17 to 1.31 1.22* 1.14 to 1.29

1460+ 656 1.6% 3,192 1.1% 1.30* 1.18 to 1.42 1.29* 1.18 to 1.41

Antipsychotic

0 39,734 97.5% 278,793 98.2% 1.00 1.00

>0-13 413 1.0% 2,356 0.8% 0.98 0.88 to 1.09 0.96 0.86 to 1.07

14-89 243 0.6% 1,175 0.4% 1.11 0.96 to 1.28 1.10 0.95 to 1.27

90-364 194 0.5% 800 0.3% 1.31* 1.12 to 1.54 1.29* 1.09 to 1.52

365-1459 134 0.3% 601 0.2% 1.20 0.99 to 1.46 1.16 0.95 to 1.42

1460+ 52 0.1% 208 0.1% 1.31 0.96 to 1.80 1.32 0.96 to 1.82

Gastro-intestinal

0 38,953 95.5% 271,876 95.8% 1.00

1.00

>0-13 1,205 3.0% 8,052 2.8% 0.95 0.90 to 1.02 0.95 0.89 to 1.01

14-89 454 1.1% 2,977 1.0% 0.91 0.82 to 1.00 0.91 0.82 to 1.01

90-364 101 0.2% 717 0.3% 0.80 0.65 to 0.99 0.82 0.66 to 1.02

365+ 57 0.1% 311 0.1% 1.09 0.82 to 1.46 1.12 0.84 to 1.49

Parkinsonian

0 40,500 99.3% 282,982 99.7% 1.00 1.00

>0-13 30 0.1% 125 0.0% 1.29 0.95 to 1.74 1.19 0.88 to 1.61

14-89 60 0.1% 207 0.1% 1.40 0.94 to 2.07 1.27 0.86 to 1.89

90-364 65 0.2% 220 0.1% 1.44 1.08 to 1.92 1.27 0.95 to 1.70

Page 43 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


365-1459 88 0.2% 301 0.1% 1.56* 1.22 to 2.01 1.35 1.05 to 1.74

1460+ 27 0.1% 98 0.0% 1.61 1.04 to 2.50 1.39 0.89 to 2.18

Respiratory

0 36,768 90.2% 258,738 91.1% 1.00 1.00

>0-13 1,997 4.9% 12,966 4.6% 1.02 0.97 to 1.08 1.02 0.97 to 1.07

14-89 1,511 3.7% 9,321 3.3% 1.05 1.00 to 1.12 1.05 0.99 to 1.11

90-364 293 0.7% 1,809 0.6% 1.02 0.90 to 1.15 1.01 0.89 to 1.15

365-1459 169 0.4% 893 0.3% 1.18 1.00 to 1.40 1.18 1.00 to 1.40

1460+ 32 0.1% 206 0.1% 1.01 0.69 to 1.47 1.01 0.70 to 1.48

Urological

0 37,509 92.0% 267,060 94.1% 1.00 1.00

>0-13 295 0.7% 1,810 0.6% 1.06 0.94 to 1.20 1.02 0.90 to 1.15

14-89 1,167 2.9% 6,514 2.3% 1.15* 1.08 to 1.23 1.10* 1.03 to 1.17

90-364 773 1.9% 3,891 1.4% 1.27* 1.18 to 1.38 1.21* 1.12 to 1.31

365-1459 800 2.0% 3,550 1.3% 1.41* 1.30 to 1.53 1.35* 1.24 to 1.46

1460+ 226 0.6% 1,108 0.4% 1.28* 1.11 to 1.48 1.24* 1.07 to 1.44

Other

0 40,483 99.3% 282,179 99.4% 1.00 1.00

>0-13 83 0.2% 488 0.2% 1.06 0.84 to 1.35 1.04 0.82 to 1.32

14-89 165 0.4% 1,109 0.4% 0.89 0.75 to 1.05 0.89 0.76 to 1.06

90-364 36 0.1% 144 0.1% 1.52 1.05 to 2.20 1.55 1.07 to 2.24

DEP, drug exposure period; DDD, defined daily dose; OR, odds ratio, aOR, adjusted odds ratio; CI confidence interval; ACB Anticholinergic cognitive burden

* p<0.01

Page 44 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


eTable 3. Adjusted odds ratios of dementia by any new prescription of a probably or potently anticholinergic medication and by drug class

Any use Cases (n=35,555) Controls (n=221,772)

Measured at start of

DEP

Measured at end of

DEP

during DEP n % n % aORa 95% CI aOR

b 95% CI


No prescription 3791 10.7% 28866 13.0% 1.00 1.00

Any ACB 1 31463 88.5% 191183 86.2% 1.12* 1.08 to 1.16 1.11* 1.07 to 1.16

New ACB 2 912 2.6% 4594 2.1% 1.11* 1.03 to 1.20 1.12* 1.04 to 1.21

New ACB 3 10163 28.6% 55534 25.0% 1.13* 1.09 to 1.16 1.07* 1.04 to 1.10


No prescription 3791 10.7% 28866 13.0% 1.00 1.00

Any ACB1

Analgesic 20056 56.4% 119509 53.9% 1.02 1.00 to 1.05 1.02 0.99 to 1.05

Antidepressant 4468 12.6% 19010 8.6% 1.39* 1.34 to 1.45 1.26* 1.20 to 1.32

Antipsychotic 6534 18.4% 36702 16.5% 1.05* 1.02 to 1.09 1.04 1.01 to 1.07

Cardiovascular 24092 67.8% 148297 66.9% 0.97 0.94 to 1.00 0.98 0.95 to 1.02

Gastro-intestinal 8870 24.9% 53297 24.0% 0.96* 0.93 to 0.99 0.95* 0.93 to 0.98

Respiratory 7853 22.1% 47517 21.4% 0.99 0.96 to 1.02 0.95* 0.92 to 0.98

Other 9470 26.6% 57341 25.9% 0.95* 0.92 to 0.98 0.99 0.96 to 1.02

New ACB2

Analgesic 286 0.8% 1510 0.7% 1.08 0.95 to 1.23 1.09 0.96 to 1.24

Antipsychotic 6 0.0% 12 0.0% NA NA

Parkinsonian 39 0.1% 83 0.0% 1.55 1.04 to 2.30 1.32 0.88 to 1.96


Other 591 1.7% 3019 1.4% 1.07 0.98 to 1.18 1.10 1.00 to 1.21

New ACB3

Antidepressant 5708 16.1% 30035 13.5% 1.10* 1.06 to 1.14 1.08* 1.04 to 1.11

Antipsychotic 503 1.4% 2230 1.0% 1.14 1.03 to 1.27 1.12 1.01 to 1.25

Gastro-intestinal 1296 3.6% 7854 3.5% 0.93 0.88 to 0.99 0.93 0.87 to 0.99

Parkinsonian 91 0.3% 283 0.1% 1.32 1.03 to 1.69 1.16 0.90 to 1.49


Urological 2239 6.3% 10630 4.8% 1.22* 1.16 to 1.28 1.17* 1.11 to 1.23

Other 220 0.6% 1169 0.5% 1.03 0.89 to 1.19 1.03 0.89 to 1.20

DEP, drug exposure period; aOR, adjusted odds ratio; CI confidence interval; ACB Anticholinergic cognitive burden

a adjusted for covariates measured at the start of the period

* p<0.01

Page 45 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


eTable 4. Adjusted odds ratios of dementia by prescription of an anticholinergic medication as measured by the Anticholinergic drug scale and by drug class

Drug class

Cases (n=40,770) Controls (n=283,933) Measured at start of DEP Measured at end of DEP


b 95% CI

Any use by ADS score

No prescription 14,585 35.8% 114,814 40.4% 1.00 1.00

ADS1 21,074 51.7% 135,248 47.6% 1.07* 1.04 to 1.09 1.03 1.01 to 1.06

ADS2 7,038 17.3% 45,301 16.0% 0.98 0.95 to 1.01 0.96 0.93 to 0.99

ADS3 13,203 32.4% 79,114 27.9% 1.13* 1.10 to 1.16 1.08* 1.06 to 1.11

Any use by ADS score and class

No prescription 14,585 35.8% 114,814 40.4% 1.00 1.00

ADS1

Analgesic 5,275 12.9% 33,364 11.8% 0.96 0.93 to 1.00 0.97 0.93 to 0.00


Antipsychotic 7,602 18.6% 47,934 16.9% 1.03 1.00 to 1.06 1.02 0.99 to 1.05

Cardiovascular 4,320 10.6% 27,358 9.6% 1.02 0.98 to 1.06 1.02 0.98 to 1.06

Gastro-intestinal 832 2.0% 5,513 1.9% 0.96 0.89 to 1.04 0.96 0.89 to 1.04


Other 7,240 17.8% 49,734 17.5% 0.94* 0.91 to 0.97 0.94* 0.92 to 0.97

ADS2

Antipsychotic 367 0.9% 1,707 0.6% 1.12 1.00 to 1.26 1.10 0.97 to 1.24

Gastro-intestinal 5,820 14.3% 39,251 13.8% 0.94* 0.91 to 0.97 0.93* 0.90 to 0.96

Other 1,314 3.2% 6,688 2.4% 1.10* 1.04 to 1.18 1.07 1.00 to 1.14

ADS3


Antipsychotic 649 1.6% 3,247 1.1% 1.02 0.93 to 1.12 1.01 0.92 to 1.10

Gastro-intestinal 921 2.3% 6,176 2.2% 0.90* 0.84 to 0.97 0.91 0.85 to 0.98

Parkinsonian 270 0.7% 951 0.3% 1.48* 1.28 to 1.72 1.28* 1.10 to 1.49


Urological 3,236 7.9% 16,750 5.9% 1.24* 1.19 to 1.29 1.17* 1.12 to 1.23

DEP, drug exposure period; OR, odds ratio, aOR, adjusted odds ratio; CI confidence interval; ADS Anticholinergic drug scale

a adjusted for covariates measured at the start of the period

* p<0.01

Page 46 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


etable 5a. Adjusted odds ratios of dementia by prescription of a potent anticholinergic by drug class, by adjustment for BMI, smoking and alcohol abuse and by patients

with complete data. Adjustments made for covariates as measured at the start of the drug exposure period.

Adjusted for covariates measured at the start of the drug exposure period

All patients (40,770 cases, 283,933 controls) Patients without missing data at start of DEP (15,214 cases, 48,486 controls)

ACB score and drug class aORa

Unadjusted for alcohol abuse,

smoking, BMI

aORa

Unadjusted for alcohol abuse, smoking, BMI

95% CI aORb 95% CI 95% CI aOR

b 95% CI


No prescriptions 1.00 1.00 1.00 1.00

ACB 1 prescription 1.11* 1.07 to 1.15 1.11* 1.07 to 1.15 1.08 1.01 to 1.16 1.07 1.00 to 1.15


ACB 3 prescription 1.16* 1.13 to 1.19 1.16* 1.13 to 1.19 1.16* 1.11 to 1.21 1.16* 1.11 to 1.21


No prescription 1.00 1.00 1.00 1.00

ACB1

Analgesic 1.02 1.00 to 1.05 1.02 0.99 to 1.04 1.02 0.97 to 1.06 1.01 0.96 to 1.05

Antidepressant 1.37* 1.32 to 1.42 1.37* 1.32 to 1.42 1.43* 1.34 to 1.52 1.42* 1.34 to 1.51

Antipsychotic 1.05* 1.02 to 1.08 1.05* 1.02 to 1.08 1.08* 1.02 to 1.14 1.08* 1.02 to 1.13

Cardiovascular 0.97 0.94 to 0.99 0.96 0.94 to 0.99 0.94 0.90 to 0.99 0.93* 0.88 to 0.97

Gastro-intestinal 0.97 0.94 to 0.99 0.97 0.94 to 0.99 0.98 0.94 to 1.03 0.98 0.94 to 1.03

Respiratory 0.99 0.97 to 1.02 1.00 0.97 to 1.02 0.97 0.93 to 1.02 0.98 0.93 to 1.03

Other 0.95* 0.92 to 0.97 0.95* 0.92 to 0.97 0.94 0.90 to 0.99 0.94 0.90 to 0.99

ACB2


Antipsychotic 1.44 0.87 to 2.36 1.43 0.87 to 2.35 1.13 0.40 to 3.17 1.12 0.40 to 3.15

Parkinsonian 1.55* 1.12 to 2.14 1.56* 1.13 to 2.15 1.23 0.68 to 2.25 1.26 0.69 to 2.29


Other 1.07 1.00 to 1.15 1.07 1.00 to 1.15 1.13 0.99 to 1.28 1.13 0.99 to 1.28

ACB3




Parkinsonian 1.45* 1.25 to 1.68 1.44* 1.24 to 1.68 1.47* 1.10 to 1.97 1.46* 1.09 to 1.95


Urological 1.23* 1.18 to 1.28 1.23* 1.18 to 1.28 1.26* 1.17 to 1.36 1.26* 1.17 to 1.35

Other 0.99 0.87 to 1.13 0.99 0.87 to 1.12 0.96 0.77 to 1.19 0.96 0.77 to 1.19

DEP, drug exposure period; OR, odds ratio, aOR, adjusted odds ratio; CI, confidence interval; ACB, Anticholinergic Cognitive Burden score; BMI, body mass index

* p<0.01

Page 47 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


etable 5b. Adjusted odds ratios of dementia by prescription of a potent anticholinergic by drug class, by adjustment for BMI, smoking and alcohol abuse and by patients

with complete data. Adjustments made for covariates as measured at the end of the drug exposure period.

Adjusted for covariates measured at the end of the drug exposure period

All patients (40,770 cases, 283,933 controls) Patients without missing data at end of DEP (26,642 cases, 126,075 controls)

ACB score and drug class aORc

Unadjusted for alcohol abuse,

smoking, BMI

aORc

Unadjusted for alcohol abuse, smoking, BMI

95% CI aORd 95% CI 95% CI aOR

d 95% CI


No prescriptions 1.00

1.00

1.00

1.00



ACB 3 prescription 1.11* 1.08 to 1.14 1.10* 1.08 to 1.13 1.09* 1.06 to 1.13 1.09* 1.05 to 1.12


No prescription 1.00

1.00

1.00

1.00

ACB1



Antipsychotic 1.04 1.01 to 1.07 1.04 1.01 to 1.07 1.05* 1.02 to 1.09 1.05* 1.02 to 1.09

Cardiovascular 0.98 0.95 to 1.01 0.97 0.94 to 1.00 0.97 0.93 to 1.01 0.94 0.91 to 0.98

Gastro-intestinal 0.96* 0.93 to 0.99 0.96 0.94 to 0.99 0.97 0.94 to 1.01 0.97 0.94 to 1.01


Other 0.95* 0.92 to 0.98 0.95* 0.92 to 0.97 0.95* 0.91 to 0.98 0.94* 0.91 to 0.98

ACB2



Parkinsonian 1.32 0.96 to 1.82 1.37 1.00 to 1.89 1.40 0.93 to 2.10 1.47 0.97 to 2.21


Other 1.09 1.01 to 1.17 1.09 1.01 to 1.17 1.07 0.98 to 1.18 1.07 0.97 to 1.17

ACB3




Parkinsonian 1.29* 1.11 to 1.50 1.29* 1.11 to 1.50 1.31* 1.07 to 1.59 1.31* 1.08 to 1.59


Urological 1.18* 1.13 to 1.23 1.17* 1.12 to 1.22 1.16* 1.10 to 1.22 1.16* 1.09 to 1.22

Other 0.99 0.87 to 1.13 0.98 0.86 to 1.12 1.01 0.87 to 1.17 1.00 0.86 to 1.16

DEP, drug exposure period; OR, odds ratio, aOR, adjusted odds ratio; CI, confidence interval; ACB, Anticholinergic Cognitive Burden score; BMI, body mass index

* p<0.01

Page 48 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


1

ISAC APPLICATION FORM

PROTOCOLS FOR RESEARCH USING THE CLINICAL PRACTICE RESEARCH DATALINK (CPRD)

ISAC use only: Protocol Number Date submitted

............................. .............................

IMPORTANT

If you have any queries, please contact ISAC Secretariat:

[email protected] 1. Study Title

Benzodiazepine, non benzodiazepine derivatives, and anticholinergic use and incident dementia

2. Principal Investigator (full name, job title, organisation & e-mail address for correspondence regarding this protocol)

George Savva, Senior Lecturer, University of East Anglia, [email protected]

3. Affiliation (full address)

School of Health Sciences, University of East Anglia, Norwich Research Park, Norwich, NR4 7TJ

4. Protocol’s Author (if different from the principal investigator)

Kathryn Richardson, University of East Anglia, [email protected]

5. List of all investigators/collaborators (please list the names, affiliations and e-mail addresses* of all collaborators, other than the principal investigator)

Kathryn Richardson, University of East Anglia, [email protected]

Chris Fox, University of East Anglia, [email protected] Ian Maidment, Aston University, [email protected]

Antony Arthur, University of East Anglia, [email protected] Yoon Loke, University of East Anglia, [email protected] Nicholas Steel, University of East Anglia, [email protected] Kathleen Bennett, Trinity College Dublin, [email protected] Carol Brayne, University of Cambridge, [email protected]

Noll Campbell, Purdue University, [email protected] Phyo Myint, University of Aberdeen, [email protected]

Malaz Boustani, University of Indiana, [email protected] Carlota Grossi, University of East Anglia, [email protected]

*Please note that your ISAC application form and protocol must be copied to all e-mail addresses listed above at the time of submission of your application to the ISAC mailbox. Failure to do so will result in delays in the processing of your application.

6. Type of Institution (please tick one box below)

Academia Research Service Provider Pharmaceutical Industry NHS Government Departments Others

7. Financial Sponsor of study

Pharmaceutical Industry (please specify) Academia(please specify)

Government / NHS (please specify) None

Other (please specify) Alzheimer’s Society

8. Data source (please tick one box below) Sponsor has on-line access Purchase of ad hoc dataset Commissioned study

Other (please specify) 9. Has this protocol been peer reviewed by another Committee?

Yes* No

* Please state in your protocol the name of the reviewing Committee(s) and provide an outline of the review process and outcome. 10. Type of Study (please tick all the relevant boxes which apply) Adverse Drug Reaction/Drug Safety Drug Use Disease Epidemiology

Drug Effectiveness Pharmacoeconomic Other

11. This study is intended for:

Publication in peer reviewed journals Presentation at scientific conference

Presentation at company/institutional meetings Other Public

events /educational meetings in primary and secondary care

Page 49 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


2

12. Does this protocol also seek access to data held under the CPRD Data Linkage Scheme?

Yes No

13. If you are seeking access to data held under the CPRD Data Linkage Scheme*, please select the

source(s) of linked data being requested.

Hospital Episode Statistics Cancer Registry Data** MINAP ONS Mortality Data

Index of Multiple Deprivation/ Townsend Score

Mother Baby Link Other: (please specify)

* As part of the ISAC review of linkages, the protocol may be shared - in confidence - with a representative of the requested linked data set(s) and summary details may be shared - in confidence - with the Confidentiality Advisory Group of the Health Research Authority. **Please note that applicants seeking access to cancer registry data must provide consent for publication of their study title and study institution on the UK Cancer Registry website. Please contact the CPRD Research Team on +44 (20) 3080 6383 or email [email protected] to discuss this requirement further.

14. If you are seeking access to data held under the CPRD Data Linkage Scheme, have you already discussed your request with a member of the Research team?

Yes No*

*Please contact the CPRD Research Team on +44 (20) 3080 6383 or email [email protected] to discuss your requirements before submitting your application. Please list below the name of the person/s at the CPRD with whom you have discussed your request.

Antonis Kousoulis

15. If you are seeking access to data held under the CPRD Data Linkage Scheme, please provide the

following information: The number of linked datasets requested: 1

A synopsis of the purpose(s) for which the linkages are required:

Index of Multiple Deprivation as we are going to match on this, and

Townsend Score as we are going to adjust for this in a sensitivity analysis.

Is linkage to a local dataset with <1 million patients being requested?

Yes* No

* If yes, please provide further details:

16. If you have requested linked data sets, please indicate whether the Principal Investigator or any of the collaborators listed in response to question 5 above, have access to any of the linked datasets in a patient identifiable form, or associated with a patient index.

Yes* No

* If yes, please provide further details:

Page 50 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


3

17. Does this protocol involve requesting any additional information from GPs?

Yes* No * Please indicate what will be required: Completion of questionnaires by the GPψ Yes No

Provision of anonymised records (e.g. hospital discharge summaries) Yes No

Other (please describe)

ψ Any questionnaire for completion by GPs or other health care professional must be approved by ISAC before circulation for completion. 18. Does this protocol describe a purely observational study using CPRD data (this may include the

review of anonymised free text)?

Yes* No** * Yes: If you will be using data obtained from the CPRD Group, this study does not require separate ethics approval from an NHS Research Ethics Committee. ** No: You may need to seek separate ethics approval from an NHS Research Ethics Committee for this study. The ISAC will provide advice on whether this may be needed. 19. Does this study involve linking to patient identifiable data from other sources?

Yes No

20. Does this study require contact with patients in order for them to complete a questionnaire?

Yes No

N.B. Any questionnaire for completion by patients must be approved by ISAC before circulation for completion. 21. Does this study require contact with patients in order to collect a sample?

Yes* No

* Please state what will be collected

22. Experience/expertise available

Please complete the following questions to indicate the experience/expertise available within the team of researchers

actively involved in the proposed research, including analysis of data and interpretation of results

Previous GPRD/CPRD Studies Publications using GPRD/CPRD data

None 1-3 > 3

Yes No

Is statistical expertise available within the research team? If yes, please outline level of experience KR, GS, and KB are experienced

statisticians. KB is an Assoc Prof in Pharmacoepidemiology and co-investigator on a current CPRD ovarian cancer study. KR has experience analysing CPRD data (with paper on childhood asthma in press) and attended the LSHTM Practical

Pharmacoepidemiology course (which included CPRD training). Is experience of handling large data sets (>1 million records) available within the research team? If yes, please outline level of experience KR has used Canada’s British Columbia linked primary care data (>4 million patients) and has 5 publications using this. KB regularly works with Irish pharmacy dispensing data and has many publications on this. AA has 3 publications using QResearch data. Is UK primary care experience available within the research team?

If yes, please outline level of experience NS has significant research experience and previous UK primary care clinical experience, and has supervised a PhD student using CPRD data.

23. References relating to your study

See reference list below in section 15.

Page 51 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


4

PROTOCOL CONTENT CHECKLIST

In order to help ensure that protocols submitted for review contain adequate information for protocol evaluation, ISAC have produced instructions on the content of protocols for research using CPRD data. These instructions are available on the CPRD website (www.cprd.com/ISAC). All protocols using CPRD data which are submitted for review by ISAC must contain information on the areas detailed in the instructions. IF you do not feel that a specific area required by ISAC is relevant for your protocol, you will

need to justify this decision to ISAC.

Applicants must complete the checklist below to confirm that the protocol being submitted includes all the areas required by ISAC, or to provide justification where a required area is not considered to be relevant

for a specific protocol. Protocols will not be circulated to ISAC for review until the checklist has been completed by the applicant.

Please note, your protocol will be returned to you if you do not complete this checklist, or if you answer ‘no’ and fail to include justification for the omission of any required area.

Included in

protocol?

Required area Yes No If no, reason for omission

Lay Summary (max.200 words)

Background

Objective, specific aims and rationale

Study Type Descriptive Hypothesis Generating Hypothesis Testing

Study Design

Sample size/power calculation (Please provide justification of sample size in the protocol)

Study population (including estimate of expected number of relevant patients in the CPRD)

Selection of comparison group(s) or controls

Exposures, outcomes and covariates Exposures are clearly described Outcomes are clearly described

Use of linked data (if applicable)

Data/ Statistical Analysis Plan There is plan for addressing confounding There is a plan for addressing missing data

Patient/ user group involvement †

Limitations of the study design, data sources and analytic methods

Plans for disseminating and communicating study results

† It is expected that many studies will benefit from the involvement of patient or user groups in their planning and refinement, and/or in the interpretation of the results and plans for further work. This is particularly, but not exclusively true of studies with interests in the impact on quality of life. Please indicate whether or not you intend to engage patients in any of the ways mentioned above. Voluntary registration of ISAC approved studies: Epidemiological studies are increasingly being included in registries of research around the world, including those primarily set up for clinical trials. To increase awareness amongst researchers of ongoing research, ISAC encourages voluntary registration of epidemiological research conducted using MHRA databases. This will not replace information on ISAC approved protocols that may be published in its summary minutes or annual report. It is for the applicant to determine the most appropriate registry for their study. Please inform the ISAC secretariat that you have registered a protocol and provide the location.

Page 52 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


5

PROTOCOL

1 Lay Summary There is no effective cure for dementia and so identifying modifiable risk factors is essential.

The long term use of medications known to affect the brain for short durations may cause

permanent harm. Some of these medications are widely used among older people, and so

even small increases in individual risks could lead to large increases in the number of people

with dementia. In this project we will examine whether the use of certain medicines that affect

cognition increase the risk of dementia. We will also examine whether the risk persists or is

reduced once the medication is stopped.

Clinical trials rarely examine the long term cognitive harms of medications, and so

‘observational’ studies are needed. We will identify CPRD patients who, since April 2004, were

diagnosed with dementia aged 65-99 years. We will compare the medication history of each

patient with dementia to that of 7 people of the same age and sex who had no dementia (on

the same date). We will compare the use and dosage of medications prescribed during the

period beyond 4 years before dementia diagnosis, whilst controlling for other health and social

differences.

Our findings will contribute to prescribing guidance and practice for management of many

common long term conditions and have the potential to reduce the future risks of dementia.

2 Objectives, specific aims and rationale Our broad aim is to test whether the short term cognitive effects of commonly used

medications translate into an increased risk of dementia incidence with longer term use.

As we intend this project to be hypothesis testing (rather than hypothesis generating), we

focus on examining medications with adverse cognitive effects for which there has been (1)

consistent evidence of short-term effects, (2) published evidence of potential long-term

effects, and (3) are commonly prescribed rather than obtained over-the-counter (OTC). Hence

we aim to examine the use of benzodiazepines, non-benzodiazepine derivatives, and

medications with anticholinergic activity.

2.1 Objectives

We will conduct a nested case-control study among CPRD patients to:

1. determine whether the use of benzodiazepines (BZD), non benzodiazepine derivatives (Z-

drugs), or medications with anticholinergic activity increases the risk of dementia, and

whether risk increases with greater duration, dosage or concurrent use.

2. determine whether these effects persist beyond medication cessation.

2.2 Aims

All hypotheses were generated a priori based on published studies (see background section).

For (1) benzodiazepines, (2) Z-drugs, and (3) medications with anticholinergic activity we will:

1. test whether older people with dementia (aged 65-99 and diagnosed since April 2004)

were more likely to have been prescribed a greater quantity of medication than people not

diagnosed with dementia in the period beyond 4 years before the index date (defined by

the dementia diagnosis date), whilst adjusting for confounding factors.

Page 53 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


6

2. asses the evidence for a dose-response relationship by:

a. examining duration and quantity prescribed (i.e. defined daily doses [DDDs]);

b. and for anticholinergics - examine whether dementia risk increases with greater

anticholinergic burden

3. examine the above associations in subgroups of patients with depression, anxiety or

insomnia.

4. test whether this association for exposures beyond 4 years pre diagnosis persists once

adjusted for the exposures occurring 0-4 years pre diagnosis.

5. test if the association remains once medication use is discontinued.

6. test for how many years since discontinuation the association remains within categories of

previous use (light, medium, heavy user).

2.3 Rationale

• Aim 1 will address objective 1. Drug exposure will not be examined in the 4 years

before index date to reduce issues of protopathic bias and surveillance bias, and

because the drug exposure would be expected to cause dementia over a longer period

of time.

• Aim 2 will provide evidence of a dose-response to further support objective 1, hence

strengthen the case for a causal effect if one exists [1].

• Aim 3 also supports objective 1 by additionally reducing the impact of confounding by

indication and protopathic bias.

• Aim 4 will examine a potential mechanism by addressing whether exposure beyond 4

years before index is just a marker for 0-4 year exposure and the effect is only short

lived. Associations remaining for exposures beyond 4 years before index once adjusted

for 0-4 year exposure would indicate evidence of a cumulative effect.

• Aim 5 will address objective 2.

• Aim 6 will further examine objective 2 by looking at whether the effect of stopping

depends on prior cumulative use.

3 Background Dementia is a syndrome of progressive decline in cognitive and daily function causing

significant patient, carer and societal burden. There are estimated to be between 670,000 and

850,000 people with dementia in the UK [2,3]. The number of dementia cases worldwide are

expected to double over the next 20 years [4]. Much is still not understood about the causes

and risk factors for dementia. There is no cure for dementia and so the identification of

potentially modifiable risk factors is essential.

Many middle aged and older people take multiple medications for the prevention or

management of chronic diseases. Yet there is concern over the impact of some medication use

on cognitive health. A recent systematic review of randomised controlled trials examined the

risks of short term cognitive impairment due to benzodiazepines, anticholinergics,

antihistamines, antipsychotics and opioid use [5]. Consistent amnestic and nonamnestic

cognitive impairments were found for benzodiazepine use, but evidence was less consistent

for nonbenzodiazepine derivatives. Antihistamines and anticholinergic tricyclic antidepressants

induced non-amnestic deficits in attention and information processing. There was insufficient

evidence surrounding anticholinergic bladder relaxant drugs, opioids and antipsychotics to

draw conclusions.

Whether these potential short-term cognitive impairments lead to an increased risk in

dementia is unknown. Benzodiazepine use has been associated with increased dementia

Page 54 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


7

incidence in longer term observational studies, but these studies have been limited by their

size, adjustment for confounding, or type of database used [6–9]. They rarely adjust for the

use of other classes of medications with adverse cognitive effects. A recent Canadian study

found a greater incidence of dementia in those taking more than 90 DDDs of benzodiazepines

in the period 5-10 years before dementia diagnosis [7]. Data from Taiwanese medical records

also suggest that elevated risks persist for at least three years after benzodiazepine use

cessation, depending on the previous usage [8].

Benzodiazepines are indicated to treat anxiety and insomnia and are commonly used by older

people. Benzodiazepines are only indicated for short-term use (2 to 4 weeks) because they are

highly addictive with serious withdrawal reactions, however studies show that many people

take benzodiazepines for many years [10]. In the NHS England in 2013, 15.4 million

benzodiazepine prescriptions were dispensed, costing £73 million [11]. Benzodiazepines are

known to affect memory, concentration and coordination, but whether this continues after

cessation is controversial. More recently, non-benzodiazepines called ‘Z-drugs’ (zolpidem,

zopiclone, and zaleplon) have been developed to have an improved safety profile than

benzodiazepines in the treatment of insomnia, but their cognitive effects are not so well

understood, and the improved safety profile requires further confirmation.

Many of the medications with short term cognitive effects have anticholinergic activity. Longer

term observational studies have also linked medications with anticholinergic activity with

cognitive decline and dementia [12,13]. Many commonly used medicines including some used

for urinary incontinence, depression, asthma, and heart problems block the neurotransmitter

acetylcholine which affects many body systems and is important for forming memories.

Anticholinergic medications can cause dry mouth, constipation, urinary retention, low blood

pressure and also confusion, difficulty concentrating, agitation and memory problems. These

side effects have been assumed to be temporary, but recent studies suggest that long term

use of anticholinergics may cause cognitive decline and dementia.

Estimates of the prevalence of definite anticholinergic use in the elderly ranges from 4%-27%,

with around 42%-47% regularly using medications that possibly possess anticholinergic activity

[14–16]. Taking a few medications that are only possibly or potentially anticholinergic for a

longer duration may pose long-term cognitive risks, as our project team members have found

that people taking at least 3 possible anticholinergics for 90 days or longer had a higher risk of

mild cognitive impairment, but not dementia [17]. However, an American study found an

increased risk of dementia after 90 DDDs of strong anticholinergics [13]. A study in France of a

small sample of 372 people aged 60 and over found a higher risk of mild cognitive impairment

among people taking anticholinergic medications; there was no identified increased risk of

dementia during an eight year follow-up period, but the study lacked statistical power [18]. In

2011 we used data from 13,004 participants of the MRC Cognitive Function and Ageing Study

to show that people taking anticholinergic medication not only had worse cognition, but also

additional cognitive decline over two years compared to people who did not take such

medications [14]. So while anticholinergics are known to bring short term cognitive harms,

evidence of their long term effects is conflicted.

Antihistamine use was also shown to have short-term cognitive effects [5], but many

antihistamines have anticholinergic activity. However, as most antihistamines are purchased

OTC rather than prescribed, we will not be able to examine this medication class specifically,

but will adjust for non-anticholinergic antihistamine use in our analyses.

The main limitations to observational studies in this area are confounding, in particular

confounding by indication, and protopathic bias. For example depression is an indication for

the anticholinergic antidepressants, yet also associated with increased dementia risk [19].

Protopathic bias, by which the use of these medications is a marker of prodromal symptoms of

Page 55 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


8

dementia, is a concern. Studies have suggested that symptoms of depression and anxiety may

occur in the period before dementia diagnosis [20].

Hence we focus our hypothesis testing study on the use of benzodiazepines, Z-drugs, and

medications with anticholinergic activity as these have the greatest evidence of short- and

long-term cognitive effects. We will adjust for the other medication classes mentioned in the

literature as confounders and will aim to address confounding and protopathic bias. This

project aims to refute or confirm the effects of anticholinergic medications, benzodiazepines

and Z-drugs on dementia incidence.

4 Study Type - Hypothesis testing

5 Study Design - Nested case-control study

6 Sample size We plan to use data from all available people with dementia meeting our definition to

maximise precision. The following power calculations are based on the estimated 34,125 cases

(with 238,875 matched controls) identified during the CPRD feasibility check. This number is

substantially larger than any currently published study on this topic.

The prevalence of medication use and health conditions were estimated from the CPRD GOLD

sample dataset v1.5 in 631 people aged over 65 years. Over a 6 year period, 6%, 3%, and 11%

were given prescriptions for benzodiazepines, Z-drugs and definite anticholinergics lasting 90

days or longer. A diagnosis of depression, insomnia, and anxiety was received by 19%, 17%,

and 18% of patients, respectively.

Previous studies generally suggest an increased risk of dementia or cognitive decline after 90

days or 90 DDDs of exposure, rather than after any use of a benzodiazepine or anticholinergic

medication [7,13,17]. We have estimated the smallest detectable effect sizes for detecting a

difference in the proportion of cases and controls with 90 days of prescription exposure

assuming a 2-sided test with alpha as 0.01. The smallest odds ratios (OR) detectable are 1.08,

1.12, and 1.06 for 90 days of BZD, Z-drug, and definite anticholinergic exposure at 80% power.

7 Use of linked data (if applicable) Index of Multiple Deprivation will be used in the matching criteria, and we will adjust for

Townsend Score in a sensitivity analysis.

8 Study population First patients will be excluded if they ever have diagnoses of:

• Motor neuron disease (read code F152%), HIV/AIDS (A788%, A789%), Multiple

sclerosis (F20%), Down’s syndrome (PJ0%), Alcoholism or related diseases (see

appendix for read codes).

o Rationale: Conditions associated with a greater risk of dementia through

different pathologies, and differential use of BZDs or anticholinergics.

Dementia cases definition criteria (with rationale for decisions below):

Page 56 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


9

1. Patients with first dementia event recorded between 01/04/2006 and the practice

“Last Collection Date” (this first date defines the “Index Date”).

o Rationale: Date chosen to capture dementia coded since the start of the

Quality Outcomes Framework (QOF) and more recent medication use.

2. First dementia event defined as the first of a dementia read code (see appendix) or a

prescription for a dementia drug (memantine, donepezil, rivastigmine, or galantamine)

which is followed by a dementia code within a year.

o Rationale: Dementia read codes from QOF excluding alcohol induced

dementia. Dementia drugs added due to occasional suspected delay in

diagnosis coding (i.e. from specialists), and to be in line with other validated

definitions [21,22].

3. Patients aged between 65 and 99 years at the index date

o Rationale: Age criteria due to the vast majority of dementias occurring at this

age, more homogeneous dementia, and consistency with other publications.

4. Patients have at least 6 years of continuous follow-up with Up-To-Standard (UTS) data

(and acceptable records) before the index date.

o Rationale: This is to enable at least 1 year of drug exposure prior to 4 years

before diagnosis and to enable drug exposure to occur after at least one year

of UTS data, which gives sufficient time for confounder and indication

estimation, and recording to have stabilised for patients new to the practice

[23].

Exclusion criteria:

• A read code for “dementia annual review” before the index date (read code 6AB.00).

o Rationale: Due to dementia diagnosis date being missing.

CPRD feasibility estimates returned 34,125 potential dementia patients in CPRD (using the

June 2015 version of the database) before the health conditions above are excluded.

8.1 Selection of controls

All cases will be categorised according to the number of years of acceptable, UTS, continuously

registered data available before index date. The number of years before four years before the

index date (excluding the first year of UTS follow-up) will define the drug exposure period. See

figure 1 which summarises the study design.

Figure 1. Study design

Page 57 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


10

7 controls per case will be selected (without replacement) using incidence density sampling

[24,25] from all patients (including cases pre diagnosis) in CPRD with the following matching

criteria on the index date:

1. Same sex

2. Year of birth +/- 3 years

o Rationale: Age and sex matches to select control patients similar to dementia

patients.

3. No dementia diagnosis or dementia medication on or before the index date (see

appendix).

o Rationale: To ensure no dementia events exist before the index date for

controls.

4. The same number of years of continuous UTS follow-up before the index date as the

case.

o Rationale: To enable the same duration and calendar time for the exposure

period. The controls will be assigned the same drug exposure period as the

case.

5. Same quintile of practice level Index of Multiple Deprivation (IMD).

o Rationale: To enable matching on area level deprivation which correlates with

comorbidity, health behaviours and dementia risk.

The ratio of 7 controls per case was chosen to maximise controls (10 controls per case

previously found to give as precise estimates as use of the full cohort [26]) whilst being below

the total 300,000 CPRD patient limit. When incidence density sampling is used to select

controls, the estimated odds ratios are unbiased estimates of incidence rate ratios [24].

9 Exposures The drug exposure period is defined by the number of years before the index date excluding

the four years before index and the first year of UTS follow-up (see figure 1). The main

exposures of interest are BZDs, Z-drugs, and medications with anticholinergic activity. We will

extract all prescriptions for patients occurring during their exposure period. Extracted

information will include the date of each prescription, the drug name, the number of daily

doses instructed, the dose and quantity prescribed.

Our primary exposure variable is the number of defined daily doses (DDD) and we hypothesise

a relationship with dementia risk above 90 DDDs of exposure (based on published studies). We

will calculate the total dose of each prescription by multiplying the dose of each tablet by the

number of tablets prescribed. To enable comparison of doses across drug classes, we will

convert the total dose for each medication to a number of DDDs, defined as the assumed

average maintenance dose per day for a drug based on its main indication in adults, using the

DDD values assigned by the World Health Organisation’s Collaborating Centre for Drug

Statistics Methodology (www.whocc.no/atc_ddd_index).

We will also estimate the duration of each prescription in days by dividing the number of

tablets prescribed by the dosing instructions (e.g. number of tablets to be taken per day).

When the dosing instructions are missing, we will assume an estimate based on the median

tablets to be taken per day for the same drug in patients with dosing instructions.

Benzodiazepines will be defined as those in WHO-Anatomical Therapeutic Chemical (ATC)

categories N05BA or N05CD and the Z-drugs defined as Zopiclone, Zaleplon, or Zolpidem (ATC

N05CF).

Page 58 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


11

The anticholinergic properties of prescribed medications will be assessed using the

Anticholinergic Cognitive Burden (ACB) scale (www.agingbraincare.org/tools/abc-

anticholinergic-cognitive-burden-scale/) [27,28]. This is a frequently updated scale that

classifies, through expert consensus and literature review, the evidence for the anticholinergic

activity of medications. Medications with serum anticholinergic activity or in vitro affinity to

muscarinic receptors but with no known clinically relevant negative cognitive effects are

scored 1 ‘possible’ , while drugs with established and clinically relevant anti-cholinergic effects

are scored 2 ‘probable’ based on blood-brain penetration and 3 ‘definite’ if also have reported

associations with delirium. We will classify separately medication exposure according to

exposure to (1) possible, (2) probable, and (3) definite anticholinergics.

For a patient at a given point in time, total anti-cholinergic burden is defined as the sum of ACB

scores for all medications taken. We will estimate the cumulative anticholinergic burden as the

mean ACB score over the exposure period (by summing each medication’s duration multiplied

by its ACB score, and dividing by the length of the exposure period) and the duration taking at

least 3 possible anticholinergics simultaneously.

Anticholinergic medication exposure will also be assessed within medication classes of

antidepressants, first-generation antipsychotics, second-generation antipsychotics, urologicals,

respiratory, and cardiovascular drugs. Exposures will also be assessed for the 5 most

commonly prescribed BZDs or Z-drugs.

As a sensitivity analysis, other anticholinergic medication scales will be used instead of the

ACB. We will code the medications according to both the US based Anticholinergic Drug Scale

(ADS) and the Anticholinergic Risk Scales (ARS), which also quantify the degree of anti-

cholinergic activity exhibited by a medication into scores of 0, 1, 2, or 3 according to literature

review and expert consensus [29,30]. We will also quantify the anticholinergic component of

the Drug Burden Index over the exposure period, which estimates anticholinergic burden using

the daily dose and the minimum efficacious dose [31].

As a secondary analysis, by considering all medication exposures occurring during the exposure

period and the 4 years prior to the index date, we will quantify time since discontinuation for

(1) BZD/Z-drugs and (2) definite anticholinergics. We will categorise never users, current users

and the time since discontinuation in groups. We will also classify cumulative use by DDDs pre

discontinuation (in 3 groups of light, medium and heavy user).

10 Covariates Patients will already be matched on sex, birth year (+/- 3 years), calendar time and IMD

quintile. It is important to capture to the greatest extent possible the health status of each

patient (particularly all indications for the medications and conditions that have a greater risk

of dementia). We measure covariates at 4 years prior to dementia rather than at the index

date, as this is when the drug exposure period ends and we want to limit the amount of events

captured that are consequences of taking the medications.

The following additional covariates will be adjusted for in regression analyses.

10.1 Covariates measured at 4 years prior to the index date:

• Age (years)

• Body Mass Index, calculated as weight (in kg) divided by height2

(in m) using the latest

records and categorised [32,33].

• Smoking status (non-smoker, ex-smoker, current smoker) using most recent record

[34].

Page 59 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


12

• Previous diagnoses (at any point in records up to 4 years prior to the index date):

o Insomnia, depression and anxiety (see appendix for codes). Duration with

insomnia, depression and anxiety (time since first diagnosis, categorised),

depression severity (mild, moderate, severe), phobia, drug abuse, symptoms

of depression, symptoms of anxiety, time since first depression symptom

(categorised), time since first anxiety symptom (categorised) [35,36].

o Family history of dementia recorded (where recorded).

o Severe Mental Illnesses (SMI) - schizophrenia, bipolar disorder, delusional

disorder, depressive psychosis, schizoaffective disorder, brief psychoses, and

psychoses not otherwise specified [37,38].

o Parkinson’s disease, Parkinson’s disease severity - prescription for drugs

containing levodopa (ATC N04BA), MAO-B inhibitor prescription (N04BD),

COMT-inhibitor prescription (N04BX01, N04BX02).

o Diabetes, Diabetes with complications, dyslipidaemia, Mood affective

disorders, Meniere’s disease, restless leg syndrome, Epilepsy, Hemiplegia and

paraplegia, Neuropathic pain, migraine, Hypertension, stroke, transient

ischaemic attack, angina, Cerebrovascular disease, Congestive heart disease,

Myocardial infarction, Peripheral vascular disease, Other atherosclerotic

disease, Atrial fibrillation, Deep vein thrombosis, Acute coronary syndrome,

Coronary artery bypass graft, Percutaneous coronary intervention, Other

cardiovascular surgery, Osteoarthritis, rheumatoid arthritis, other connective

tissue disease, chronic back pain, Chronic obstructive pulmonary disease,

asthma, rhinitis, other allergic disease, GERD, reflux oesophagitis, Peptic ulcer,

gastric ulcer, Irritable bowel syndrome, Inflammatory Bowel disease, Chronic

Diarrhoea, Gastric Bypass Surgery, intestinal surgery, Mild liver disease,

Moderate or severe liver disease, Dermatitis, eczema, psoriasis, Urinary

incontinence/overactive bladder, Prostatism, Renal disease, Cancer,

Metastatic tumour [39–41].

10.2 Covariates measured during the exposure period and the year prior:

• Recent stress/bereavement [42–44].

• Recent agitation

10.3 Covariates measured during the exposure period:

• Mean alcohol consumption (categorised by units per week) [45].

• Blood pressure (categorised) - average systolic blood pressure (in mmHg) [46,47].

• Mean annual frequency of GP visits (categorised).

• Categorised DDDs and duration of exposure to non-anticholinergic medications with

evidence of cognitive effects:

o Other all CNS active medications by class:

� Hypnotics and anxiolytics (BNF 4.1)

� First-generation antipsychotics, second-generation antipsychotics (BNF

4.2)

� Antidepressants (BNF 4.3)

� Drugs used in Nausea and vertigo (BNF 4.6)

� Analgesics - Opioids (BNF 4.7.2), non-opioids (rest of BNF 4.7)

� Antiepileptic drugs (BNF 4.9)

� Drugs used in Parkinson’s disease (BNF 4.9)

o Antihistamines (BNF 3.4.1) not listed on the anticholinergic scale.

• Any statin prescription, antiplatelet agent prescription, ACE inhibitor prescription.

• Average number of all other cardiovascular and non-cardiovascular medications

prescribed (excluding topical medications and eye drops).

Page 60 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


13

11 Data/ Statistical Analysis Stata (version 14) will be used for data management and statistical analysis. All reported P

values reported will be two tailed, with P<0.01 defining significance. Alpha of 0.01 was chosen

as being equivalent to a Benforroni correction for the five primary exposure outcomes.

First, we will explore the internal validity of the dementia diagnosis in the cases. In line with

other dementia studies [21], we will examine the proportion of cases with a second

(confirmatory) dementia-related code, e.g. additional dementia medication prescriptions,

additional dementia diagnosis codes, a dementia annual review, a specific dementia test prior

to diagnosis (e.g. Mini Mental State Examination, Clock Drawing Test, or Abbreviated Mental

Test [7-Minute Screen]), a referral to a specialist (e.g. neurologist, geriatrician or psycho-

geriatrician), or an assessment based on neuroimaging technique (e.g. magnetic resonance

imaging, computed tomography, or single photon emission computed tomography), or

dementia symptoms (e.g. memory impairment, aphasia, apraxia, or agnosia). Dementia cases

will be also be described according to their first coded sub-type.

Second, the characteristics 4 years prior to the index date of the cases and controls will be

described (percentages, means (SD), median (IQR) where relevant). We will summarise the

usage patterns of medications with adverse cognitive effects over the exposure period.

11.1.1 Primary analysis

We will assess the association between benzodiazepine exposure, anticholinergic exposure

and dementia using conditional logistic regression. Odds ratios (and 95% confidence intervals)

will be provided unadjusted and adjusted for exposure to each class of medications with

adverse cognitive effects and the covariates listed above. Where covariates are categorised

they will be grouped into sensible categories according to the distribution of the data, i.e.

tertiles or quintiles for continuous measures.

The association between dementia and the primary exposure of DDDs will be examined, for (i)

benzodiazepines, (ii) Z-drugs, and medications with (iii) possible, (iv) probable and (v) definite

anticholinergic activity. The model will adjust for the covariates as well as the number of DDDs

of the other medications with potential adverse cognitive effects.

The association between dementia and the following secondary exposures will be separately

examined:

• Duration of prescriptions. Adjusting for the covariates as well as the duration of

prescriptions of the other medications with adverse cognitive effects.

• Any prescription. Adjusting for the covariates as well as any prescriptions for the other

medications with adverse cognitive effects.

• For anticholinergics - average ACB sum (categorised) and duration taking 3 or more

possible anticholinergics concurrently (categorised).

11.1.1.1 Sensitivity analyses

1) For the primary analysis above, the interactions between medication use and both age and

sex will be tested, with subgroup results reported if p<0.05.

2) For a potential anticholinergic synergistic effect we will test for an interaction between

definite and possible anticholinergic use (defined as the DDDs of simultaneous use of the

definite and possible anticholinergics).

Page 61 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


14

3) We will repeat the primary analysis:

a) In cases (and their matched controls) with a second dementia-related code or

medication as a subgroup with ‘confirmed’ dementia due to the potential for cases

with only one code to be misclassified.

b) In cases (and their matched controls) with the index date defined as the first diagnosis

code for dementia.

c) In three subgroups: those with a diagnosis of insomnia, anxiety and depression.

d) Excluding patients with Parkinson’s disease, Severe Mental Illness, or drug abuse.

e) Using a covariate selection procedure, i.e. retaining those that affect the odds ratio by

10% or more.

f) Adjusting for patient level deprivation in those with data linkage, using the Townsend

deprivation score in deciles [48].

g) Adjusting only for confounders measured up to the start of the drug exposure period,

so to not adjust for variables potentially on the causal pathway during the drug

exposure period. However, this analysis has the limitation of potentially under

adjusting for the medication indications.

4) The anticholinergic analyses will be repeated with the ADS and ARS scales instead of the

ACB. As the Drug Burden Index does not categorise anticholinergic burden, we will only

compare the average score of the anticholinergic component over the exposure period.

5) To examine whether associations observed for exposures beyond 4 years pre index are

only due to an increased propensity to continue taking those medications closer to

diagnosis, we will also adjust for exposures in the 0-4 years before the index date, and

shall also split exposures in the period beyond 4 years into several 2-year periods.

6) Anticholinergic medications will also be assessed within medication classes of

antidepressants, first-generation antipsychotics, second-generation antipsychotics,

urologicals, respiratory, and cardiovascular drugs. The effect of the 5 most commonly

prescribed BZDs or Z-drugs will also be assessed. The Benjamini-Hochberg procedure will

be applied to these analyses to control the false discovery rate at 5% to take into account

the multiple tests being conducted [49].

11.1.2 Secondary analysis

Finally, we will examine the association between medication cessation and dementia

incidence. Here we will consider all medication exposures occurring during the exposure

period and in the 4 years before the index date. The association between time since

discontinuation (categorised) and dementia will be examined separately using conditional

logistic regression for (1) BZD/Z-drugs and (2) definite anticholinergics, adjusting for

confounders up to 4 years prior to the index date. Associations will also be provided according

to cumulative use pre discontinuation (in three groups of light, medium and heavy user,

defined by DDDs).

In sensitivity analyses we will adjust for confounders measured up to the index date (this will

include adjusting for new drug use in that period), due to being unable to adjust for

confounders at the exact date of medication cessation.

11.1.3 Missing data

In the primary analyses, patients with missing covariate data will be coded in a missing data

category. For covariates with at least 10% of patients with missing data, we will summarise the

characteristics of those with and without missing data, and examine sensitivity analyses using

(1) multiple imputations (5 sets) based on reasonable assumptions and subject to

computational feasibility, (2) single imputation, and (3) restricting to those with complete data.

Page 62 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


15

11.2 Limitations of the study design, data sources and analytic methods

11.2.1 Study design limitations

• The nested case-control study design includes prevalent medication users rather than

only new users of the medications so may over-represent those who tolerate the drugs

[50]. However, the nested case-control study design is best placed to deal with

complex time-varying and longer term medication exposure [51]. We also include

indications in a wide period including at least one year before the prescriptions, and

preliminary data analysis reveals that this is more likely to capture the medication

indications than through a new user design.

• Our results may suffer from protopathic bias, in which the use of these medications is

a marker of prodromal symptoms of dementia [20]. Examining medication use many

years before dementia diagnosis will help to limit this bias as well as subgroup analyses

of those who have depression, anxiety and insomnia whilst also adjusting for time

since diagnosis.

11.2.2 Analytic methods limitations

• Our dementia cases and controls could be misclassified. However, we will use read

codes included in the QOF business rules for dementia and only include dementia

diagnosed since this was introduced in the QOF, thus improving the quality of

dementia reporting. We also used an algorithm similar to one validated by GPs

confirming dementia in 95% of cases [21]. Using older CPRD data, another study found

a sensitivity and specificity of 100% in 150 cases and 50 controls sampled and sent GP

letters [22]. We will also assess what proportion of our cases have more than one

dementia related code including dementia diagnoses, symptoms, medications,

specialist referrals, neuroimaging assessments, or cognitive tests. We will also perform

two sensitivity analysis with more specific dementia definitions.

• We cannot be sure of the timing of dementia onset. By definition diagnosis in primary

care records is delayed as the patient will have had symptoms for some time before.

As a slowly developing degenerating disease, the disease onset is not a precise point in

time. Not examining medication exposures in the period 4 years prior to diagnosis for

our primary analysis reduces the impact of any errors in the recording of this timing.

• Dementia is also under-diagnosed in primary care [3], hence we could misclassify some

early dementias as controls which would have the effect of reducing the strength of

our associations.

• Our analysis will be subject to confounding by indication. We will address this by (1)

adjusting for many indications of the medications, (2) analysing subgroups with just

the main indication, and (3) additionally adjusting for many diseases and related

symptoms to best possibly capture the reasons for prescription, as many medications

will be prescribed outside of their guidelines [38].

• The ACB scale used to evaluate anticholinergic activity of medications has not been

validated against in vitro measures of anticholinergic activity, although it is uncertain

whether assays reflect anti-cholinergic activity in the brain, and not all relevant drugs

have been assayed in vitro [52]. The ACB scale was chosen as it is commonly use and

has been recently updated with respect to mediations currently used in the UK [27].

Due to some medications potentially being misclassified using this scale, we will also

perform sensitivity analyses using different anticholinergic scales.

11.2.3 Data source limitations

• We cannot be sure the patients take the prescribed medications, and adherence

would be difficult to ascertain. This would be more of a concern if our analysis was

examining patients prescribed one medication, as they may not have taken it, but as

Page 63 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


16

our primary hypotheses concentrate on those receiving multiple prescriptions (i.e. >90

days) we can be more sure the patient has taken at least some of them [53,54]. Having

prescription data also enables detailed and accurate information on dosing and

prescribing instructions.

• We may under-estimate medication use as we will lack data on medications purchased

over-the-counter and illegally, and those prescribed in secondary care. However, in

most cases of prescriptions given in secondary care the responsibility of longer term

prescribing is transferred to primary care via shared care arrangements, or equivalent.

Also, the majority of the medications we are interested in are prescription only and

the bulk of prescribing should be in primary care. We may have limited ability to

extrapolate the risks of OTC medication use with anticholinergic activity.

• Some confounder data will be missing for some patients (e.g. smoking, BMI, alcohol

use, blood pressure), however we will examine appropriate techniques to address this.

• There is always the possibility of residual confounding. We are missing information on

certain disease severities (e.g. depression scores), genetics (e.g. APOE E4 status), and

cognitive scores. However, CPRD contains information on all of the most important

confounding variables in these hypotheses.

• There could be issues over the quality of data recording in primary care records, but

we will use UTS data only and examine medication exposures after at least one year of

UTS to enable recording to have stabilised [23]. The CPRD has also been well validated

for a range of conditions [40,55], and we will use QOF definitions or validated read

code lists where possible.

12 Patient/ user group involvement The funding for this study was awarded by the Alzheimer’s Society grant advisory board which

has substantial lay representation.

As the funder of this study, The Alzheimer’s Society has appointed three Research Network

Volunteers to act as study monitors and service user representatives on our study steering

committee. These individuals have all acted as carers of people with dementia.

The monitors contribute to our protocol development by sharing their experiences of

psychoactive medication use and their view of the balance between the benefits of medication

use and potential cognitive decline. Monitors meet the study team to discuss study progress

every six months. They will be particularly concerned with the dissemination phase of our

study, making sure that lay summaries of results are accessible and avoid possible

misinterpretation.

13 Plans for disseminating and communicating study results Our communication plan is designed to gain maximum impact for our work among clinicians,

patient groups and researchers, and the Alzheimer’s Society are partners to ensure the public

and patient message can be developed. Specific dissemination activities are:

Academic dissemination. We will disseminate our findings though one or two peer-reviewed

academic papers in high impact medical journals, as well as including the findings in a planned

and funded systematic review and discussion articles to link our work to that of others at the

end of the project. Academic conference presentations will be made to clinical and public

health audiences. We aim to analyse and publish our findings within one year of receiving the

Page 64 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


17

CPRD data. Funding for this project currently ends in July 2017, and although we will also apply

for an Alzheimer’s Society dissemination grant, activity will be restricted beyond this point.

Medical practitioners, planners and policy makers. These groups have already supported our

study and asked to be kept informed of progress. We will write articles for professional

periodicals and give updates about the study on relevant websites (e.g. that of the British

Geriatrics Society, NHS Primary Care Commissioning website, and NHS Evidence) and via our

regional and professional networks. We will contribute study findings to the professional

training programmes at our Universities. Our previous experience of Department of Health

consultations will help us to contact key policy groups.

For patient and public groups. We will contribute to newsletters including those for carers and

the voluntary sector (e.g. Living with Dementia - Alzheimer's Society newsletter) and websites

(e.g. Dementia UK, Alzheimer’s Society). We will also present to lay groups advocating for

dementia and older people’s issues. We have comprehensive service user involvement and

through this we will explore all opportunities to disseminate study information and

implications to relevant people without dementia, people with dementia and their carers.

Website/social media. We have a project website at www.uea.ac.uk/drug-safety-and-

dementia/ to draw attention to the study and provide progress updates. We will use personal,

academic, institutional and study specific Twitter accounts to provide details of the

publication(s).

International dissemination. We have established links to groups such as the American

Geriatrics Society, the American Association of Geriatric Psychiatry, the European Delirium

Association and international public health bodies and we will produce briefings for these.

14 Protocol peer review

This exact protocol has not been peer reviewed by a committee; however an earlier simplified

version was peer reviewed by our study funder, the Alzheimer’s Society, when submitted as

part of a larger project. Our research grant application underwent scientific peer review and

lay review by their Research Network volunteers. We provided replies and the application was

then considered by a Grant Advisory Board. The scientific peer reviewers provided some

specific comments regarding the CPRD analysis, such as that not all dementia patients will

have a diagnosis, concerns about protopathic bias, and that anticholinergic burden scales

differ. The project was successful, particularly at the Grant Advisory Board stage, and was

funded in full.

Page 65 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


18

15 References

1 Howick J, Glasziou P, Aronson JK. The evolution of evidence hierarchies: what can Bradford

Hill’s ‘guidelines for causation’ contribute? JRSM 2009;102:186–94.

doi:10.1258/jrsm.2009.090020

2 Alzheimer’s Society. Dementia UK: Update [Internet]. Available from:

http://www.alzheimers.org.uk/site/scripts/download_info.php?fileID=2323 (accessed 22

Jan2015).

3 Matthews FE, Arthur A, Barnes LE, et al. A two-decade comparison of prevalence of

dementia in individuals aged 65 years and older from three geographical areas of England:

results of the Cognitive Function and Ageing Study I and II. The Lancet 2013;382:1405–12.

doi:10.1016/S0140-6736(13)61570-6

4 Ferri CP, Prince M, Brayne C, et al. Global prevalence of dementia: a Delphi consensus

study. The Lancet 2006;366:2112–7.

5 Tannenbaum C, Paquette A, Hilmer S, et al. A Systematic Review of Amnestic and Non-

Amnestic Mild Cognitive Impairment Induced by Anticholinergic, Antihistamine, GABAergic

and Opioid Drugs. Drugs Aging 2012;29:639–58.

6 Billioti de Gage S, Begaud B, Bazin F, et al. Benzodiazepine use and risk of dementia:

prospective population based study. BMJ 2012;345:e6231–e6231. doi:10.1136/bmj.e6231

7 Billioti de Gage S, Moride Y, Ducruet T, et al. Benzodiazepine use and risk of Alzheimer’s

disease: case-control study. BMJ 2014;349:g5205–g5205. doi:10.1136/bmj.g5205

8 Wu C-S, Ting T-T, Wang S-C, et al. Effect of Benzodiazepine Discontinuation on Dementia

Risk. Am J Geriatr Psychiatry 2011;19:151–9. doi:10.1097/JGP.0b013e3181e049ca

9 Lagnaoui R, Bégaud B, Moore N, et al. Benzodiazepine use and risk of dementia: A nested

case–control study. J Clin Epidemiol 2002;55:314–8. doi:10.1016/S0895-4356(01)00453-X

10 Bartlett G, Abrahamowicz M, Tamblyn R, et al. Longitudinal patterns of new

Benzodiazepine use in the elderly. Pharmacoepidemiol Drug Saf 2004;13:669–82.

doi:10.1002/pds.908

11 Prescribing and Primary Care team, Health and Social Care Information Centre.

Prescription Cost Analysis England 2013. 2014.

12 Fox C, Smith TO, Maidment I, et al. An Updated Systematic Review of the Impact of

Medications with Anti-Cholinergic Activity on Cognitive function and Mortality. In:

Alzheimer’s Association International Conference. Boston, USA: 2013.

http://www.rcpsych.ac.uk/pdf/Posters%20Final.pdf (accessed 26 Jul2013).

13 Gray SL, Anderson ML, Dublin S, et al. Cumulative use of strong anticholinergics and

incident dementia: A prospective cohort study. JAMA Intern Med Published Online First: 26

January 2015. doi:10.1001/jamainternmed.2014.7663

14 Fox C, Richardson K, Maidment ID, et al. Anticholinergic Medication Use and Cognitive

Impairment in the Older Population: The Medical Research Council Cognitive Function and

Ageing Study. J Am Geriatr Soc 2011;59:1477–83. doi:10.1111/j.1532-5415.2011.03491.x

Page 66 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


19

15 Richardson K, Bennett K, Maidment I, et al. Use of medications with anti-cholinergic

activity and self-reported injurious falls in the older aged community-dwelling population. J

Am Geriatr Soc 2015;(In Press).

16 Ness J, Hoth A, Barnett MJ, et al. Anticholinergic medications in community-dwelling

older veterans: Prevalence of anticholinergic symptoms, symptom burden, and adverse

drug events. Am J Geriatr Pharmacother 2006;4:42–51.

doi:10.1016/j.amjopharm.2006.03.008

17 Cai X, Campbell N, Khan B, et al. Long-term anticholinergic use and the aging brain.

Alzheimers Dement Published Online First: November 2012. doi:10.1016/j.jalz.2012.02.005

18 Ancelin ML, Artero S, Portet F, et al. Non-degenerative mild cognitive impairment in

elderly people and use of anticholinergic drugs: longitudinal cohort study. BMJ

2006;332:455–9. doi:10.1136/bmj.38740.439664.DE

19 Ownby RL, Crocco E, Acevedo A, et al. Depression and risk for Alzheimer disease:

systematic review, meta-analysis, and metaregression analysis. Arch Gen Psychiatry

2006;63:530–8. doi:10.1001/archpsyc.63.5.530

20 Masters MC, Morris JC, Roe CM. ‘Noncognitive’ symptoms of early Alzheimer disease A

longitudinal analysis. Neurology 2015;:10.1212/WNL.0000000000001238.

doi:10.1212/WNL.0000000000001238

21 Imfeld P, Brauchli Pernus YB, Jick SS, et al. Epidemiology, Co-Morbidities, and

Medication Use of Patients with Alzheimer’s Disease or Vascular Dementia in the UK. J

Alzheimers Dis 2013;35:565–73. doi:10.3233/JAD-121819

22 Dunn N, Holmes C, Mullee M. Does lithium therapy protect against the onset of

dementia? Alzheimer Dis Assoc Disord 2005;19:20–2.

23 Lewis JD, Bilker WB, Weinstein RB, et al. The relationship between time since

registration and measured incidence rates in the General Practice Research Database.

Pharmacoepidemiol Drug Saf 2005;14:443–51. doi:10.1002/pds.1115

24 Flanders WD, Louv WC. The Exposure Odds Ratio in Nested Case-Control Studies with

Competing Risks. Am J Epidemiol 1986;124:684–92.

25 Beaumont JJ, Steenland K, Minton A, et al. A Computer Program for Incidence Density

Sampling of Controls in Case-Control Studies Nested Within Occupational Cohort Studies.

Am J Epidemiol 1989;129:212–9.

26 Steenland K, Deddens JA. Increased Precision Using Countermatching in Nested Case-

Control Studies. Epidemiology 1997;8:238–42.

27 Campbell NL, Maidment I, Fox C, et al. The 2012 Update to the Anticholinergic

Cognitive Burden Scale. J Am Geriatr Soc 2013;61:S1–232. doi:10.1111/jgs.12263

28 Boustani M, Campbell N, Munger S, et al. Impact of anticholinergics on the aging brain:

a review and practical application. Aging Health 2008;4:311–20.

doi:10.2217/1745509X.4.3.311

29 Rudolph JL, Salow MJ, Angelini MC, et al. The anticholinergic risk scale and

anticholinergic adverse effects in older persons. Arch Intern Med 2008;168:508–13.

doi:10.1001/archinternmed.2007.106

Page 67 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


20

30 Carnahan RM, Lund BC, Perry PJ, et al. The Anticholinergic Drug Scale as a Measure of

Drug-Related Anticholinergic Burden: Associations With Serum Anticholinergic Activity. J

Clin Pharmacol 2006;46:1481–6. doi:10.1177/0091270006292126

31 Narayan SW, Hilmer SN, Horsburgh S, et al. Anticholinergic Component of the Drug

Burden Index and the Anticholinergic Drug Scale as Measures of Anticholinergic Exposure in

Older People in New Zealand: A Population-Level Study. Drugs Aging 2013;30:927–34.

doi:10.1007/s40266-013-0111-y

32 Bhaskaran K, Forbes HJ, Douglas I, et al. Representativeness and optimal use of body

mass index (BMI) in the UK Clinical Practice Research Datalink (CPRD). BMJ Open

2013;3:e003389. doi:10.1136/bmjopen-2013-003389

33 Bhaskaran K, Douglas I, Forbes H, et al. Body-mass index and risk of 22 specific cancers:

a population-based cohort study of 5·24 million UK adults. The Lancet 2014;384:755–65.

doi:10.1016/S0140-6736(14)60892-8

34 Lewis JD, Brensinger C. Agreement between GPRD smoking data: a survey of general

practitioners and a population-based survey. Pharmacoepidemiol Drug Saf 2004;13:437–41.

doi:10.1002/pds.902

35 Martinez C, Rietbrock S, Wise L, et al. Antidepressant treatment and the risk of fatal

and non-fatal self harm in first episode depression: nested case-control study. BMJ

2005;330:389. doi:10.1136/bmj.330.7488.389

36 Coupland C, Dhiman P, Morriss R, et al. Antidepressant use and risk of adverse

outcomes in older people: population based cohort study. BMJ 2011;343:d4551.

37 Osborn DPJ, Levy G, Nazareth I, et al. Relative risk of cardiovascular and cancer

mortality in people with severe mental illness from the United Kingdom’s General Practice

Rsearch Database. Arch Gen Psychiatry 2007;64:242–9. doi:10.1001/archpsyc.64.2.242

38 Marston L, Nazareth I, Petersen I, et al. Prescribing of antipsychotics in UK primary

care: a cohort study. BMJ Open 2014;4:e006135. doi:10.1136/bmjopen-2014-006135

39 Khan NF, Perera R, Harper S, et al. Adaptation and validation of the Charlson Index for

Read/OXMIS coded databases. BMC Fam Pract 2010;11:1. doi:10.1186/1471-2296-11-1

40 Khan NF, Harrison SE, Rose PW. Validity of diagnostic coding within the General

Practice Research Database: a systematic review. Br J Gen Pract 2010;60:e128–36.

doi:10.3399/bjgp10X483562

41 Martinez C, Assimes TL, Mines D, et al. Use of venlafaxine compared with other

antidepressants and the risk of sudden cardiac death or near death: a nested case-control

study. BMJ 2010;340:c249.

42 Martín-Merino E, Ruigómez A, Wallander M-A, et al. Prevalence, incidence, morbidity

and treatment patterns in a cohort of patients diagnosed with anxiety in UK primary care.

Fam Pract 2009;:cmp071. doi:10.1093/fampra/cmp071

43 Shah SM, Carey IM, Harris T, et al. Initiation of Psychotropic Medication after Partner

Bereavement: A Matched Cohort Study. PLoS ONE 2013;8:e77734.

doi:10.1371/journal.pone.0077734

Page 68 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


21

44 Yang C, Jick SS, Jick H. LIpid-lowering drugs and the risk of depression and suicidal

behavior. Arch Intern Med 2003;163:1926–32. doi:10.1001/archinte.163.16.1926

45 Khadjesari Z, Marston L, Petersen I, et al. Alcohol consumption screening of newly-

registered patients in primary care: a cross-sectional analysis. Br J Gen Pract 2013;63:e706–

12. doi:10.3399/bjgp13X673720

46 Delaney JAC, Moodie EEM, Suissa S. Validating the effects of drug treatment on blood

pressure in the General Practice Research Database. Pharmacoepidemiol Drug Saf

2008;17:535–45. doi:10.1002/pds.1553

47 Davies NM, Kehoe PG, Ben-Shlomo Y, et al. Associations of Anti-Hypertensive

Treatments with Alzheimer’s Disease, Vascular Dementia, and Other Dementias. J

Alzheimers Dis 2011;26:699–708. doi:10.3233/JAD-2011-110347

48 Townsend P, Phillimore P, Beattie A, et al. Health and deprivation: inequality and the

North. Croom Helm London 1988.

49 Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and

Powerful Approach to Multiple Testing. J R Stat Soc Ser B Methodol 1995;57:289–300.

doi:10.2307/2346101

50 Ray WA. Evaluating Medication Effects Outside of Clinical Trials: New-User Designs. Am

J Epidemiol 2003;158:915–20. doi:10.1093/aje/kwg231

51 Essebag V, Genest J, Suissa S, et al. The nested case-control study in cardiology. Am

Heart J 2003;146:581–90. doi:10.1016/S0002-8703(03)00512-X

52 Lampela P, Lavikainen P, Garcia-Horsman JA, et al. Anticholinergic Drug Use, Serum

Anticholinergic Activity, and Adverse Drug Events Among Older People: A Population-Based

Study. Drugs Aging 2013;30:321–30. doi:10.1007/s40266-013-0063-2

53 Fischer MA, Stedman MR, Lii J, et al. Primary Medication Non-Adherence: Analysis of

195,930 Electronic Prescriptions. J Gen Intern Med 2010;25:284–90. doi:10.1007/s11606-

010-1253-9

54 Weich S, Pearce HL, Croft P, et al. Effect of anxiolytic and hypnotic drug prescriptions

on mortality hazards: retrospective cohort study. BMJ 2014;348:g1996–g1996.

doi:10.1136/bmj.g1996

55 Herrett E, Thomas SL, Schoonen WM, et al. Validation and validity of diagnoses in the

General Practice Research Database: a systematic review. Br J Clin Pharmacol 2010;69:4–14.

doi:10.1111/j.1365-2125.2009.03537.x

56 Wallander M-A, Johansson S, Ruigómez A, et al. Morbidity Associated With Sleep

Disorders in Primary Care: A Longitudinal Cohort Study. Prim Care Companion J Clin

Psychiatry 2007;9:338–45.

Page 69 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


22

16 Appendix

16.1 Code lists

16.1.1 Alcoholism or alcohol related disease

The alcoholism code lists includes alcoholism, alcoholic dementia, alcoholic liver disease, and

alcoholic cirrhosis.

Alcoholism read codes: 1462, E01, E23..00, E23..11, E23z.00, E230, E231, Eu102, Eu103, Eu104,

Eu1051, Eu10611, Eu10711, F11x000, F11x011, F144000, F375.00, F394100, G555.00,

G852300, J153.00, J610.00, J611.00, J612.00, J612000, J613.00, J613000, J617.00, J617000,

J671000, ZV11300

16.1.2 Dementia

The dementia code list is the QOF definition excluding alcohol induced dementia, and is very

similar to lists published and validated [21].

Dementia diagnosis read codes:

E00, E02y1, E041, Eu00, Eu01, Eu02, Eu041, F110, F111, F112, F116, Fyu3000

16.1.3 Depression

The depression code list is the QOF definition excluding depression in dementia, but also

including ‘Depressive psychoses’.

Depression read codes:

E11..12, E112, E113, E118., E11y2, E11z2, E130., E135., E2003, E291., E2B.., E2B1, Eu204,

Eu251, Eu32, Eu33, Eu341, Eu412

16.1.4 Anxiety

The anxiety codes are essentially the ones used in a published and validated study, but

excluding anxiety symptoms, interventions and phobias [42].

Anxiety read codes: 1466, E200, E292000, Eu054, Eu34114, Eu41, Eu51511, Eu60600.

16.1.5 Insomnia

The insomnia/sleep disturbance codes are essentially the ones used in a published study, but

including related sleep disorders/disturbance and excluding sleep apnoea [56].

Insomnia read codes: 1B1B (excluding 1B1B.12), 1B1Q, E274, Eu51, Fy0..00, Fy00, Fy01, Fy02,

R005 (excluding R0051, R0053).

Page 70 of 70


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Documents

BMJ · 2018-04-30 · Malaz Boustani, School of Medicine, University of Indiana, Indiana United States Professor of Medicine Louise Robinson, Newcastle University Institute for Ageing,