Body mass index in adults with intellectual disability: distribution, associations and service implications: a population-based prevalence study

Body mass index in adults with intellectual disability:distribution, associations and service implications: apopulation-based prevalence study

S. Bhaumik,1,2 J. M. Watson,1,2 C. F. Thorp,2 F. Tyrer2 & C.W. McGrother2

1 Learning Disability Service, Leicestershire Partnership NHS Trust, Leicester Frith Hospital, Leicester, UK2 Department of Health Sciences, University of Leicester, Leicester, UK

Abstract

Background Previous studies of weight problemsin adults with intellectual disability (ID) have gen-erally been small or selective and given conflictingresults. The objectives of our large-scale studywere to identify inequalities in weight problemsbetween adults with ID and the general adultpopulation, and to investigate factors associatedwith obesity and underweight within the IDpopulation.Methods We undertook a population-based preva-lence study of 1119 adults with ID aged 20 andover on the Leicestershire Learning Disability Reg-ister who participated in a programme of universalhealth checks and home interviews with theircarers. We performed a cross-sectional analysis ofthe register data and compared the observed andexpected prevalences of body mass index categoriesin the ID and general populations using indirect

standardisation for age. We used logistic regressionto evaluate the association of a range of probabledemographic, physical, mental and skills attributeswith obesity and underweight.Results In those aged 25 and over, the standard-ised morbidity ratio (SMR) for obesity was 0.80

(95% CI 0.64–1.00) in men and 1.48 (95% CI1.23–1.77) in women. The SMR for underweightwas 8.44 (95% CI 6.52–10.82) in men and 2.35

(95% CI 1.72–3.19) in women. Among those aged20 and over, crude prevalences were 20.7% forobesity, 28.0% for overweight, 32.7% for normalweight and 18.6% for underweight. Obesity wasassociated with living independently/with family,ability to feed/drink unaided, being female, hyper-tension, Down syndrome and the absence of cere-bral palsy. Underweight was associated withyounger age, absence of Down syndrome and nottaking medication.Conclusion Obesity in women and underweight inboth men and women was more common in adultswith ID than in the general population after con-trolling for differences in the age distributionsbetween the two populations. The associated factorssuggest opportunities for targeting high-risk groupswithin the ID population for lifestyle and behaviourmodification.

Correspondence: Dr Sabyasachi Bhaumik, Learning DisabilityService, Leicestershire Partnership NHS Trust, Leicester FrithHospital, Groby Road, Leicester, LE3 9QF, UK (e-mail:[email protected]).Ethics approval: The Leicestershire Learning Disability Registerhas Leicestershire Ethics Committee approval. Details of funding:This study was funded by Leicestershire NHS Partnership Trustand the Department of Health, Policy Research Programme.

Journal of Intellectual Disability Research doi: 10.1111/j.1365-2788.2007.01018.x

volume 52 part 4 pp 287–298 april 2008287

© 2007 The Authors. Journal Compilation © 2007 Blackwell Publishing Ltd

mailto:[email protected]

Keywords body mass index, health inequality,intellectual disability, morbidity, obesity,underweight

Introduction

There are major concerns about the impact ofobesity on health and health inequalities globally(Haslam & James 2005) and in the UK (NationalAudit Office 2001; Department of Health 2003).Obesity increases the risk of developing a broadspectrum of cardiovascular, pulmonary, metabolicand neoplastic diseases, osteoarthritis, impaired fer-tility and complications of pregnancy and anaesthe-sia (National Audit Office 2001; Haslam & James2005; Haslam et al. 2006). Obesity has additionalpsychological, social and economic consequences.Among preventable causes of disease and prematuredeath, obesity is overtaking smoking (Haslam et al.2006). Risk factors for obesity in the general popu-lation include age, sex, ethnicity, socio-economicgroup, education, diet, physical inactivity, smokingstatus and certain drugs (National Audit Office2001; Haslam & James 2005). Body mass index(BMI) is a general indicator of health risks due toexcess weight (World Health Organisation ExpertCommittee 1995).

In the general population, 21% are obese and 4%underweight (Department of Health 1998). Untilrecently, it was unclear whether weight problemswere more common in people with intellectual dis-ability (ID) (Rimmer et al. 1995; Gravestock 2000;Hove 2004; Moran et al. 2005). However, recentevidence suggests that obesity and underweight maybe more prevalent than in the general population.One large UK study of adults in supported accom-modation found that obesity was more common inwomen with ID; and overweight was more prevalentin both men and women with ID (Emerson 2005).A study of adults with self-reported ID in the USAfound that obesity was more prevalent in both menand women (Yamaki 2005). Studies suggest that25–48% of adults with Down syndrome are obese(Prasher 1995; Gravestock 2000; Melville et al.2005). A wide range of estimates has been reportedfor weight problems in the whole ID population:from 2% to 35% for obesity and 5% to 43% for sig-nificant underweight (Gravestock 2000; Emerson2005; Yamaki 2005). Few studies have investigated

factors associated with obesity and underweight inthe ID population.

People with ID have significantly more healthproblems than the general population (The Foun-dation for People with Learning Disabilities 2001;Department of Health 2001). About half have amajor psychiatric or behaviour problem requiringspecialist help. Around 25% have epilepsy. At least33% have a sensory impairment. Approximately40% have associated major physical disabilities ofmobility and continence, 50–90% have communica-tion difficulties. Most have limited understanding ofhealthy lifestyle choices and are dependent oncarers to meet their often complex needs.

Health inequalities in the ID population havemultiple causes, including intellectual and otherimpairments, socio-economic factors, which limitopportunities and lifestyle options, lack of resourcesfor services, and poor standards of health care (Dis-ability Rights Commission 2006; Mencap 2007).Equal Treatment: Closing the Gap (Disability RightsCommission 2006) and Death by Indifference(Mencap 2007) have exposed widespread ignoranceand discrimination throughout healthcare servicestowards people with ID and their carers. A keyprinciple of Tackling Health Inequalities (Departmentof Health 2003) is working through the mainstreamby making services more responsive to the needs ofdisadvantaged populations. The government nowrecognises the importance of disability as well associal deprivation in relation to inequalities. Thus,the needs of people with ID and their carers are nolonger the sole concern of specialist services, butare also relevant to mainstream healthcare profes-sionals, especially those in primary care.

The overall aim of our study was to describe thedistribution, associations and service implications ofBMI in a population of adults with ID.The objec-tives were to identify inequalities in weight problemsbetween adults with ID and the general adult popu-lation, to investigate factors associated with obesityand underweight in the ID population, and tosuggest ways of addressing these health inequalities.

Method

This was a population-based prevalence study usingthe Leicestershire Learning Disability Register. The

288Journal of Intellectual Disability Research volume 52 part 4 april 2008

S. Bhaumik et al. • BMI in adults with ID


register is an open cohort of adults with ID aged 19

and over living within the unitary authorities ofLeicester, Leicestershire and Rutland (Watson2003). The register was set up in 1987 to provide asampling frame for research and epidemiologicalinformation for service planning and delivery. ID isdefined as a moderate, severe or profound develop-mental intellectual impairment (World HealthOrganisation 1992) with associated adaptive behav-iour problems (Heber 1959) and a need or likelyneed for long-term support. A comprehensivenetwork of specialist services continually notifieseligible adults to the register, which has an esti-mated 95% ascertainment (Watson 2003). The reg-ister supports a rolling programme of in-depthstructured home interviews with carers every five toseven years, which incorporate questionnaires devel-oped for this population, including the DisabilityAssessment Schedule (Holmes et al. 1982). Uptakeis 95% for first interviews, 96% for second ones and98% for third (Watson 2003).

Between 1987 and 1999, the register offered four-yearly health checks to adults on the register aged20 and over. A community physician used a stan-dardised form to collect demographic details andscreen for a wide range of physical and mentalhealth problems. Whenever possible, the medicalexamination included measuring an individual’sheight, weight and blood pressure, and checking aurine sample for glycosuria. The acceptance ratewas 85%. Unfortunately, funding for health checksceased in 1999 and no subsequent data werecollected.

The total study population comprised all theadults on the register who had had a health checkwithin two years of 1 July 1998 and a home inter-view within three years of that health check. Weexcluded individuals from the study if any of thekey variables were missing: age, sex, height, weightand type of accommodation. We collected interviewdata about ethnicity, level of understanding, physi-cal disabilities, epilepsy, relevant skills deficits,eating problems, other behaviour problems, autism,participation in sports, smoking status and takingmedication (Box 1). We obtained additional datafrom the health checks about comorbidity.

We calculated the BMI for each individual usingthe Quetelet formula (World Health OrganisationExpert Committee 1995). We categorised males and

females separately as underweight, normal weight(desirable weight), overweight and obese as definedin Box 1. We derived measures of prevalence bycounting the numbers of males and females withineach BMI category and calculating proportions. Wethen restricted the study population to individualsaged 25 and over to match the national statisticsand compared our results with data for the generalpopulation given in the Health Survey for England1998 (Department of Health 1998). We ascertaineddifferences in the proportions of men and womenwithin the populations using the chi-squared test.For each BMI category, we derived age-adjustedstandardised morbidity ratios (SMRs) stratified bysex and Down syndrome. We report SMRs as ratiosrather than percentages.

Using Stata version 9 software, we applied logis-tic regression modelling to the data for the wholestudy population to derive odds ratios (OR) and95% confidence intervals (CI) for the relative risksof obesity and underweight compared with normalweight for a range of possible associations. Theseincluded demographic characteristics, skills levels,disabilities, eating and other behaviour problems,comorbidity and other health indicators (Box 1). Wecalculated levels of significance (P-values) by apply-ing the likelihood ratio test. Empirical evidence sug-gests that age, sex, ethnicity, Down syndrome andsmoking status are related to BMI (Department ofHealth 1998; National Audit Office 2001; Haslam &James 2005). Thus, we entered all these factors intoa multiple logistic regression model to identify anyindependent associations with obesity and under-weight. The remaining variables were added to themodel only if they were significant at the 5% level(P � 0.05) in the univariate analysis.

Results

Characteristics of the ID study population aged20 and over

A total of 661 (59%) men and 458 (41%) womenon the Leicestershire Learning Disability Registerwere eligible for inclusion in the study (Table 1). Asexpected, our study population was relativelyyoung, with 74% aged under 50. 88% were white,59% lived in residential care. Half the group haddifficulty preparing simple foods independently and




Box 1 Definitions used in this study

Data from the home interviews with carersEthnicity White British, Irish or other European

South Asian Indian, Pakistani or BangladeshiOther Black, Chinese, mixed or other

Level of understanding Fairly good Understanding information about things outsidehis/her immediate experience, e.g. major itemsof current news

Poor/moderate Understanding comments, questions and instructionsrelated to his/her personal needs and experiences

Very poor Understanding a few simple commands at mostRelevant skills deficits Cannot manage the following without help:

• Feeding/drinking • Preparing simple foodsPhysical disabilities Poor vision Blind, almost blind or partially sighted

Poor hearing Deaf or poor hearingDifficulty with walking Unable to walk half a mile unaidedWetting Wetting at least once 1 weekSoiling Soiling at least once 1 week

Epilepsy A positive response to the question ‘Does he/she suffer from epilepsy?’ or reportedexperience of seizures occasionally or more often or reported taking of medication toprevent or control seizures

Eating problems Self-induced vomitingContinual or psychogenic eating/drinkingPica: compulsive eating of non-nutritive substances, e.g. dirt, hair, paint

Behaviour problems Any of the following if severe and/or frequent (at least three episodes 1 week)• Physical aggression to others • Destructive behaviour• Excessive activity • Seeks attention• Self-injury • Wanders or runs away• Screams or makes disturbing noises • Tantrums or verbal abuse• Disturbs others at night • Scatters or throws objects• Antisocial or delinquent behaviour • Untruthful• Inappropriate sexual behaviour • Uncooperative• Difficult or offensive personal habits • Other behaviour problems

Autism Four or more of the following autistic traits:• Poor speech • Poor quality of social interaction • Lack of empathy• Obsessive behaviour • Simple stereotypies

Sports participation • None • Occasional • RegularSmoker Smokes cigarettes at all at time of interviewTaking medication Reported taking of medication for anxiety, depression, epilepsy, behaviour problems and/or

sleep problemsData from the health checks

Type of accommodation attime of health check

Lives independently/with family Lives alone or with spouse/partner, friends, parents,other relatives, guardian, foster carer

Residential care Lives in staffed home, hostel or hospital (NHS, SocialServices, private or voluntary sector)

Body mass index Using the Quetelet formula4:BMI = weight (kilograms)

height2 (metres2)

Underweight BMI � 20

Normal weight BMI 20.1 to 25

Overweight BMI 25.1 to 30

Obese BMI > 30

Hypertension Systolic blood pressure �140 mmHg and/or diastolic blood pressure �90 mmHgGlycosuria Urine sample tested showed more than a trace of sugarComorbidity ICD-9/ICD-10 criteria for specific conditions, e.g. Down syndrome, cerebral palsy

NHS, National Health Survey; Leicestershire Learning Disability Register.




Table 1 General characteristics of the study population

Characteristics Males (n = 661) % Females (n = 458) % Persons (n = 1119) %

Age at health check20–29 25.9 25.8 25.830–39 23.4 25.1 24.140–49 22.3 25.3 23.6�50 28.3 23.8 26.5

EthnicityWhite 89.6 86.7 88.4South Asian 8.6 10.0 9.2Other 1.8 3.3 2.4

Accommodation at health checkWith family/alone 37.4 46.3 41.0Residential care 62.6 53.7 59.0

Level of understandingFairly good 43.6 43.2 43.4Poor/moderate 46.4 47.4 46.8Very poor 10.0 9.4 9.7

Skills deficitsFeeding/drinking 12.0 14.4 13.0Preparing food 50.5 48.3 49.6

Physical disabilitiesPoor vision 10.7 9.8 10.4Poor hearing 7.0 9.6 8.0Difficulty walking 15.0 19.2 16.7Wetting 13.3 18.6 15.5Soiling 4.5 6.6 5.4

Behaviour problemsSelf-induced vomiting 0.8 1.1 0.9Continual eating/drinking 2.6 1.5 2.1Pica 2.1 0.4 1.4Other 48.3 44.3 46.6

ComorbidityDown syndrome 18.5 22.5 20.1Cerebral palsy 3.8 4.8 4.3Epilepsy 26.5 26.0 26.3Autism 4.1 3.5 3.8Hypertension 40.1 33.8 37.5*Glycosuria 4.2 2.8 3.6†

Other health indicatorsSports participation:

none 71.4 75.3 73.0occasional 10.0 9.0 9.6regular 18.6 15.7 17.4

Smoker 37.4 20.3 30.7‡

Taking medication 51.1 44.3 48.3BMI categories

Underweight 20.7 15.5 18.6Normal weight 34.6 29.9 32.7Overweight 30.0 25.1 28.0Obese 14.7 29.5 20.7

* Missing data for 27 males, 20 females, 47 (4%) persons; numerator for % is all those for whom there are data on blood pressure.† Missing data for 132 males, 107 females, 239 (21%) persons; numerator for % is all those for whom there are data on glycosuria.‡ Missing data for 418 males, 300 females, 718 (64%) persons; numerator for % is all those for whom there are data on smoking status.BMI, body mass index.




13% feeding/drinking unaided. 4% had an eatingproblem and 47% had other behaviour problems. Afifth had Down syndrome. Almost half were takingmedication.

Overall, 18.6% of the study population wereunderweight, 32.7% normal weight, 28.0% over-weight and 20.7% obese. Obesity was almost twiceas prevalent in women (29%) as in men (15%)(P < 0.001). Underweight was more prevalent inmen (21%) than in women (16%) (P = 0.03). Therewere no significant sex differences in the propor-tions of men and women who were normal weight(P = 0.10) or overweight (P = 0.08).

With increasing age, the prevalence of under-weight decreased in both sexes. 32% of men aged20–29 were underweight compared with 15% ofthose aged 50 and over. The equivalent figures forwomen were 25% and 9% respectively.

Comparison between the ID and generalpopulations aged 25 and over

In those aged 25 and over, 15% of men with IDwere obese compared with 19% of men in thegeneral population. The corresponding prevalencesfor women were 32% and 23% respectively. Incontrast, 19% of men with ID were underweight

compared with 2% of men in the general popula-tion. The corresponding prevalences for womenwere 12% and 5%. In both populations aged 25 andover, a higher proportion of women was obese thanmen; this was particularly apparent in women withID.

These crude differences were reflected in thefindings after standardisation for age. The SMR forobesity was slightly lower in men with ID comparedwith the general population (SMR 0.80; 95% CI0.64–1.00), but elevated significantly in women withID (SMR 1.48; 95% CI 1.23–1.77) (Fig. 1). TheSMR for underweight was eight times higher thanexpected in men with ID (SMR 8.44; 95% CI 6.52–10.92) and twice as high in women with ID (SMR2.35; 95% CI 1.72–3.19). Adults with Down syn-drome were more likely than expected to be obese(SMR 1.39; 95% CI 1.06–1.82) or underweight(SMR 2.17; 95% CI 1.32–3.58).

Comparisons within the ID study population

Factors associated with obesity

In the univariate analysis, a range of factors wasassociated with obesity, including relatively mild IDbut not age (Table 2). After adjustment, independent

0.5

1.0

1.5

2.0

2.5

4.0

5.0

11.0

Sta

nd

ard

ized

mo

rbid

ity

rati

o

3.0

6.07.08.09.0

Generalpopulation

MalesFemalesDown syndromeAll persons

63796221Down syndrome

208277316151All persons

12310311846Females

85174198105Males

ObeseOverweightNormal weightUnderweightNumbers in IDpopulation Figure 1 Standardised morbidity ratios*

for body mass index among adults withintellectual disability (ID) aged 25 andover based on the Health Survey forEngland 1998 (Department of Health1998). *With 95% confidence intervals.




associations with obesity were found for type ofaccommodation, feeding skills, sex, hypertension,Down syndrome and cerebral palsy.Those livingindependently/with family were more than threetimes as likely to be obese than those in residentialcare (OR 3.28; 95% CI 2.12–5.08).Women werealmost three times as likely to be obese as men (OR0.36; 95% CI 0.25–0.53 for men). Adults with Downsyndrome had an increased risk of obesity (OR 2.30;95% CI 1.40–3.80), as did those with hypertension(OR 2.67; 95% CI 1.76–4.06). Conversely, adultswith cerebral palsy were considerably less likely to beobese (OR 0.07; 95% CI 0.01–0.59).

Factors associated with underweight

In the univariate analysis, underweight was directlyassociated with living independently/with family,needing help with feeding/drinking, difficulty withwalking, younger age and very poor level of under-standing (Table 3). Hypertension, Down syndromeand taking medication were inversely associatedwith underweight.

After adjustment, younger age, Down syndromeand taking medication were the only factors associ-ated with underweight. The risk of underweightdecreased with increasing age, showing a gradient

Table 2 Factors associated with obesity in adults with intellectual disability* (sample limited to 598 adults who were obese or normal weightonly)

Characteristics Total

Obesity Unadjusted data Adjusted data†

n (%) OR (95% CI) P-value OR (95% CI) P-value

598 232 (39%)Sex

Male 326 97 (50%) 0.43 (0.31–0.60) <0.001 0.36 (0.25–0.53) <0.001Female 272 135 (30%) 1.00 (reference) 1.00 (reference)

Accommodation at health checkWith family/alone 254 137 (54%) 3.07 (2.18–4.32) <0.001 3.28 (2.12–5.08) <0.001Residential care 344 95 (28%) 1.00 (reference) 1.00 (reference)

Level of understandingFairly good 259 114 (44%) 1.00 (reference) <0.001 1.00 (reference) 0.34Poor/moderate 293 111 (38%) 0.78 (0.55–1.09) 0.83 (0.53–1.28)Very poor 46 7 (15%) 0.23 (0.10–0.53) 0.48 (0.17–1.34)

Skills deficitsFeeding/drinking 72 13 (18%) 0.31 (0.17–0.58) <0.001 0.40 (0.19–0.86) 0.02Preparing food 283 93 (33%) 0.62 (0.44–0.86) 0.005 0.99 (0.63–1.55) 0.96

Behaviour problemsOther 272 93 (34%) 0.70 (0.50–0.98) 0.03 0.97 (0.64–1.46) 0.87

ComorbidityDown syndrome 125 63 (50%) 1.83 (1.23–2.72) 0.003 2.30 (1.40–3.80) 0.001Cerebral palsy 20 1 (5%) 0.08 (0.01–0.59) <0.001 0.07 (0.01–0.59) <0.001Hypertension:

present 232 116 (50%) 2.15 (1.52–3.03) <0.001 2.67 (1.76–4.06) <0.001absent 340 108 (32%) 1.00 (reference) 1.00 (reference)missing data 26 8 (31%) 0.95 (0.40–2.26) 1.51 (0.56–4.12)

Other health indicatorsTaking medication 298 102 (34%) 0.68 (0.49–0.95) 0.02 1.37 (0.88–2.14) 0.17

* All variables described in Box 1 were entered into the unadjusted analysis; only those that were significant at the 5% level (P < 0.05) areshown here.† We adjusted the multivariable model for age, sex, ethnicity, Down syndrome, smoking status and all the remaining significant factorsshown in the table.OR, odds ratio; CI, confidence interval.




effect from 1.00 for those in their 20s to 0.37 forthose aged 50 and over. Individuals with Down syn-drome (OR 0.46; 95% CI 0.26–0.81) or takingmedication (OR 0.57; 95% CI 0.38–0.87) were lesslikely to be underweight.

Discussion

Compared with the general population, women withID (but not men) were more likely to be obese(SMR 1.48), as were people with Down syndrome(SMR 1.39). Underweight was also relativelycommon among adults with ID. Contrary to previ-ous reports (Gravestock 2000; Hove 2004; Moran

et al. 2005), we found the largest differences in BMIbetween the ID and general populations in theunderweight category. The SMR for underweightwas eight times higher than expected in men, twiceas high in women and four times higher overall. TheSMR for underweight was twice as high as expectedin adults with Down syndrome.

Within the ID population, the crude prevalenceestimate for obesity was 21%, which lies in the mid-range of previous estimates, which ranged from 2%to 35%. Similarly, the crude prevalence estimatefor underweight was 19% compared with 5% to43% in previous studies. Obesity was twice asprevalent in women as men and associated with

Table 3 Factors associated with underweight in adults with intellectual disability* (sample limited to 574 adults who were underweight ornormal weight only)

Characteristics Total

Underweight Unadjusted data Adjusted data†

n (%) OR (95% CI) P-value OR (95% CI) P-value

574 208 (36%)Age at health check

20–29 163 84 (52%) 1.00 (reference) <0.001 1.00 (reference) 0.00430–39 134 47 (35%) 0.51 (0.32–0.81) 0.58 (0.35–0.96)40–49 126 39 (31%) 0.42 (0.26–0.69) 0.51 (0.37–0.87)�50 151 38 (25%) 0.32 (0.20–0.51) 0.37 (0.21–0.65)

Accommodation at health checkWith family/alone 207 90 (43%) 1.62 (1.14–2.31) 0.007 1.26 (0.81–1.97) 0.31Residential care 367 118 (32%) 1.00 (reference) 1.00 (reference)

Level of understandingFairly good 227 82 (36%) 1.00 (reference) 0.02 1.00 (reference) 0.16Poor/moderate 269 87 (32%) 0.85 (0.58–1.23) 1.03 (0.67–1.56)Very poor 78 39 (50%) 1.77 (1.05–2.97) 1.88 (0.93–3.82)

Skills deficitsFeeding/drinking 107 48 (45%) 1.56 (1.02–2.39) 0.04 1.19 (0.66–2.12) 0.56

Physical disabilitiesDifficulty walking 112 50 (45%) 1.55 (1.02–2.36) 0.04 1.42 (0.85–2.37) 0.18

ComorbidityDown syndrome 83 21 (25%) 0.55 (0.32–0.93) 0.02 0.46 (0.26–0.81) 0.006Hypertension:

Present 158 42 (27%) 0.54 (0.36–0.81) 0.01 0.59 (0.38–0.93) 0.06Absent 388 156 (40%) 1.00 (reference) 1.00 (reference)Missing data 28 10 (36%) 0.83 (0.37–1.84) 0.78 (0.33–1.88)

Other health indicatorsTaking medication 289 93 (32%) 0.70 (0.50–0.99) 0.04 0.57 (0.38–0.87) 0.008*

* All variables described in Box 1 were entered into the unadjusted analysis; only those that were significant at the 5% level (P < 0.05) areshown here.† We adjusted the multivariable model for age, sex, ethnicity, Down syndrome, smoking status and all the remaining significant factorsshown in the table.n, number; OR, odds ratio; CI, confidence interval.




living independently/with family, ability to feed/drink unaided, being female, hypertension, Downsyndrome and the absence of cerebral palsy. Under-weight was more prevalent in men than women,and associated with younger age, absence of Downsyndrome and not taking medication.

The prevalence rate of 21% that we have reportedfor obesity is lower than that in Emerson’s andYamiki’s studies, which reported prevalences of 27%and 35%, respectively in the 1997–2000 period(Emerson 2005; Yamaki 2005). Both studiesincluded a higher proportion of adults at the milderend of the ID spectrum, particularlyYamaki’s, whichsampled the entire general population for self-reported ID and therefore included a broader rangeof ID as a whole.Thus, the differences betweenstudies could be explained by differences in the rep-resentativeness of the study populations and by agreater risk of obesity in adults with mild ID.

We found that adults with ID who were livingindependently/with family were more likely to beobese than those in residential care, which is consis-tent with earlier, small-scale studies (Prasher 1995;Rimmer et al. 1995; Gravestock 2000). A more struc-tured residential environment may enable closersupervision of food intake and stricter monitoring ofweight (Prasher 1995; Rimmer et al. 1995). People inthe community with limited incomes are less likelyto eat healthily or have opportunities to engage inphysical activities (National Audit Office 2001;Department of Health 2003; Disability Rights Com-mission 2006).Within the ID population, it is diffi-cult to identify and interpret meaningful associationswith socio-economic group and other deprivationfactors. Our clinical experience has shown that somefamily carers have limited knowledge about healthyeating and the risks of obesity, often because of theirown isolation. Many lack the education and practicalsupport that they need to improve the lifestyle of theindividual with ID. Some ‘overprotect’ the individualby restricting involvement in community activities,sometimes with compensatory overfeeding.Whereindividuals have challenging behaviour, it can be dif-ficult to resist their demands for food if this disruptsthe peace.

Our univariate analysis showed an association ofobesity with relatively mild ID, consistent withsome (Rimmer et al. 1995; Hove 2004) but not all(Prasher 1995) previous studies. Underweight was

associated with relatively profound ID, as reportedelsewhere (Rimmer et al. 1995; Department ofHealth 1998; Emerson 2005). Other studies sug-gested that obesity increases with increasing age(Gravestock 2000; Moran et al. 2005), but oursonly confirmed the association of underweight withyounger age. However, individuals in both the IDand general populations may move in and out ofobesity (Moran et al. 2005).

Within the ID population, people with Downsyndrome were more likely to be obese than others.Earlier studies variously found that obesity wasmore prevalent in both sexes with Down syndrome(Prasher 1995; Gravestock 2000; Moran et al. 2005;Rimmer & Wang 2005) or just in women withDown syndrome (Melville et al. 2005). We foundthat obesity was independently associated with bothfemale sex and Down syndrome. The cause ofobesity in Down syndrome is unknown, but isprobably multifactorial. Low-basal metabolic rate,hypotonia and endocrine abnormalities such ashypothyroidism (Prasher 1995; The Foundation forPeople with Learning Disabilities 2001; Melvilleet al. 2005) may compound factors that predisposethe whole ID population to obesity. These includelimited understanding about health risks, restrictedopportunities to gain appropriate knowledge, pooreating behaviour, high calorie intake, reduced exer-cise, fewer opportunities for physical and socialactivities, and low income (Rimmer et al. 1995; TheFoundation for People with Learning Disabilities2001; Hove 2004; Disability Rights Commission2006).

Four per cent of our study population werereported as having a problem with continual eating/drinking, self-induced vomiting or pica. These arepart of the broad spectrum of eating disorderscommon among adults with ID (Gravestock 2000).They are associated with considerable physical,behavioural, psychiatric and social morbidity andneed to be addressed alongside obesity (Gravestock2000). Our study population had additional practi-cal difficulties in feeding/drinking and preparingsimple foods unaided.

Strengths and limitations of the study

Ours was a large, population-based study using along-standing epidemiological register with a high




ascertainment level and standardised schedules forthe home interviews and health checks. The registerprovides a unique source of high-quality data. Ourcomparisons of the ID and general populationswere adjusted for age using SMRs. The adjustedanalyses in the logistic regression modelling enabledus to identify specific factors that were indepen-dently associated with obesity and underweight. Thecross-sectional design precluded us from reportingon any predictive associations.

A comparison of the characteristics of the studypopulation with a cross section of registrants withID over the same period showed similar distribu-tions. However, people with greater levels of under-standing were under-represented in the studypopulation. This is likely to underestimate slightlythe prevalence of obesity and to overestimateslightly the prevalence of underweight when com-pared with the ID population as a whole. Further,the prevalence of obesity in the general populationhas risen substantially since 1998, when our refer-ence population was measured (Rennie & Jebb2005). This increase appears to be at least as highin the adult ID population (Rimmer & Yamaki2006) and may be even higher in the UK, where anumber of residents of NHS long-term care facili-ties have been relocated to community settingswhere they are less closely monitored and are at agreater risk of obesity. Again, this may lead to aslight underestimate in the prevalence of obesitythat we have reported.

Service implications

Body weight status is largely a function of lifestyle,reflecting factors such as diet and physical activity(Gravestock 2000; National Audit Office 2001;Haslam & James 2005; Haslam et al. 2006).Obesity is generally preventable and treatablethrough a range of public health measures andperson-centred clinical interventions (Departmentof Health 2001; National Audit Office 2001;Department of Health 2003; Haslam & James2005; Disability Rights Commission 2006; Haslamet al. 2006; National Institute for Health andClinical Excellence 2006), even in the rarer caseswhere obesity results primarily from a genetic pre-disposition to gaining weight (Gravestock 2000).The rationale for reducing obesity – and under-

weight – is the same in the ID population as inthe general population: to reduce chronic diseaseand disability, to increase quality of life, to reducehealth inequalities and to attenuate future healthcare costs (National Audit Office 2001; Depart-ment of Health 2003). However, the evidence basefor successfully managing obesity through diet,exercise, behaviour therapy, drugs and evensurgery (National Audit Office 2001; Haslam et al.2006; National Institute for Health & ClinicalExcellence 2006) relates to the general populationand needs to be tested in the ID population. Thefew intervention studies that have been done inthe ID population in primary care have mostlybeen small, non-randomized and short-term, butthese suggest that direct practitioner input is ben-eficial and that carers need to be involved(Chapman et al. 2005; Disability Rights Commis-sion 2006).

Methods for managing weight problems need tobe adapted for the ID population and political,organisational and attitudinal barriers challengedat every level of health care (Department ofHealth 2001, 2007; Healthcare Commission 2005;Disability Rights Commission 2006; Mencap2007). People with ID have often been treated asa low priority and excluded from mainstream gov-ernment targets (Disability Rights Commission2006; Mencap 2007). Many healthcare profession-als do not know much about ID or understandthe law around capacity and consent to treatment(Disability Rights Commission 2006; Mencap2007). Few fully consult and involve the familyand carers of people with ID. Diagnostic overshad-owing is common (Disability Rights Commission2006; Mencap 2007). The ID population is lesslikely to receive expected evidence-based checksand treatment than the general population (Dis-ability Rights Commission 2006; Mencap 2007).People with ID and diabetes have fewer BMImeasurements than others with diabetes (DisabilityRights Commission 2006). The Disability RightsCommission (2006) identified some impressiveexamples of positive practice in primary care, butalso extensive evidence of services not making the‘reasonable adjustments’ required by the Disa-bility Discrimination Act. Recommendationsinclude more ID training for all medical staff,annual health checks for people with ID,




improved access to and support within mainstreamservices, and a confidential enquiry into prematuredeaths (Disability Rights Commission 2006;Mencap 2007).

Primary prevention of obesity would preventmuch secondary disease (Haslam et al. 2006).Better screening, documentation and long-termmonitoring of weight status are required for thewhole ID population (Department of Health 2001,2007; Disability Rights Commission 2006; Haslamet al. 2006), with attention to underweight as wellas obesity. Our finding of an association betweenobesity and hypertension suggests potential forimprovement in important health outcomes. Ourstudy suggests that intervention might initially focuson women, those living independently/with familyand those with Down syndrome for obesity, espe-cially those able to feed themselves, and on youngeradults for underweight, especially those not takingmedication.

Recent policy initiatives have highlighted the needto reduce health and social inequalities in peoplewith ID, including supporting them to leadhealthier lifestyles (Department of Health 2001,2007; Healthcare Commission 2005). Valuing Peopleadvocates the role of health facilitators in develop-ing health action plans for all individuals with ID(Department of Health 2001). Promoting Equalitysupports regular, comprehensive health checks,endorsing them as the most effective way ofimproving the health of people with ID (Depart-ment of Health 2007). These processes need tobecome fully integrated within primary care. Thegeneral practitioner contract now provides incen-tives for establishing practice registers of patientswith ID, and access to annual health checks is agovernment priority for the future (HealthcareCommission 2005; Department of Health 2007).Professionals in primary care need training todevelop their skills and confidence in providingholistic health care for people with ID, includingmanaging any coexisting behaviour problems. Spe-cialist ID professionals have an important role toplay in supporting and educating mainstream staffand facilitating service development (Department ofHealth 2001, 2007). Crucially, to achieve all this,resources are required to support a comprehensivelong-term strategy to promote the health of the IDpopulation.

What this paper adds

What is already known on this topic

• Small-scale studies suggest that adults with ID aremore likely to be obese than the general populationand also more likely to be underweight• Individuals with Down syndrome are more likelyto be obese than others with ID.

What this study adds• Our large-scale, population-based study con-firmed these health inequalities using standardisedmorbidity data:– women with ID were 1.5 times more likely to beobese than women in the general population andtwice as likely to be underweight;– men with ID were eight times more likely to beunderweight than men in the general population• In the ID population, obesity was associated withliving independently/with family, ability to feed/drink unaided, being female, hypertension, Downsyndrome and the absence of cerebral palsy• In the ID population, underweight was associatedwith younger age, absence of Down syndrome andnot taking medication.

Acknowledgements

The authors acknowledge the contribution of DrAppiah, who carried out the health checks, and theLeicestershire Learning Disability Register interviewteam.

References

Chapman M. J., Cravel M. J. & Chadwick D. D. (2005)Fighting fit? An evaluation of health practitioner inputto improve healthy living and reduce obesity for adultswith learning disabilities. Journal of Intellectual Disabili-ties 9, 131–44.

Department of Health (1998) Health Survey for England1998. Available at: http://www.archive.official-documents.co.uk/document/doh/survey98/hset3-41.htm(retrieved 20 June 2007).

Department of Health (2001) Valuing People. A New Strat-egy for Learning Disability for the 21st Century. Her Maj-esty’s Stationery Office, London.

Department of Health (2003) Tackling Health Inequalities.A Programme for Action. Her Majesty’s Stationery Office,London.




http://www.archive.official-documents.co.uk/document/doh/survey98/hset3-41.htm



Department of Health (2007) Promoting Equality. Responsefrom the Department of Health to the Disability RightsCommission Report ‘Equal Treatment: Closing the Gap’.Department of Health, London.

Disability Rights Commission (2006) Equal Treatment:Closing the Gap. A Formal Investigation into PhysicalHealth Inequalities Experienced by People with LearningDisabilities and/or Mental Health Problems. DisabilityRights Commission, London.

Emerson E. (2005) Underweight, obesity and exerciseamong adults with intellectual disabilities in supportedaccommodation in Northern England. Journal of Intel-lectual Disability Research 49, 134–43.

Gravestock S. (2000) Eating disorders in adults with intel-lectual disability. Journal of Intellectual Disability Research44, 625–37.

Haslam D., Sattar N., Lean M. (2006) ABC of obesity.Obesity – time to wake up. British Medical Journal 333,640–2.

Haslam D. W. & James W. P. (2005) Obesity. Lancet 366,1197–209.

Healthcare Commission (2005) Draft Three-Year StrategicPlan for Assessing and Encouraging Improvement in theHealth and Healthcare of Adults with Learning Disabilities2006–2009. Commission for Healthcare Audit andInspection, London.

Heber R. (1959) A manual on terminology and classifica-tion in mental retardation. American Journal of MentalDeficiency 64, 1–11.

Holmes N., Shah A. & Wing L. (1982) The disabilityassessment schedule: a brief screening device for use withthe mentally retarded. Psychological Medicine 12, 879–90.

Hove O. (2004) Weight survey on adults persons withmental retardation living in the community. Research inDevelopmental Disabilities 25, 9–17.

Melville C. A., Cooper S.-A., McGrother C. W., Thorp C.F. & Collacott R. (2005) Obesity in adults with Downsyndrome: a case-control study. Journal of IntellectualDisability Research 49, 125–33.

Mencap (2007) Death by Indifference. Mencap, London.

Moran R., Drane W., McDermott S., Dasari S., ScurryJ. B. & Platt T. (2005) Obesity among people with and

without mental retardation across adulthood. ObesityResearch 13, 342–9.

National Audit Office (2001) Tackling Obesity in England.Report by the Comptroller and Auditor General. Her Maj-esty’s Stationary Office, London.

National Institute for Health and Clinical Excellence(2006) Obesity: Full Guideline. Available at: http://www.nice.org.uk/guidance/CG43 (retrieved 20 June2007).

Prasher V. P. (1995) Overweight and obesity amongstDown’s syndrome adults. Journal of Intellectual DisabilityResearch 39, 437–41.

Rennie K. L. & Jebb S. A. (2005) Prevalence of obesity inGreat Britain. Obesity Reviews 6, 11–2.

Rimmer J. H. & Wang E. (2005) Obesity prevalence amonga group of Chicago residents with disabilities. Archives ofPhysical Medicine and Rehabilitation 86, 1461–4.

Rimmer J. H. & Yamaki K. (2006) Obesity and intellec-tual disability. Mental Retardation and Developmental Dis-abilities Research Reviews 12, 22–7.

Rimmer J. H., Braddock D. & Marks B. (1995) Healthcharacteristics and behaviors of adults with mentalretardation residing in three living arrangements.Research in Developmental Disabilities 16, 489–99.

The Foundation for People with Learning Disabilities(2001) Learning Disabilities:The Fundamental Facts. TheMental Health Foundation, London.

Watson J. M. (2003) Valuing people valuing resources. TheLeicestershire learning disabilities register. Frontline 53,24–5.

World Health Organisation (1992) International StatisticalClassification of Diseases and Related Health Problems,10th revision. World Health Organisation, Geneva.

World Health Organisation Expert Committee (1995)Physical Status:The Use and Interpretation of Anthropom-etry. WHP Technical Report Series 854. World HealthOrganisation, Geneva.

Yamaki K. (2005) Body weight status among adults withintellectual disability in the community. Mental Retarda-tion 43, 1–10.

Accepted 17 October 2007




http://www.nice.org.uk/guidance/CG43

http://www.nice.org.uk/guidance/CG43

Documents

Body mass index in adults with intellectual disability: distribution, associations and service implications: a population-based prevalence study