Research in Developmental Disabilities 34 (2013) 1958–1965

Contents lists available at SciVerse ScienceDirect

Research in Developmental Disabilities

The Vienna Frailty Questionnaire for Persons with Intellectual

Disabilities—Revised

Barbara Brehmer-Rinderer 1,*, Elisabeth Lucia Zeilinger 1, Ana Radaljevic 1,Germain Weber 1

Faculty of Psychology, University of Vienna, Austria

A R T I C L E I N F O

Article history:

Received 19 December 2012

Received in revised form 4 March 2013

Accepted 5 March 2013

Available online 10 April 2013

Keywords:

Frailty

Intellectual disabilities

Ageing

Vienna Frailty Questionnaire-Intellectual

Disabilities – Revised (VFQ-ID-R)

A B S T R A C T

Frailty is a theoretical concept used to track individual age-related declines. Persons with

intellectual disabilities (ID) often present with pre-existing deficits that would be

considered frailty markers in the general population. The previously developed Vienna

Frailty Questionnaire for Persons with ID (VFQ-ID) was aimed at assessing frailty in this

specific population. However, findings of the initial frailty study in 2007 revealed some

weaknesses. This paper reports on the development of the Vienna Frailty Questionnaire for

Persons with ID – Revised (VFQ-ID-R) as well as its first application and psychometric

evaluation. The authors re-administered the VFQ-ID-R to participants with ID who had

been assessed with the VFQ-ID in 2007. The goal was to study the factor structure and

reliability of the revised test. Internal consistency of the VFQ-ID-R was found to be very

good for the entire scale and was moderate for the four domains of the scale. Inter-rater

reliability and retest reliability were found to be good. The revisions made to the VFQ-ID

will be useful in assessing and supporting ageing individuals with ID.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Frailty is an evolving concept in gerontology research. It is especially useful to track age-related declines on an individuallevel and to predict adverse health outcomes (Lacas & Rockwood, 2012). Different theoretical approaches to frailty exist.There is neither an official definition nor an agreed-upon way of measuring it (Hogan, MacKnight, & Bergman, 2003; Lacas &Rockwood, 2012; Rockwood, 2005). Most theories of frailty focus primarily on physical deficits (Ferrucci et al., 2004; Friedet al., 2001; Romero-Ortuno, Walsh, Lawlor, & Kenny, 2010), while others include broader aspects like mood or social contact(e.g. Ravaglia et al., 2008). One well-known theory is the Frailty Index (FI, Mitnitski, Mogilner, & Rockwood, 2001), whichassesses frailty by mere accumulation of different deficits. Although methods for creating an FI were proposed (Searle,Mitnitski, Gahbauer, Gill, & Rockwood, 2008), the deficits chosen for the computation of an FI are not relevant for predictingadverse health outcomes (Mitnitski, Song, & Rockwood, 2004). They can even be chosen randomly (Rockwood, Andrew, &Mitnitski, 2007).

Persons with ID often present pre-existing deficits, which are in turn considered frailty markers in the general population(e.g., deficits in the Activities of Daily Living, ADL). Thus, the premorbid level of frailty-related deficits in persons with IDdiffers compared to the general population. This can bias the measurement of frailty, e.g., the computation of an FI in persons

* Corresponding author at: University of Vienna, Faculty of Psychology, Liebiggasse 5, A-1010 Vienna, Austria. Tel.: +43 1427722347.

E-mail address: barbara.brehmer@univie.ac.at (B. Brehmer-Rinderer).1 All authors contributed equally to this article.

0891-4222/$ – see front matter � 2013 Elsevier Ltd. All rights reserved.

http://dx.doi.org/10.1016/j.ridd.2013.03.004

B. Brehmer-Rinderer et al. / Research in Developmental Disabilities 34 (2013) 1958–1965 1959

with ID. Using methods that were designed for the general population in persons with ID can lead to flawed results. Thisargumentation is comparable to the discussion on diagnosing dementia in persons with ID. There it is widely accepted thatinstruments for the general population are not suitable for persons with ID due to their differing premorbid level (e.g.Nieuwenhuis-Mark, 2009; Prasher, Farooq, & Holder, 2004). Just like diagnostic tools for dementia, frailty instruments needto be specially designed in order to gain accurate results.

The issue of adapting existing frailty measurement tools has already been addressed (Brehmer & Weber, 2010) andconsequently an instrument to assess frailty in the ID population was designed; the Vienna Frailty Questionnaire for Persons

with ID (VFQ-ID). This questionnaire focuses on changes in specific aspects and not on the mere existence of deficits, whichmakes it applicable for all persons with or without ID, irrespective of their premorbid level of frailty-related deficits.

The VFQ-ID was first used 2007 on 190 persons with ID and its psychometric properties were examined (Brehmer &Weber, 2010). Some weaknesses related to its content and to the method of conduction were revealed in the process. The aimof the present study was to redefine and improve the VFQ-ID by describing and reflecting on these encountered weaknessesand to introduce the VFQ-ID-Revised (VFQ-ID-R). This paper reports on the first application of the VFQ-ID-R and itspsychometric evaluation. We report on structural validity and reliability measures of the new instrument. Furthermore, thisstudy will give an example to demonstrate the use of the VFQ-ID-R in practice as a useful instrument to track age-relateddecline in persons with ID.

2. Material and methods

2.1. Participants

All persons (n = 190), who had participated in the first frailty study (Brehmer & Weber, 2010), were contacted and invitedto the follow up study. Out of these, 43 persons denied participation due to the following reasons: (a) 24 persons did not wantto be interviewed again, (b) ten persons could not be contacted for various reasons (e.g. moved to another province ofAustria), (c) four persons could not participate due to their deteriorated health status, (d) the caregivers of three respondentswith ID decided not to take part without giving any reasons, (e) one person died, and (f) one interview set was not returnedby the informant. The final sample of this follow-up consisted of 147 persons with ID, including 69 women (46.9%; seeTable 1).

The age range was 20–72 years (M = 43). Regarding living arrangements, 61.9% (n = 91) lived in an institution or grouphome and 38.1% (n = 56) either lived alone or with their family. Nearly two thirds (64.6%; n = 95) of the sample was living inan urbanised area and 35.4% (n = 52) in a rural area. With respect to support required, 108 persons with ID (73.5%) were ableto speak for themselves during the interview. Only 39 caregivers (26.5%) had to speak for their clients, who could notcommunicate verbally. Following the recommendation of the ethics committee for the first frailty study, once again we didnot objectively measure the level of ID to keep the respondents load low and the conduction rather simple, but asked forrecords on previous ID testing. This information on ID diagnosis in addition to the main carer’s appraisal of the intensity ofsupport needed by the participant with ID in the ADLs was used to estimate the participant’s ID level. The coding of mild andmoderate ID equals low intensity of support in the ADLs, whereas severe and profound ID equals high intensity of supportand a respective diagnosis (see Table 1). The sample parameters (age, living conditions, level of ID and sex) are similarlydistributed compared to the original study by Brehmer and Weber (2010).

To assess inter-rater reliability, 17 persons with ID (nine men and eight women) were voluntarily questioned a secondtime by another interviewer. The mean age of this group was 48 years. 70.6% (n = 12) lived in an institution or group homeand 29.4% (n = 5) either lived alone or with their family. The majority (82.4%; n = 14) were living in an urbanised area and onlythree persons (17.6%) in a rural area. All persons were able to speak for themselves.

To explore retest reliability of the final VFQ-ID-R, another 17 persons, including ten women, were questioned again fourweeks after the first interview. The mean age of this sample was 41 years. A total of 76.5% (n = 13) resided in an institution orgroup home and 23.6% (n = 4) either lived alone or with their family. 82.4% (n = 14) were living in an urbanised area and 17.6%

Table 1

Sample characteristics.

Age Mild/moderate ID (n = 114) Severe/profound ID (n = 33)

Urban (n = 72) Rural (n = 42) Urban (n = 23) Rural (n = 10)

IS (n = 44) Non IS (n = 28) IS (n = 23) Non IS (n = 19) IS (n = 21) Non IS (n = 2) IS (n = 3) Non IS (n = 7)

20–29 7 8 2 2 4 2 – 5

30–39 12 6 1 3 5 – 1 1

40–49 13 11 6 7 6 – 1 1

50–59 5 – 8 4 5 – – –

60+ 7 3 6 3 1 – 1 –

Note. Mild/moderate ID equals low intensity of support. Severe/profound ID equals high intensity of support. IS illustrating figures for persons living in

institutional settings (e.g. group homes) and Non IS illustrating figures for persons in non-institutionalised settings (e.g. living with family members o

independently).

r

B. Brehmer-Rinderer et al. / Research in Developmental Disabilities 34 (2013) 1958–19651960

(n = 3) in a rural area. Ten persons with ID (58.8%) were able to speak for themselves and seven (41.2%) caregivers had tospeak for their nonverbal clients.

2.2. Procedure

Since the Ethics Committee had agreed to the first frailty study, we adhered to these procedures for the follow-up study.Procedures for sampling and data collection are described in detail elsewhere (Brehmer & Weber, 2010). Informed consentwas gained from the participating service providers, carers and persons with ID stepwise. First, respondents of the 2007study were informed with an easy-to-read informational letter explaining the background, aim, participant’s rights (i.e., theycould withdraw at any time of the interview, have a break during the interviews, refuse to answer, and be supported by apersonal assistant), and content of this project. Second, face-to- face meetings were set up in order to organise alladministrative details in person (e.g., time and place for the interviews). Similar to the first study, the consent procedureended with an agreement on the date and time for the formal study interview.

With respect to the re-administration of the VFQ-ID, the training of interviewers was extended to provide specificinstructions on how to gain information on health changes in the past years from persons with ID. All interviewers, masterstudents of clinical psychology, practiced the use of the VFQ-ID-R in various role-plays. They also observed at least seveninterviews conducted by one of the authors (BBR or ELZ), who are trained clinical and health psychologists, experienced inworking with persons with ID and had been interviewers in the original study. The average duration of one interview washalf an hour. Data was collected from April until August 2009.

2.3. Materials

Based on our findings from the first frailty study (Brehmer & Weber, 2010), the adaptation of the existing questionnairefollowed two principles: First, verbally communicating persons with ID should be able to respond to all questions withoutassistance. Therefore the wording and the content of the VFQ-ID were modified. Second, all items should be strongly relatedto frailty or age-related declines according to current research findings. Therefore literature was searched for recentpublications on frailty symptoms and instruments. All four frailty domains (social, physical, psychological and cognitive) of theVFQ-ID were analysed and modified as follows:

2.3.1. Social domain

In the social section, the item club membership was found to be practically irrelevant as no participant had been a memberof any kind of social club. Besides, this item had not been mentioned to be of relevance by any other frailty study.Additionally, the adjusted item-scale correlation was very low, justifying the exclusion of this item from the scale. Instead,items on the quality and quantity of spare time activities and partnership were added to give a thorough picture of possiblechanges in the social life of a person (see Table 2).

2.3.2. Cognitive domain

The cognitive section of the VFQ-ID consisted of six Instrumental Activities of Daily Living (IADL). However, the analysis ofthe questionnaire revealed that changes in the IADLs could have been caused by either cognitive or physical declines. Somepersons only mentioned physical impairments, but stayed cognitive active, whilst others experienced only cognitivedeteriorations. Depending on these developments, different preventive strategies to delay further declines are needed.Therefore, to differentiate between these two possible trends, we added a follow-up question asking for the reason (cognitiveor physical) of the experienced change. To give an example, participants were asked if they had been able to use thetelephone. If not, they were asked ‘Since when?’ and ‘For what reasons?’.

2.3.3. Psychological domain

The psychological section contained a list of psychological/emotional symptoms, missing an item on the mood stateanger. Thus, this symptom was frequently mentioned in the first study and therefore was included as it proved to be ofrelevance to describe age-related changes.

2.3.4. Physical domain

The physical section was expanded with three items (sleeping, balance, appetite), which were not included in the VFQ-ID,but were of interest in many other studies examining healthy ageing (e.g. Cleaver, Hunter, & Ouellette-Kuntz, 2009;Nikolaus, 2001). Additionally, the item physical strength was split up into two variables (physical strength in the arms andphysical strength in the legs) to record the individual change concerning mobility more precisely. Thereby, it can be specifiedwhether physical deteriorations experienced by a person concentrate on the upper limbs (e.g., losing grip strength, strengthin the arms) or the lower limbs (e.g., losing balance, strength in the legs).

2.3.5. Further adaptations

Despite these modifications relating to the content, the newly adapted questionnaire had to be restructured. The previous33 items of the VFQ-ID had been thematically integrated into the POMONA II protocol (Perry et al., 2010) to allow an

Table 2

Comparison of all tested items (study 1 vs. follow-up).

Frailty Domain Study 1 (2006–2008) Study 2 (2009)

Social Negative changes in work Work – change in quantity

Participation in a club/association Work – change in quality

Negative changes in contact with friends Spare time activities – change in quantity

Negative changes in contact with family Spare time activities – change in quality

Friends – change in quantity

Friends – change in quality

Family – change in quantity

Family – change in quality

Sexual relationship – change in quantity

Sexual relationship – change in quality

Physiological General health decline General health decline

ADL: washing ADL: washing

ADL: dressing ADL: dressing

ADL: eating ADL: eating

ADL: using the toilet ADL: using the toilet

ADL: transfers ADL: transfers

Decrease of vision Decrease of vision

Decrease of hearing Decrease of hearing

Decrease of mobility Change of sleeping habitsUnintended weight loss Decrease of mobility

Decrease of appetit Decrease of balanceIncrease of medication Decrease of physical strength in the arms

decrease of physical strength in the legsLoss of grip strength Unintended weight loss

Falls Decrease of appetitIncontinence Increase of medication

Use of walking frame Loss of grip strength

Falls

Incontinence

Use of walking frame

Cognitive (IADL) Using telephone Using telephone

Preparing meals Preparing meals

Handling money Handling money

Shopping Shopping

Easy household chores Easy household chores

Complex household chores Complex household chores

Medication use Medication use

Psychological Fear of falling Fear of falling

Decrease of memory skills Decrease of memory skills

Increase of exhaustion Increase of exhaustion

Increase of nervousness and anxiety Increase of nervousness and anxiety

Increase of sadness Increase of sadness

Other psychological changes Other psychological changes

Increase of anger

Total Item 33 items 44 items

Note: Newly added items are highlighted in boldface.

B. Brehmer-Rinderer et al. / Research in Developmental Disabilities 34 (2013) 1958–1965 1961

economic data collection procedure with a substantial reduction of respondent’s load. In the original study, the participantswith ID answered some items themselves, while for non-verbal communicating persons with ID the respondent was themain caregiver. The analysis of the psychometric properties of the VFQ-ID revealed that items answered by persons with IDshowed vastly different adjusted item-scale correlations compared to those items answered by caregivers. Therefore, wewanted to have only one respondent for the whole questionnaire: the person with ID. We reformulated and restructured theexisting questionnaire accordingly. Mainly the physical domain was rephrased. Finally, the adapted questionnaire was givento three self-advocates, seeking their comments and improvements with respect to its content and its style of language. Theirproposed changes related to the wording of the items were incorporated into the final VFQ-ID. Table 2 illustrates the items ofthe first frailty study as well as the 44 items that were tested in this follow-up study.

The final set of 34 items of the VFQ-ID-R depicted in Table 3 was defined by two principles: (a) being an important item offrailty according to the literature, and (b) good psychometric properties. Following these principles, eleven items had to beexcluded from further analysis (decrease of memory skills, the ADLs eating and transfers, the IADL medication use, decrease of

hearing, unintended weight loss, decrease of appetite, use of walking frame, increase of medication, incontinence, and change of

sleeping habits). As the VFQ-ID-R contains only one additional item, it was not necessary to adapt the two-folded frailtycriterion by Brehmer and Weber (2010).

Table 3

Rational for final set of frailty variables, the new VFQ-ID-R and its psychometric properties.

Frailty domain Frailty variables Central according

to literature

Corrected item-total

correlations

AC1 Spearman

correlation

Social Retirement from work – .05 1.00 1.00*

Reduction in working hours – .36 .93 .23Less spare time activities – .28 .61 .19Less practice of spare time activities .17 1.00 .69

Less friends – .25 .91 .38Less contact to friends – .14 .94 1.00

Change in quality of friendships – .33 1.00 1.00

No family anymore – .05 1.00 .54

Less contact to family members – .17 .80 .31No partnership/sexual relationship anymore – .20 .93 1.00

Less contact to partner – .08 1.00 1.00

Physiological General health decline – .16 .92 1.00

ADL: washing U .31 1.00 1.00

ADL: dressing U .32 1.00 1.00

ADL: using the toilet U .16 1.00 1.00

Decrease of vision – .26 .93 .79

Decrease of mobility U .47 .90 .79

Decrease of balance U .57 1.00 .23Decrease of physical strength in the arms U .42 1.00 �.91Decrease of physical strength in the legs U .47 .94 .79

Loss of grip strength U .41 1.00 1.00

Falls U .32 .57 .29IADL: using telephone U .22 1.00 1.00

Cognitive (IADL) Preparing meals U .15 1.00 1.00

Handling money U .34 1.00 1.00

Shopping U .37 1.00 �.63Easy household chores U .22 1.00 1.00

Complex household chores U .24 1.00 1.00

Psychological Fear of falling U .30 .90 .35Increase of exhaustion U .18 .63 .57

Increase of nervousness and anxiety U .20 .74 .79

Increase of sadness – .19 .76 .23Other psychological changes – .32 1.00 .49

Increase of anger – .05 .92 .69

Total 34 Items

Note: Spearman correlations for items with a variance of zero are set to 1.00. Critical values are in boldface.

B. Brehmer-Rinderer et al. / Research in Developmental Disabilities 34 (2013) 1958–19651962

In general, the VFQ-ID-R contains this list of items divided in four domains (physical, psychological, cognitive, social). In thefirst step, for each item the current status (e.g. good/bad balance) is recorded as perceived by the respondents. Second, theparticipant is asked if he/she experienced changes in this status (better, worse, no change) in the past years therebyanswering the first frailty criterion. Third, the respondent is asked to specify when it (e.g. balance) started changing. Inaddition to the VFQ-ID-R, the PAS-ADD screening (Moss, 2002) and the Aberrant Behaviour Checklist (ABC; Aman & Singh,1994) were administered to the main carers of the ID respondents.

2.4. Data analysis

Principal component analysis with orthogonal rotation (varimax) was conducted to explore structural validity anddetermine the number of factors for the VFQ-ID-R. Internal consistency of the VFQ-ID-R was explored in two steps usingCronbach’s alpha: first for the whole instrument and in the second step for the four frailty domains separately. Inter-raterreliability was calculated using the AC1 coefficient (Gwet, 2008) instead of the more frequently used coefficient Cohen’s Kappa.While the Kappa statistic tends to underestimate the ‘true’ inter-rater reliability for small samples, the bias of the AC1 statistic

is smaller and has proven to be most valuable in data with a variance of zero (Zeilinger, Weber, & Haveman, 2011). The AC1

values can range from 0 to 1 with the latter indicating total agreement. The AC1 coefficient was calculated for every variable ofthe VFQ-ID-R. To determine consistency among raters for the overall VFQ-ID-R scale, the median of the AC1 coefficients of thevariables was built. Spearman’s rho correlation was used for calculating the retest reliability of the VFQ-ID-R. Correlationswere built for every variable and the median was calculated for the retest reliability of the overall scale.

3. Results

This section illustrates the factor structure and reliability measures of the VFQ-ID-R. Additionally, the prevalence of frailtyis presented. Finally, the practical use of this instrument is shown by depicting a case study.

B. Brehmer-Rinderer et al. / Research in Developmental Disabilities 34 (2013) 1958–1965 1963

3.1. Principal component analysis

A principal component analysis was conducted on the 34 frailty variables of the VFQ-ID-R. Twelve components showedeigenvalues >1 and explained 70.25% of total variance. In the second step, 19 variables with item-scale correlations below0.30 were excluded and the principal component analysis was calculated again on the remaining fifteen items. Four factorswith eigenvalues >1 explained 64.33% of total variance. Factor 1 represented the physiological domain, including decrease of

strength in legs, decrease of strength in arms, decrease of mobility, and decrease of balance and loss of grip strength. Factor 2contained two ADLs (washing and dressing) and two social variables (less spare-time activities and reduction in working hours).Factor 3 mainly represented the cognitive domain including two IADL variables, shopping and handling money, one social

variable (change in quality of friendship) and one psychological variable other psychological changes. The fourth factorhighlighted two main frailty symptoms: falls and fear of falling.

3.2. Reliability measures

Cronbach’s alpha for the overall scale was 0.74, which indicates a good reliability. Eighteen variables resulted in a lowadjusted item-scale correlation with values below 0.30 (see Table 3), but as none of them would substantially affectreliability if deleted, they were not excluded from this scale. Cronbach’s alpha for the four subscales of the VFQ-ID-R was 0.75for the physiological, 0.57 for the psychological, 0.47 for the social and 0.41 for the cognitive domain.

Table 3 illustrates the AC1 coefficients for the 34 variables. The median of the 34 AC1 coefficients was 1.00, which indicates aperfect inter-rater reliability for the overall scale of the VFQ-ID-R. Additionally, Table 3 depicts the Spearman rank-order

correlations computed on the basis of a subsample with n = 17. The median value of 0.79 indicates a good retest-reliability forthe overall scale of the VFQ-ID-R.

3.3. Prevalence of frailty

The presence of frailty was coded using a twofold frailty criterion: (1) negative health outcomes in three or four domains

based on Rockwood’s multisystemic frailty concept (Rockwood, Hogan, & MacKnight, 2000); and (2) a minimum of six

negative health outcomes in accordance with Steverink, Slaets, Schuurmans, and van Lis (2001) multisystemic frailty criteria.Persons fulfilling only one of the two frailty criteria were characterised as prefrail.

Considering frailty, 64.6% (n = 95) were characterised as non-frail, 17.7% (n = 26) as prefrail and 17.7% (n = 26) as frail.Compared to the first frailty study, both the percentage of frail (9%; n = 17) and prefrail (12%; n = 22) persons with ID waslower. Therefore, the prevalence of frailty had increased in the sample within two years between these two projects. For thegeneral population aged 65 and older, it is estimated, that 10–25% are frail (Ostir, Ottenbacher, & Markides, 2004). Similar tothe first frailty study, a Spearman’s rho correlation between frailty and age was rather small (rs = 0.29), but statisticallysignificant (p < 0.001). We found no sex or mental health related differences with respect to frailty.

3.4. Case study

The case study depicted here should give an example on the kind of information that can be gathered with the VFQ-ID-R.This information is valuable for tracking age-related changes, for personal planning, as well as for detecting changes in thekind and intensity of support needs of an ageing person with ID.

In 2007, a 60-year-old independently living man with mild ID and his caregiver were interviewed using the VFQ-ID. Sincethey both reported five negative changes in three out of the four frailty domains, the man was coded as prefrail. Two yearslater, he answered the revised interview protocol alone and mentioned fourteen negative changes in all four frailty domains:In the past years, he had experienced a decline in mobility, balance and reduced strength in his arms and legs. His gripstrength had declined as well. Shopping and handling money had become more difficult. He had chosen more passive sparetime activities and opted for reduced working hours due to his physical state. He said he had become increasingly sad andangry, but also more anxious in the past two years. Additionally, he reported to have fewer contacts to friends and familymembers.

4. Discussion

The aim of this study was to redefine and improve the VFQ-ID and to introduce the VFQ-ID-R. In accordance with previousresearch findings, the four frailty domains were restructured and reformulated with the assistance of self-advocates. Wedescribed the adaptation and changes of the questionnaire and its content. The final VFQ-ID-R consists of 34 items andappears to be suitable for detecting and describing age-related changes experienced by persons with ID. With respect to thepsychometric properties of the VFQ-ID-R, internal consistency was very good for the overall scale. The factor structuresupports the four frailty domains suggested, although internal consistency was quite low on subscale level for thepsychological, social and cognitive domain. As expected, the more homogenously designed physiological domain wasindicated as the most important factor of this frailty measure. Besides, falls and fear of falling are central when measuringfrailty. This result is in line with those reported in previous publications (e.g. Fried et al., 2001).

B. Brehmer-Rinderer et al. / Research in Developmental Disabilities 34 (2013) 1958–19651964

Inter-rater and retest reliability results were good, although sample sizes, with only seventeen participants, were small.Since most of these respondents did not show age-related declines, future research needs to discover whether thesepsychometric results are different for frail or prefrail persons. However, these findings support the conduction of an extendedinterviewer training and the operationalisation of the frailty concept. Further research should examine additionalpsychometric properties and measurement characteristics that are known to be of importance in instruments assessinghealth status in persons with ID (Zeilinger et al., in press).

We examined prevalence in this sample of 147 persons with ID and found an increase of frailty from 9% to 17.7% andan increase of prefrail persons from 12% to 17.7% since the first study in 2007. Compared to the general population,results indicate that persons with ID do not show age-related declines more frequently. Previous findings on therelations between frailty, age and sex were observed again. Future research needs to determine precisely whether theresults are different according to ID aetiology and ID level. Besides, the VFQ-ID-R needs to be tested among differentethnic groups as service structures included in this project mainly support white Austrian citizens. No differences withrespect to behavioural or mental disorders were found, possibly because the sample of frail persons was so small.Clearly, further research on a large, representative population is needed to define the impact of comorbid disorders onfrail ageing persons with ID.

In conclusion, this VFQ-ID-R proved to be useful in practice to follow and depict age-related declines in persons with ID. Itthereby helps to identify actual support needs and can indicate future trends. This may help, for example, healthprofessionals in planning the care for persons with ID who recently moved into a new place of accommodation. It may alsohelp clinicians to systematically and thoroughly follow up on the client’s abilities. The VFQ-ID-R assesses different aspects ofa person, including physical, cognitive and psychological elements as well as his/her social life. Finally, the personal view ofthe person with ID can be captured and also be compared to the perception of caregivers. Currently, the VFQ-ID-R can only beadministered to verbally communicating persons with ID. Non-verbally communicating clients need to be assisted by theircarers. Information gathered with this instrument is not biased by the existence of pre-morbid deficits and therefore is areliable measure of frailty and age-related changes in persons with ID.

Acknowledgements

The authors would like to acknowledge the contributions and commitment to this study by Claudia Oppenauer, ChristineStimakovits, Denise Gruseck, Markus Ebertowsky, Nora Hickel, Viktoria Kantor, Philipp Weber and Lucija Zigrovic.Additionally, we want to thank all the participants and their caregivers, who supported them before, during and after theinterview.

For additional information on the previous European project POMONA II, see http://www.pomonaproject.org.

References

Aman, M. G., & Singh, N. N. (1994). Aberrant behavior checklist community, Supplementary manual. New York: Slosson Educational Publications.Brehmer, B., & Weber, G. (2010). Frailty vs. disability distinctions in people with intellectual disabilities. Journal of Policy and Practice in Intellectual Disabilities, 7,

49–58.Cleaver, S., Hunter, D., & Ouellette-Kuntz, H. (2009). Physical mobility limitations in adults with intellectual disabilities: A systematic review. Journal of Intellectual

Disability Research, 53, 93–105.Ferrucci, L., Guralnik, J. M., Studenski, S., Fried, L. P., Cutler, G. B., Walston, J. D., et al. (2004). Designing randomized, controlled trials aimed at preventing or

delaying functional decline and disability in frail, older persons: A consensus report. Journal of the American Geriatrics Society, 52, 625–634.Fried, L. P., Tangen, C. M., Walston, J., Newman, A. B., Hirsch, C., Gottdiener, J., et al. (2001). Frailty in older adults: Evidence for a phenotype. Journals of Gerontology

Series A: Biological Sciences and Medical Sciences, 56, M146–M156.Gwet, K. L. (2008). Computing inter-rater reliability and its variance in the presence of high agreement. British Journal of Mathematical & Statistical Psychology, 61,

29–48.Hogan, D. B., MacKnight, C., & Bergman, H. (2003). Models, definitions, and criteria of frailty. Aginging Clinical Experimental Research, 15, 1–29.Lacas, A., & Rockwood, K. (2012). Frailty in primary care: A review of its conceptualization and implications for practice. BMC Medicine, 10, 4.Mitnitski, A. B., Mogilner, A. J., & Rockwood, K. (2001). Accumulation of deficits as a proxy measure of aging. The Scientific World Journal, 1, 323–336.Mitnitski, A. B., Song, X. W., & Rockwood, K. (2004). The estimation of relative fitness and frailty in community-dwelling older adults using self-report data.

Journals of Gerontology Series. A, Biological Sciences and Medical Sciences, 59, 627–632.Moss, S. (2002). The PAS-ADD checklist (revised). Brighton, UK: Pavilion Publishing.Nieuwenhuis-Mark, R. E. (2009). Diagnosing Alzheimer’s dementia in Down syndrome: Problems and possible solutions. Research in Developmental Disabilities, 30,

827–838.Nikolaus, T. (2001). Comprehensive geriatric assessment. Zeitschrift fuer Gerontologie und Geriatrie, 34, 36–42.Ostir, G. V., Ottenbacher, K. J., & Markides, K. S. (2004). Onset of frailty in older adults and the protective role of positive affect. Psychology and Aging, 19,

402–408.Perry, J., Linehan, C., Kerr, M., Salvador-Carulla, L., Zeilinger, E., Weber, G., et al. (2010). The P15-a multinational assessment battery for collecting data on health

indicators relevant to adults with intellectual disabilities. Journal of Intellectual Disability Research, 54, 981–991.Prasher, V., Farooq, A., & Holder, R. (2004). The Adaptive Behaviour Dementia Questionnaire (ABDQ): Screening questionnaire for dementia in Alzheimer’s disease

in adults with Down syndrome. Research in Developmental Disabilities, 25, 385–397.Ravaglia, G., Forti, P., Lucicesare, A., Pisacane, N., Rietti, E., & Patterson, C. (2008). Development of an easy prognostic score for frailty outcomes in the aged. Age and

Ageing, 37, 161–166.Rockwood, K. (2005). Frailty and its definition: A worthy challenge. Journal of the American Geriatrics Society, 53, 1069–1070.Rockwood, K., Andrew, M., & Mitnitski, A. (2007). A comparison of two approaches to measuring frailty in elderly people. Journals of Gerontology. Series A, Biological

Sciences and Medical Sciences, 62, 738–743.Rockwood, K., Hogan, D. B., & MacKnight, C. (2000). Conceptualisation and measurement of frailty in elderly people. Drugs and Aging, 17, 295–302.Romero-Ortuno, R., Walsh, C. D., Lawlor, B. A., & Kenny, R. A. (2010). A frailty instrument for primary care: Findings from the Survey of Health. Ageing and

Retirement in Europe (SHARE). BMC Geriatrics, 10, 57.Searle, S. D., Mitnitski, A., Gahbauer, E. A., Gill, T. M., & Rockwood, K. (2008). A standard procedure for creating a frailty index. BMC Geriatrics, 8, 24.

B. Brehmer-Rinderer et al. / Research in Developmental Disabilities 34 (2013) 1958–1965 1965

Steverink, N., Slaets, J. P. J., Schuurmans, H., & van Lis, M. (2001). Measuring frailty: Developing and testing the GFI (Groningen frailty indicator). Gerontologist, 41,236–237.

Zeilinger, E. L., Nader, I. W., Brehmer-Rinderer, B., Koller, I., & Weber, G. (in press). CAPs-IDD: Characteristics of assessment instruments for psychiatric disorders inpersons with intellectual developmental disorders. Journal of Intellectual Disability Research.

Zeilinger, E. L., Weber, G., & Haveman, M. J. (2011). Psychometric properties and norms of the German ABC-Community and PAS-ADD Checklist. Research inDevelopmental Disabilities, 32, 2431–2440.