Back to Journals » Clinical Interventions in Aging » Volume 15

The Prediction of Mortality by Disability Among Dutch Community-Dwelling Older People

Authors Gobbens RJJ, van der Ploeg T

Received 11 July 2020

Accepted for publication 9 August 2020

Published 5 October 2020 Volume 2020:15 Pages 1897—1906

DOI https://doi.org/10.2147/CIA.S271800

Checked for plagiarism Yes

Review by Single anonymous peer review

Peer reviewer comments 2

Editor who approved publication: Dr Richard Walker

Download Article [PDF] 

Robbert JJ Gobbens,1– 3 Tjeerd van der Ploeg4

1Faculty of Health, Sports and Social Work, Inholland University of Applied Sciences, Amsterdam, the Netherlands; 2Zonnehuisgroep Amstelland, Amstelveen, the Netherlands; 3Department of Primary and Interdisciplinary Care, Faculty of Medicine and Health Sciences, University of Antwerp, Antwerp, Belgium; 4Faculty of Engineering, Design and Computer Technology, Inholland University of Applied Sciences, Alkmaar, the Netherlands

Correspondence: Robbert JJ Gobbens
Faculty of Health, Sports and Social Work, Inholland University of Applied Sciences, De Boelelaan 1109, Amsterdam 1081 HV, the Netherlands
Tel +31 6 21115578
Email [email protected]

Objective: To predict mortality by disability in a sample of 479 Dutch community-dwelling people aged 75 years or older.
Methods: A longitudinal study was carried out using a follow-up of seven years. The Groningen Activity Restriction Scale (GARS), a self-reported questionnaire with good psychometric properties, was used for data collection about total disability, disability in activities in daily living (ADL) and disability in instrumental activities in daily living (IADL). The mortality dates were provided by the municipality of Roosendaal (a city in the Netherlands). For analyses of survival, we used Kaplan–Meier analyses and Cox regression analyses to calculate hazard ratios (HR) with 95% confidence intervals (CI).
Results: All three disability variables (total, ADL and IADL) predicted mortality, unadjusted and adjusted for age and gender. The unadjusted HRs for total, ADL and IADL disability were 1.054 (95%-CI: [1.039;1.069]), 1.091 (95%-CI: [1.062;1.121]) and 1.106 (95%-CI: [1.077;1.135]) with p-values < 0.001, respectively. The AUCs were < 0.7, ranging from 0.630 (ADL) to 0.668 (IADL). Multivariate analyses including all 18 disability items revealed that only “Do the shopping” predicted mortality. In addition, multivariate analyses focusing on 11 ADL items and 7 IADL items separately showed that only the ADL item “Get around in the house” and the IADL item “Do the shopping” significantly predicted mortality.
Conclusion: Disability predicted mortality in a seven years follow-up among Dutch community-dwelling older people. It is important that healthcare professionals are aware of disability at early stages, so they can intervene swiftly, efficiently and effectively, to maintain or enhance the quality of life of older people.

Keywords: disability, mortality, community-dwelling older people, activities of daily living

Introduction

The Department of Economic and Social Affairs of the United Nations forecasts that by 2050 33.2% of the population in the Netherlands will be 60 years of age or older.1 Population ageing is accompanied by an increase in the number of people with disability. Prevalence figures of disability among Dutch people aged ≥75 years range from 25.2% to 34.8%, in samples of 377 and 234 individuals, respectively.2,3 There are many different conceptual definitions of disability in circulation. Frequently, disability is defined as having difficulty in carrying out activities of daily living (ADL) and/or instrumental activities of daily living (IADL).4,5 Examples of ADL are washing and drying your whole body, feeding yourself and going up and down the stairs. Activities such as preparing dinner, doing household activities and making the beds belong to IADL. In general, IADL disability precedes ADL disability and represents a less severe form of disability;6 this was demonstrated in a study showing that 54.6% and 67.4% of 377 older people had at least one ADL and IADL disability, respectively.2 Besides a greater age,5,7 gender (women), and low educational level were predictors of disability.8

Previous studies have shown that people with ADL and/or IADL disability have a higher risk of adverse outcomes. Well-known adverse outcomes are an increase in the use of care and related costs,9 and a lower quality of life.2,8,10 In addition, disability is associated with premature death.1118 Only one study has been conducted in the Netherlands;13 disability was measured using three disability indicators (ADL, mobility and the Organization for economic co-operation and development measure) representing different levels of severity of disability in a sample of 15,208 people aged ≥55 years. People suffering from severe disability (ADL, mobility) had an especially higher risk for death compared to people without a disability using a follow-up period of five years.13 It is therefore important that healthcare professionals pay attention to older people with disability and use interventions to prevent disability, and do delay its further progress.

The aim of the present study is to predict mortality by disability in a sample of Dutch community-dwelling older people. Our study distinguished itself from previous studies in several areas. First, the sample only included people aged 75 years or older. Second, the follow-up period was seven years. Third, unlike other studies, we selected the Groningen Activity Restriction Scale (GARS) for assessing disability;19,20 because the GARS contains two subscales (ADL and IADL), we had the opportunity to determine the predictive value of total, ADL and IADL disability for mortality. Finally, we also examined whether mortality could be predicted by individual ADL and IADL items of the GARS.

Methods

Study Population and Data Collection

In June 2008, a questionnaire including validated measures about disability, frailty, quality of life and questions about healthcare utilisation and socio-demographic characteristics was sent to a sample of 1154 community-dwelling individuals aged 75 years or older living in Roosendaal, a municipality with 78,000 inhabitants. The sample was randomly drawn from the register of this municipality. A total of 484 people filled out the questionnaire and returned it, reflecting a response rate of 42%.21

Measures

Disability

As mentioned in the introduction we chose to use the GARS to determine disability.19,20 The GARS is a self-reported questionnaire which contains two subscales. One subscale is focused on ADL disability; this scale includes 11 items; the other subscale measures IADL disability, this scale contains seven items. Each of the 18 items has four response categories: 1) able to perform the activity without any difficulty, 2) able to perform the activity with some difficulty, 3) able to perform the activity with great difficulty, and 4) unable to perform the activity independently. The score for total disability ranges from 18 (no disability) to 72 (maximum disability). For the ADL and IADL subscales the scores ranges from 11 to 44 and from 7 to 28, respectively. A cut-off point of 29 has been established for total disability.22 No cut-off points are known for the ADL and IADL subscales. The GARS has been validated in the Netherlands and demonstrated having good psychometric properties to determine disability among older people.19,20

Mortality

In August 2015, the municipality of Roosendaal provided the dates of death of the people who filled out the questionnaire in June 2008; this created a follow-up period of approximately seven years.

Statistical Analyses

We used descriptive statistics to analyse the baseline characteristics of the participants. Categorical variables were presented as numbers and percentages. Continuous variables were presented as means together with the standard deviations. We defined a time-to-event outcome using the date of mortality of the participants. The time in days was set to 0 (zero) at the time point the first participant died and the time was set to 2613 for participants who were still alive.

For bivariate and multivariate analyses of survival, we used Kaplan–Meier analyses and Cox regression analyses to calculate hazard ratios (HR) with 95% confidence intervals (CI). Predictors in these analyses included the total GARS score, the ADL score, the IADL score and associated items. Since there are no cut-off points for both subscales of the GARS (ADL and IADL), we determined data-driven cut-off points by using a grid of cut-off values for each of these scores. For each cut-off value of a score, sensitivity (se) and specificity (sp) for the prediction of mortality with Cox regression was calculated. The cut-off value that minimised was then defined as the best cut-off value.23 The comparison of the survival curves was carried out using the Log rank test. In additional analyses, we adjusted for age and gender. For all analyses we considered a p-value <0.05 as significant. Variables with a bivariate p-value <0.20 were used in the multivariate analysis.24

The predictive performance of the models was measured using the AUC, the area under the receiver operating characteristics (ROC) curve. An AUC >0.7 was considered as an indication for good predictive performance.24 For the analyses we used R version 3.4.4.25

Ethical Considerations

For this study, medical ethics approval was not necessary as particular treatments or interventions were not offered or withheld from respondents. The integrity of respondents was not encroached upon as a consequence of participating in this study which is the main criterion in medical-ethical procedures in the Netherlands.26 Informed consent in relation to detailing the study and maintaining confidentiality was observed.

Results

Participants Characteristics

As in a previous study using the same sample the data of five participants were not used, because they had too many missing values;27 this resulted in carrying out our analyses on 479 participants. At baseline, the mean age of the participants was 80.3 years (sd=3.8); the majority of the sample was female (56.8%), and 238 participants (49.8%) were married or cohabiting. See Table 1 for more detailed information. Table 1 also presents the scores of the participants for total disability (GARStot), ADL disability (GARSADL) and IADL disability (GARSIADL) at the item level. The items “Dress yourself” to “Take care of your feet and toenails’ belong to ADL and the items” Prepare breakfast or “lunch” through “Do the shopping” belong to IADL. The most prevalent ADL disability was “Take care of your feet and toenails” (38.6%), followed by “Go up and down the stairs” (10.6%). As far as IADL disability is concerned, the participants had the most substantial problems with doing “heavy” household activities (43.6%) and making the beds (26.2%).

Table 1 Participant Characteristics

Prediction of Mortality by Total, ADL and IADL Disability Scores

Table 2 shows the HRs for total, ADL and IADL disability with related 95%-CIs. All three disability variables predicted mortality, unadjusted and adjusted for age and gender, with all p-values <0.001. The unadjusted and adjusted AUCs ranged from 0.630 (ADL) to 0.668 (IADL) and 0.680 (ADL) to 0.696 (IADL), respectively.

Table 2 HRs, CIs, p-values and AUCs for Mortality

Prediction of Mortality by Total, ADL and IADL Disability Based on Cut-off Points

The cut-off values were determined as described in Statistical analyses. For GARStot, GARSADL and GARSIADL the cut-off-values were 25, 14, and 11, respectively. Scores higher then these cut-off values indicated disability. Figure 1 presents the survival plots distinguishing disability from non-disability. The p-values of the Log rank test for the comparison of the survival curves are shown in each plot. In each plot, the survival curves differed significantly (p-values <0.001).

Figure 1 Survival plots distinguishing disability from non-disability.

Prediction of Mortality by the Individual Items of Disability

Visualisation of Bivariate Associations

To get insight into the bivariate associations between the 18 items of the GARS and the association of these items with mortality, we generated a web graph. The thickness of the lines in the graph is based on Cramers V (CV), a statistic derived from the Chi-square statistic, that ranges from 0 to 1 (values towards 1 indicate stronger association). Figure 2 presents the associations between the 18 items of the GARS and mortality with CV ≥0.25. See Table 1 for interpreting the numbers in the graph. The graph shows that mortality was associated with the items “8. Get around in the house (if necessary, with a cane)”, “14. Do ‘light’ household activities”, and “18. Do the shopping” (CV-values 0.27, 0.28, and 0.30 respectively).

Figure 2 Association mortality and GARS items.

Multivariate Analyses

First, we imputed data for the missing values with regard to the total disability items using the MICE package in the R software (m=5 and methods= “polyreg”).28 Subsequently, bivariate analyses were carried out which aimed to determine which of the 18 individual items had a p-value <0.20 with respect to mortality; the items that complied were included in the multivariate analyses, which applied to all disability items. The Cox regression including all 18 items showed an unadjusted AUC of 0.682 and an adjusted AUC of 0.715. The multivariate analyses revealed that only the disability item “Do the shopping” predicted mortality. The unadjusted HR was 1.288 [1.050; 1.580], with p-value 0.015 and the adjusted HR was 1.320 [1.069; 1.631] with p-value 0.010. For the HRs of the other disability items we refer to Table 3.

Table 3 HRs, CIs and p-values for Mortality per Item for Total Disability and Domains

For the determination of the predictive value of the disability items belonging to the GARS ADL subscale and the GARS IADL subscale we followed exactly the same steps as aforementioned. The results are also presented in Table 3. For ADL, “Get around in the house (if necessary with a cane)” was the only item that predicted mortality. The unadjusted HR was 1.694 [1.141; 2.514] with p-value 0.009 and the adjusted HR was 1.871 [1.239; 2.825] with a p-value 0.003. For IADL, not surprisingly “Do the shopping” was the item that predicted mortality. The unadjusted and adjusted HR was 1.244 [1.041; 1.486] with p-value 0.016 and 1.280 [1.064; 1.540] with p-value 0.009, respectively. No other IADL item predicted mortality. See Table 3 for further details.

Discussion

Our study showed that all three disability variables (total, ADL and IADL) predicted mortality, unadjusted and adjusted for age and gender. This finding was supported by comparing the prediction of mortality by disabled and non-disabled older people. All

AUCs were <0.7, ranging from 0.630 (ADL) to 0.668 (IADL). As mentioned in the introduction, to the best of our knowledge, just one other study carried out in the Netherlands has examined the prediction of mortality by disability.13 This study also showed that ADL disabled people had a higher risk of mortality than ADL non-disabled people; their life expectancy reduced by 10 years. However, it should be noted that a good comparison of the results is not possible because the samples are very different in age (≥75 years versus ≥55 years) and the instruments used for data collection varied (GARS versus an ADL measure including 5 items: eating and drinking, dressing, washing oneself completely, washing face and hands, transfer from a chair).

Other previous studies showing that disability predicted mortality were in particular conducted in the USA.11,12,14,15,18 For instance, in a sample consisting of 1495 patients aged ≥70 years or older discharged from a hospital ADL disability predicted 1-year mortality, after adjustment for other independent variables in a prognostic index (e.g. gender, medical diagnoses, laboratory values as creatinine level).12 In addition, ADL disability, as well as IADL disability stages, predicted three-year mortality in Medicare beneficiaries of 65 years or older; the HRs between the two types were similar.15 Moreover, in a sample of 4516 community-dwelling older people (≥70 years), ADL and IADL disability predicted two-year mortality.14 Finally, analyses based on data from the National Health and Nutrition Examination Survey, including 1834 people in the age range 60–84 years demonstrated an HR 2.23 (95%-CI: 1.29–3.85), after adjustment for potential confounders (e.g. age, gender) during a follow-up of 5.7 years.18 The adjusted HRs for ADL and IADL were 1.74 (95%-CI: 0.72–4.16) and 1.57 (95%-CI: 0.76–3.27), respectively.

Our results are also supported by studies carried out in countries other than the Netherlands and the USA. In Spain, in a sample of 598 older persons (≥65 years) a significant association was found between disability in ADL and IADL and mortality.16 In Brazil, using data from 1333 community-dwelling persons aged 60 years or older the overall mortality rate appeared to be 46.1 per 1000 person-years at risk. The HRs for ADL disability in men and women were 1.65 (95%-CI: 1.11–2.45) and 1.43 (95%-CI: 1.05–1.95), respectively. For IADL, the HR for men was 2.07 (95%-CI: 1.53–2.79) and for women 1.43 (95%-CI: 1.11–1.84).17

In addition, multivariate analyses including all 18 disability items demonstrated AUCs of 0.682 (unadjusted) and 0.715 (adjusted) and revealed that only the item “Do the shopping” predicted mortality. Moreover, multivariate analyses focusing on 11 ADL items and 7 items separately, showed that only the ADL item “Get around in the house (if necessary with a cane)” and the IADL item “Do the shopping” significantly predicted mortality. Multivariate analyses in a USA study showed for “Dependency in shopping” an adjusted OR of 1.9 (95%-CI: 1.5–2.5) for predicting two-year mortality.14 In the same study, bivariate analyses demonstrated that “Walking across the room” was associated with two-year mortality (OR 4.6 95%-CI: 3.6–5.3).14

A few limitations of our study should be noted. Firstly, disability was assessed subjectively, because these data were based on self-report. However, we used an instrument with good psychometric properties, the GARS, for measuring disability.19,20 Secondly, sampling was carried out by only one municipality in the Netherlands; therefore, there may be doubts about the generalisability of the results. Thirdly, we only adjusted for age and gender. Other predictors of mortality such as educational level and frailty were disregarded.29,30

In conclusion, in this longitudinal study we found that disability (total, ADL and IADL) predicted mortality in a seven years follow-up among Dutch community-dwelling people aged 75 years or older. It is important that healthcare professionals are aware of disability at an early stages, so they can intervene swiftly, efficiently and effectively, with the aim to maintain the independence and autonomy of older individuals and prevent or delay adverse effects of disability, which do not necessarily affect their quality of life. In particular, attention should be paid to the IADL disability item “Do the shopping”, because this item was significantly associated with mortality, and besides that, a substantial part of the participants indicated they were longer able to do this activity themselves (15.2%).

Disclosure

The authors report no conflicts of interest for this work.

References

1. United Nations Department of Economic and Social Affairs. Population division. world population prospects. the 2015 revision. key findings and advance tables. Working paper No. ESA/P/WP.241. New York. 2015. Available from: http://esa.un.org/unpd/wpp/publications/files/key_findings_wpp_2015.pdf. Accessed July 19 2017.

2. Gobbens RJ. Associations of ADL and IADL disability with physical and mental dimensions of quality of life in people aged 75 years and older. PeerJ. 2018;6:e5425. doi:10.7717/peerj.5425

3. Gobbens RJ, van Assen MA, Luijkx KG, Wijnen-Sponselee MT, Schols JM. The tilburg frailty indicator: psychometric properties. J Am Med Dir Assoc. 2010;11(5):344–355.

4. Tas U, Verhagen AP, Bierma-Zeinstra SM, et al. Incidence and risk factors of disability in the elderly: the Rotterdam Study. Prev Med. 2007;44(3):272–278.

5. Tas U, Verhagen AP, Bierma-Zeinstra SM, Odding E, Koes BW. Prognostic factors of disability in older people: a systematic review. Br J Gen Pract. 2007;57(537):319–323.

6. Wong CH, Weiss D, Sourial N, et al. Frailty and its association with disability and comorbidity in a community-dwelling sample of seniors in Montreal: a cross-sectional study. Aging Clin Exp Res. 2010;22(1):54–62. doi:10.1007/BF03324816

7. Fujiwara Y, Yoshida H, Amano H, et al. Predictors of improvement or decline in instrumental activities of daily living among community-dwelling older Japanese. Gerontology. 2008;54(6):373–380. doi:10.1159/000151221

8. den Ouden ME, Schuurmans MJ, Mueller-Schotte S, Brand JS, van der Schouw YT. Domains contributing to disability in activities of daily living. J Am Med Dir Assoc. 2013;14(1):18–24. doi:10.1016/j.jamda.2012.08.014

9. Fried LP, Ferrucci L, Darer J, Williamson JD, Anderson G. Untangling the concepts of disability, frailty, and comorbidity: implications for improved targeting and care. J Gerontol a Biol Sci Med Sci. 2004;59(3):255–263. doi:10.1093/gerona/59.3.M255

10. Akosile CO, Mgbeojedo UG, Maruf FA, Okoye EC, Umeonwuka IC, Ogunniyi A. Depression, functional disability and quality of life among Nigerian older adults: prevalences and relationships. Arch Gerontol Geriatr. 2018;74:39–43. doi:10.1016/j.archger.2017.08.011

11. Manton KG. A longitudinal study of functional change and mortality in the United States. J Gerontol. 1988;43(5):S153–S161. doi:10.1093/geronj/43.5.S153

12. Walter LC, Brand RJ, Counsell SR, et al. Development and validation of a prognostic index for 1-year mortality in older adults after hospitalization. JAMA. 2001;285(23):2987–2994. doi:10.1001/jama.285.23.2987

13. Majer IM, Nusselder WJ, Mackenbach JP, Klijs B, van Baal PH. Mortality risk associated with disability: a population-based record linkage study. Am J Public Health. 2011;101(12):e9–e15. doi:10.2105/AJPH.2011.300361

14. Carey EC, Walter LC, Lindquist K, Covinsky KE. Development and validation of a functional morbidity index to predict mortality in community-dwelling elders. J Gen Intern Med. 2004;19(10):1027–1033. doi:10.1111/j.1525-1497.2004.40016.x

15. Hennessy S, Kurichi JE, Pan Q, et al. Disability stage is an independent risk factor for mortality in medicare beneficiaries aged 65 years and older. PM R. 2018;74(12):1215–1225. doi:10.1016/j.pmrj.2015.05.014

16. Millan-Calenti JC, Tubio J, Pita-Fernandez S, et al. Prevalence of functional disability in activities of daily living (ADL), instrumental activities of daily living (IADL) and associated factors, as predictors of morbidity and mortality. Arch Gerontol Geriatr. 2010;50(3):306–310. doi:10.1016/j.archger.2009.04.017

17. Nascimento CM, Oliveira C, Firmo JOA, Lima-Costa MF, Peixoto SV. Prognostic value of disability on mortality: 15-year follow-up of the Bambui cohort study of aging. Arch Gerontol Geriatr. 2018;74:112–117. doi:10.1016/j.archger.2017.10.011

18. Wu LW, Chen WL, Peng TC, et al. All-cause mortality risk in elderly individuals with disabilities: a retrospective observational study. BMJ Open. 2016;6(9):e011164. doi:10.1136/bmjopen-2016-011164

19. Kempen GI, Miedema I, Ormel J, Molenaar W. The assessment of disability with the groningen activity restriction scale. Conceptual framework and psychometric properties. Soc Sci Med. 1996;43(11):1601–1610. doi:10.1016/S0277-9536(96)00057-3

20. Suurmeijer TP, Doeglas DM, Moum T, et al. The Groningen Activity Restriction Scale for measuring disability: its utility in international comparisons. Am J Public Health. 1994;84(8):1270–1273. doi:10.2105/AJPH.84.8.1270

21. Gobbens RJ, van Assen MA, Luijkx KG, Wijnen-Sponselee MT, Schols JM. Determinants of frailty. J Am Med Dir Assoc. 2010;11(5):356–364. doi:10.1016/j.jamda.2009.11.008

22. Ormel J, Rijsdijk FV, Sullivan M, van Sonderen E, Kempen GI. Temporal and reciprocal relationship between IADL/ADL disability and depressive symptoms in late life. J Gerontol B Psychol Sci Soc Sci. 2002;57(4):P338–P347. doi:10.1093/geronb/57.4.P338

23. Unal I. Defining an optimal cut-point value in ROC analysis: an alternative approach. Comput Math Methods Med. 2017;2017:3762651. doi:10.1155/2017/3762651

24. Steyerberg E. Clinical Prediction Models. A Practical Approach to Development, Validation, and Updating. 1 ed. New York: Springer-Verlag; 2009.

25. R Core Team. R: a language and environment for statistical computing. Available from: https://www.R-project.org. Accessed February 3 2020.

26. Central Committee on Research Involving Human Subjects. Available from: http://www.ccmo.nl/en/your-research-does-it-fall-under-the-wmo. Accessed June 20 2016.

27. Gobbens RJ, van Assen MA, Luijkx KG, Schols JM. The predictive validity of the Tilburg Frailty Indicator: disability, health care utilization, and quality of life in a population at risk. Gerontologist. 2012;52(5):619–631. doi:10.1093/geront/gnr135

28. van Buuren S, Groothuis-Oudshoorn K. Mice: multivariate imputation by chained equations in R. J Stat Softw. 2011;45(i03): 1–68.

29. Zhou S, Zou G, Chen X, et al. Educational attainment and mortality: results from the sixth population census in China. J Glob Health. 2019;9(2):020604. doi:10.7189/jogh.09.020604

30. Shamliyan T, Talley KM, Ramakrishnan R, Kane RL. Association of frailty with survival: a systematic literature review. Ageing Res Rev. 2013;12(2):719–736. doi:10.1016/j.arr.2012.03.001

Creative Commons License This work is published and licensed by Dove Medical Press Limited. The full terms of this license are available at https://www.dovepress.com/terms.php and incorporate the Creative Commons Attribution - Non Commercial (unported, v3.0) License. By accessing the work you hereby accept the Terms. Non-commercial uses of the work are permitted without any further permission from Dove Medical Press Limited, provided the work is properly attributed. For permission for commercial use of this work, please see paragraphs 4.2 and 5 of our Terms.

Download Article [PDF]