Abstract
Background
We observed a lack of population-based longitudinal research examining the association of disadvantaged childhood socioeconomic circumstances (CSC) and disability [activities of daily living (ADL) and instrumental activities of daily living (IADL)] in older age, and whether socioeconomic attainments in adulthood can compensate for a poor socioeconomic start in life.
Methods
Data on 24 440 persons aged 50–96 in 14 European countries (Survey of Health, Ageing and Retirement in Europe) were used to measure the associations between CSC and limitations with ADL and with IADL, using mixed-effects logistic regression models. Models stratified by gender were adjusted for education during young adulthood, main occupation during middle age, ability to make ends meet during old age and potential confounding and control variables.
Results
Risks of ADL and IADL limitations increased with age and were different between women and men. For women, a gradient across CSC strata was observed, showing that the more disadvantaged the CSC, the higher the risk of ADL and IADL limitations in old age, even after adjustment for adult socioeconomic indicators. For men, the association between CSC and disability was mediated by the main occupation in middle age and the ability to make ends meet at older age.
Conclusion
Women who grew up in socioeconomically disadvantaged households were at higher risk of disability in older age and this disadvantage was not attenuated by favourable adult socioeconomic conditions. Men were more likely to make up for a disadvantaged start in adulthood.
Introduction
The risk of disability increases with age, peaks at very old age,1 and affects women more than men.2 At least 5.5% of the elderly in Western countries encounter difficulties with activities of daily living (ADL), and at least 15% of them have problems with instrumental activities of daily living (IADL).3,4
The disablement process framework posits that biological, psychological and social factors contribute to the development of physical disability.5 More specifically, socioeconomic circumstances were found to be a latent cause of functional limitations.6 Most studies focus on socioeconomic indicators in middle adulthood, although the development of ADL and IADL limitations seems to be linked to circumstances in early life, which subsequently influence both health and socioeconomic life trajectories.7,8 Only a few studies investigated the association between disability in late adulthood and childhood socioeconomic circumstances (CSC).9–11 Some found the relationship between CSC and disability at old age to be partly mediated by socioeconomic attainments in mid-life.12,13 Yet, there is no clear evidence if this link can be completely broken by educational and professional opportunities, and whether mediation by adult socioeconomic attainments differ between men and women, assuming that women in older cohorts had more restricted access to education14 and employment15 during their life course.
While CSC seems to be a promising predictor of disability in later life, longitudinal data research is still lacking in Europe. Hence, the objectives of this study are to examine the associations between CSC and the risk of ADL and IADL limitations at older age, using a large European longitudinal survey [Survey of Health, Ageing and Retirement in Europe (SHARE)]. A potential mediation by adult socioeconomic indicators was tested.
Methods
Study design
This study used repeated measurements of health and socioeconomic circumstances on participants aged between 50 and 96 years living in 14 European countries of the longitudinal SHARE.16 The data was collected between 2004 and 2015 at two-year intervals. In wave 3, respondents answered a retrospective questionnaire (SHARELIFE), which focussed on their life histories.17 The study sample included individuals who participated in wave 3 and who provided one or more outcome measurements (ADL or IADL) in any of the other five waves (waves 1, 2, 4, 5 and 6). The analytical sample consisted of 24 440 participants (detailed flow chart in Supplementary material—Supplementary figure S1).
Dependent variables
Functional limitations in ADL and IADL served to measure disability. The validated ADL measure18 implemented in SHARE includes six everyday self-care activities: dressing (including putting on shoes and socks), walking across a room, bathing or showering, eating, getting in or out of bed and using the toilet (including getting up or down).19 For our analyses, we categorized individuals as having ‘no ADL limitations’ or ‘1 or more ADL limitations’.
The IADL index20 defines the number of limitations with instrumental activities of everyday life. The validated IADL measure implemented in SHARE includes the following seven activities: using a map to figure out how to get around in a strange place, preparing a hot meal, shopping for groceries, making telephone calls, taking medications, doing work around the house or garden and managing money (such as paying bills and keeping track of expenses). The variable was dichotomized into ‘no IADL limitations’ and ‘1 or more IADL limitations’.
Childhood socioeconomic circumstances
Four retrospective indicators of socioeconomic circumstances at the age of 10 (SHARELIFE; wave 3), assessing the long-term effects of CSC on health,21–24 were used to compute the CSC index.25 Including binary measures of the occupational position of the main breadwinner,26,27 the number of books at home,23 household overcrowding,24 and housing quality,22 CSC resulted in a five-category variable ranging from ‘most disadvantaged’ to ‘most advantaged’ (details on variable composition in Supplementary materials— Supplementary table S1).
Adult socioeconomic indicators
Three adult socioeconomic indicators were added as potential mediators. First, based on the International Standard Classification of Education classification, participants’ highest educational attainment was integrated by grouping participants with a tertiary education level into ‘tertiary’, those who reached at least secondary level into ‘secondary’, and those who finished only primary education into ‘primary’. Second, participants’ main occupational position was constructed in the same manner as the occupation for the main breadwinner during childhood, with the exception that individuals, who had never done paid work, were grouped separately as ‘never worked’. This classification was chosen to account for physically demanding work, rather than opting for a proxy of social class, which can also be influenced by other household members. Third, the mode of the answers to the question ‘Is the household able to make ends meet?’, ranging from 1 (‘with great difficulty’) to 4 (‘easily’), was calculated across waves (1, 2, 4, 5 and 6) for each participant.
Confounders and controls
Due to their potential association with disability,28–30 the following two confounding variables were included in all models: birth cohort [1919–28, 1929–38 (Great Depression), 1939–45 (World War II) and after 1945), and country of residence (Austria, Belgium, Czech Republic, Denmark, France, Germany, Greece, Ireland, Italy, Netherlands, Poland, Spain, Sweden and Switzerland). Additionally, all models were adjusted with two control variables due to their observed association with disability:31 living with biological parents at the age of 10 (both parents, one biological parent, without biological parents) and participant attrition (no dropout, dropout, death).
Covariates
The following three covariates were included: partnership status (living with a partner vs. alone),32 chronic disease score (<2 vs. 2 or more of the following conditions: stroke, heart attack, hypertension, diabetes, cancer, Parkinson’s disease and asthma),13 and body mass index (BMI, average of participants’ measures over the waves).5
Statistical analysis
Mixed-effects logistic regression models were estimated with ADL and IADL as dependent variables. Observations (i.e. Level 1) were nested within participants (i.e. Level 2). Different nested models with various random effects were tested to identify the best random structure using Bayesian information criteria and likelihood ratio tests (data not shown). The best random structure included a random intercept for participants. Analyses were stratified by gender following the literature on disability trajectories at older age2. Interactions between gender and CSC were significant.
Model 1 examined the association between CSC and ADL, adjusting for prior confounders (birth cohorts and countries) and control variables (living with biological parents, participant attrition). Age was centred at the midpoint of the sample’s age range (73 years). To test whether CSC moderated the effect of ageing on ADL, the models included interaction terms between CSC and age. Sequentially, main effects and interactions of age with educational attainment (Model 2), main occupation (Model 3) and ability to make ends meet (Model 4) were added as potential mediators. Finally, partnership status, number of chronic diseases and BMI were included as covariates (Model 5). Moreover four sensitivity analyses were performed: (i) excluding participants who died during follow-up; (ii) excluding participants who dropped out during follow-up; (iii) excluding participants with visual impairment, using the validated measure eyesight,33 to check if visual acuity was on the pathway between CSC and disability; (iv) adjusting with an index of the following binary health behaviours: physical inactivity, unhealthy eating, smoking and alcohol consumption.34 The same sequential analytic strategy was used to examine IADL. A robustness analysis with the index of CSC was conducted using a less restrictive threshold for the indicator of overcrowding (≥1 bedroom per household member). Results (not shown) were similar compared to main analysis. Multicollinearity was quantified using variance inflations factors (VIF). In all final models (i.e. IADL and ADL both in men and women), the highest VIF across socioeconomic predictors was lower than 10 (the highest VIF was 3.38). R, and the lme4 and lmerTest packages were used to estimate the models and parameters.
Results
Participants’ characteristics
The sample of 24 440 participants stratified by gender is described in table 1. The distribution in CSC strata was similar between women and men. Overall, men had better socioeconomic conditions in adulthood (education, main occupational position, ability to make ends meet) compared to women. At baseline, women reported more limitations with ADL and IADL compared to men.
Table 1.
Characteristics of participants
| Women | Men | |
|---|---|---|
| Age, years, mean(SD) | 62.36(9.2) | 62.75(8.9) |
| N(%) | N(%) | |
| Birth cohort | ||
| Between 1919 and 1928 | 1386(10.2) | 990(9.2) |
| Between 1929 and 1938 | 3049(22.3) | 2660(24.7) |
| Between 1939 and 1945 | 3052(22.4) | 2628(24.4) |
| After 1945 | 6161(45.1) | 4514(41.8) |
| Childhood socioeconomic circumstances at the age of 10 years | ||
| Most disadvantaged | 2498(18.3) | 2059(19.1) |
| Disadvantaged | 3473(25.5) | 2693(25.0) |
| Middle | 4452(32.6) | 3399(31.5) |
| Advantaged | 2469(18.1) | 2019(18.7) |
| Most advantaged | 756(5.5) | 622(5.8) |
| Living with biological parents at the age of ten | ||
| Both parents | 12 339(90.4) | 9785(90.7) |
| One biological parent | 1035(7.6) | 815(7.6) |
| Without biological parent | 274(2.0) | 192(1.8) |
| Level of educational achievement | ||
| Primary | 3878(28.4) | 2479(23.0) |
| Secondary | 7434(54.5) | 5780(53.6) |
| Tertiary | 2336(17.1) | 2533(23.5) |
| Skills of main occupational position during active life | ||
| High-skilled | 2237(16.4) | 3362(31.2) |
| Low-skilled | 9444(69.2) | 7341(68.0) |
| Never worked | 1967(14.4) | 89(0.8) |
| Ability to make ends meet | ||
| Easily | 4798(35.2) | 4257(39.5) |
| Fairly easily | 4189(30.7) | 3353(31.1) |
| With some difficulty | 3091(22.7) | 2195(20.3) |
| With great difficulty | 1570(11.5) | 987(9.2) |
| Partnership status | ||
| Alone | 4315(31.6) | 1787(16.6) |
| Living with a partner | 9333(68.4) | 9005(83.4) |
| Number of chronic diseases | ||
| 2 or more | 6129(44.9) | 4071(37.7) |
| Less than 2 | 7519(55.1) | 6721(62.3) |
| Body mass index (continuous) | 26.44(4.5) | 26.99(3.7) |
| Countries | ||
| Austria | 495 (3.6) | 338(3.1) |
| Belgium | 1381(10.1) | 1141(10.6) |
| Czech Republic | 920(6.74) | 671(6.2) |
| Denmark | 1065(7.8) | 893(8.3) |
| France | 1204(8.8) | 900(8.3) |
| Germany | 898(6.6) | 782(7.3) |
| Greece | 1537(11.3) | 1194(11.1) |
| Ireland | 338(2.5) | 264(2.5) |
| Italy | 1270(9.3) | 1057(9.8) |
| Netherlands | 1039(7.6) | 870(8.1) |
| Poland | 920(6.7) | 702(6.5) |
| Spain | 1002(7.3) | 747(6.9) |
| Sweden | 924(6.8) | 734(6.8) |
| Switzerland | 655(4.8) | 499(4.6) |
| Attrition | ||
| No dropout | 9766(71.6) | 7312(67.8) |
| Dropout | 2881(21.1) | 2333(21.6) |
| Death | 1001(7.3) | 1147(10.6) |
| Disability | ||
| 1 or more ADL limitations, at baseline | 1223(9.0) | 709(6.6) |
| 1 or more IADL limitations, at baseline | 2221(16.3) | 904(8.4) |
Abbreviations: ADL, activities of daily living; IADL, instrumental activity of daily living; SD, standard deviation.
Childhood socioeconomic circumstances and limitations with ADL
ADL limitations increased slowly with age among men and women in all adjusted models (table 2). For men, CSC did not change the effect of ageing on ADL trajectories (interactions of CSC and age were not significant; Supplementary figure S3 and table 2), whereas for women the differences between CSC groups increased with age (Supplementary figure S3 and table 2).
Table 2.
Associations between childhood socioeconomic circumstances and trajectories of disability at older age, ADL
| Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| OR (95%CI) | OR (95%CI) | OR (95%CI) | OR (95%CI) | OR (95%CI) | ||||||
| Women | ||||||||||
| Age | 1.19 (1.16–1.21) | *** | 1.18 (1.16–1.21) | *** | 1.18 (1.15–1.22) | *** | 1.16 (1.13–1.20) | ** | 1.15 (1.11–1.18) | ** |
| Childhood socioeconomic circumstances (ref. Most disadvantaged) | ||||||||||
| Disadvantaged | 0.68 (0.54–0.86) | ** | 0.69 (0.55–0.87) | ** | 0.70 (0.56–0.88) | ** | 0.78 (0.63–0.97) | * | 0.87 (0.72–1.06) | |
| Middle | 0.54 (0.43–0.69) | *** | 0.58 (0.45–0.74) | *** | 0.59 (0.46–0.75) | *** | 0.69 (0.54–0.88) | ** | 0.76 (0.62–0.95) | * |
| Advantaged | 0.43 (0.32–0.57) | *** | 0.48 (0.36–0.65) | *** | 0.51 (0.38–0.68) | *** | 0.59 (0.45–0.79) | *** | 0.71 (0.55–0.92) | * |
| Most advantaged | 0.30 (0.19–0.46) | *** | 0.37 (0.24–0.59) | *** | 0.39 (0.25–0.61) | *** | 0.47 (0.31–0.73) | ** | 0.65 (0.44–0.97) | * |
| Age × childhood socioeconomic circumstances (ref. Most disadvantaged) | ||||||||||
| Age × disadvantaged | 0.97 (0.95–0.99) | ** | 0.97 (0.95–0.99) | ** | 0.97 (0.95–0.99) | ** | 0.97 (0.95–0.99) | ** | 0.98 (0.96–0.99) | * |
| Age × middle | 0.97 (0.95–0.99) | ** | 0.97 (0.95–0.99) | ** | 0.97 (0.95–0.99) | ** | 0.97 (0.96–1.00) | * | 0.98 (0.96–1.00) | |
| Age × advantaged | 0.96 (0.94–0.99) | ** | 0.96 (0.94–0.99) | ** | 0.96 (0.94–0.99) | ** | 0.96 (0.94–0.99) | ** | 0.97 (0.95–0.99) | * |
| Age × most advantaged | 0.96 (0.93–1.00) | * | 0.96 (0.92–1.00) | * | 0.96 (0.92–0.99) | * | 0.96 (0.93–1.00) | 0.98 (0.95–1.02) | ||
| Education (ref. Primary) | ||||||||||
| Secondary | – | 1.01 (0.83–1.21) | 1.00 (0.83–1.21) | 1.06 (0.88–1.27) | 1.11 (0.94–1.30) | |||||
| Tertiary | – | 0.57 (0.42–0.77) | *** | 0.63 (0.45–0.89) | ** | 0.76 (0.55–1.06) | 0.90 (0.67–1.21) | |||
| Age × education (ref. Primary) | ||||||||||
| Age × secondary | – | 1.00 (0.98–1.02) | 1.00 (0.99–1.02) | 1.01 (0.99–1.02) | 1.00 (0.99–1.02) | |||||
| Age × tertiary | – | 1.00 (0.97–1.02) | 1.00 (0.97–1.03) | 1.00 (0.97–1.03) | 1.00 (0.98–1.03) | |||||
| Main occupational position (ref. High-skilled) | ||||||||||
| Low-skilled | – | – | 1.34 (1.00–1.78) | 1.25 (0.94–1.65) | 1.27 (0.98–1.64) | * | ||||
| Never worked | – | – | 1.64 (1.15–2.34) | ** | 1.41 (1.00–1.99) | 1.45 (1.06–1.99) | * | |||
| Age × main occupational position (ref. High-skilled) | ||||||||||
| Age × low-skilled | – | – | 0.99 (0.97–1.02) | 1.00 (0.97–1.02) | 1.00 (0.98–1.02) | |||||
| Age × never worked | – | – | 1.03 (0.99–1.06) | 1.03 (1.00–1.06) | * | 1.03 (1.00–1.06) | * | |||
| Ability to make ends meet (ref. Easily) | – | – | – | |||||||
| Fairly easily | – | – | – | 1.34 (1.10–1.64) | ** | 1.14 (0.95–1.37) | ||||
| With some difficulty | – | – | – | 2.09 (1.67–2.63) | *** | 1.56 (1.27–1.92) | *** | |||
| With great difficulty | – | – | – | 5.62 (4.22–7.48) | *** | 3.31 (2.56–4.29) | *** | |||
| Age × ability to make ends meet (ref. Age × easily) | – | – | – | |||||||
| Age × fairly easily | – | – | – | 1.02 (1.01–1.04) | * | 1.02 (1.00–1.04) | * | |||
| Age × with some difficulty | – | – | – | 1.00 (0.99–1.02) | 1.00 (0.99–1.02) | |||||
| Age × with great difficulty | – | – | – | 1.00 (0.98–1.02) | 1.00 (0.98–1.02) | |||||
| Men | ||||||||||
| Age | 1.24 (1.21–1.27) | *** | 1.23 (1.20–1.27) | *** | 1.23 (1.18–1.28) | *** | 1.24 (1.19–1.30) | *** | 1.22 (1.17–1.26) | *** |
| Childhood socioeconomic circumstances (ref. Most disadvantaged) | ||||||||||
| Disadvantaged | 0.75 (0.53–1.06) | 1.01 (0.71–1.43) | 0.86 (0.61–1.22) | 0.93 (0.66–1.30) | 0.93 (0.69–1.25) | |||||
| Middle | 0.73 (0.51–1.04) | 1.07 (0.74–1.54) | 0.95 (0.66–1.38) | 1.00 (0.70–1.44) | 1.03 (0.76–1.42) | |||||
| Advantaged | 0.64 (0.42–0.96) | 1.06 (0.69–1.64) | 0.93 (0.61–1.44) | 1.02 (0.67–1.57) | 1.00 (0.69–1.45) | |||||
| Most advantaged | 0.65 (0.36–1.16) | 0.99 (0.54–1.83) | 1.04 (0.56–1.94) | 1.18 (0.64–2.17) | 1.14 (0.67–1.94) | |||||
| Age × childhood socioeconomic circumstances (ref. Most disadvantaged) | ||||||||||
| Age × disadvantaged | 0.98 (0.95–1.01) | 0.97 (0.94–1.00) | 0.97 (0.94–1.00) | 0.97 (0.94–1.01) | 0.98 (0.95–1.01) | |||||
| Age × middle | 0.98 (0.95–1.01) | 0.97 (0.94–1.00) | * | 0.97 (0.94–1.00) | 0.97 (0.94–1.00) | 0.98 (0.95–1.00) | ||||
| Age × advantaged | 0.97 (0.94–1.01) | 0.96 (0.93–1.00) | 0.97 (0.93–1.00) | 0.96 (0.93–1.00) | * | 0.97 (0.94–1.00) | * | |||
| Age × most advantaged | 1.00 (0.95–1.06) | 0.98 (0.93–1.04) | 1.00 (0.94–1.05) | 0.99 (0.94–1.05) | 0.98 (0.94–1.03) | |||||
| Education (ref. Primary) | ||||||||||
| Secondary | – | 0.75 (0.55–1.00) | 0.78 (0.58–1.05) | 0.82 (0.61–1.09) | 0.93 (0.72–1.20) | |||||
| Tertiary | – | 0.46 (0.31–0.67) | *** | 0.55 (0.36–0.83) | ** | 0.59 (0.39–0.89) | * | 0.70 (0.49–1.01) | ||
| Age × education (ref. Primary) | ||||||||||
| Age × secondary | – | 1.01 (0.99–1.04) | 1.01 (0.99–1.04) | 1.01 (0.98–1.03) | 1.01 (0.98–1.03) | |||||
| Age × tertiary | – | 1.03 (1.00–1.07) | 1.03 (0.99–1.07) | 1.02 (0.98–1.06) | 1.02 (0.98–1.05) | |||||
| Main occupational position (ref. High-skilled) | ||||||||||
| Low-skilled | – | – | 1.41 (1.05–1.91) | * | 1.25 (0.93–1.67) | 1.28 (0.99–1.66) | ||||
| Never worked | – | – | 2.92 (0.80–10.66) | 2.65 (0.75–9.39) | 1.67 (0.58–4.84) | |||||
| Age × main occupational position (ref. High-skilled) | ||||||||||
| Age × low-skilled | – | – | 1.00 (0.98–1.03) | 1.00 (0.97–1.02) | 1.00 (0.97–1.02) | |||||
| Age × never worked | – | – | 1.01 (0.91–1.13) | 1.00 (0.90–1.11) | 0.96 (0.88–1.05) | |||||
| Ability to make ends meet (ref. Easily) | – | – | – | |||||||
| Fairly easily | – | – | – | 1.41 (1.06–1.88) | * | 1.34 (1.04–1.72) | * | |||
| With some difficulty | – | – | – | 2.05 (1.44–2.92) | *** | 1.83 (1.35–2.49) | *** | |||
| With great difficulty | – | – | – | 4.45 (2.76–7.17) | *** | 3.97 (2.63–5.99) | *** | |||
| Age × ability to make ends meet (ref. Age × easily) | – | – | – | |||||||
| Age × fairly easily | – | – | – | 1.00 (0.97–1.02) | 0.99 (0.97–1.01) | |||||
| Age × with some difficulty | – | – | – | 0.99 (0.96–1.01) | 0.99 (0.97–1.02) | |||||
| Age × with great difficulty | – | – | – | 1.00 (0.96–1.04) | 1.00 (0.96–1.03) |
Abbreviations: ADL, activities of daily living; OR, odds ratio; CI, confidence interval.
Model 1: Adjusted for confounders and controls: birth cohorts, countries, attrition and living with biological parents.
Model 2: model 1 adjusted for young adulthood socioeconomic circumstances (education).
Model 3: model 2 adjusted for adulthood socioeconomic circumstances (main occupation).
Model 4: model 3 adjusted for old age socioeconomic circumstances (current ability to make ends meet with the household income).
Model 5: model 4 adjusted for covariates: partnership status, number of chronic diseases and body mass index.
P-values:
P<0.001,
P<0.01,
P<0.05.
Among women, a gradient across CSC strata was observable, showing that the more advantaged the CSC, the lower the risk of ADL limitations (Model 1). This association was robust across models, although attenuated by covariates (Model 5). Furthermore, difficulties to make ends meet were associated with a greater risk of ADL limitations in the fully adjusted model (Model 5), but it did not completely mediate the association with CSC.
Among men, CSC were not associated with the risk of ADL limitations (Model 1). In Model 4, greater difficulty in making ends meet was associated with a higher risk of ADL limitations, whereas holding a tertiary degree was associated with a lower risk. Only difficulty in making ends meet remained associated after adjusting for partnership and health status (Model 5).
Childhood socioeconomic circumstances and limitations with IADL
All five models showed a slow increase of IADL limitations with age among men and women (table 3). The risk of IADL limitations increased with ageing more slowly among women who grew up in middle and advantaged CSC compared to those who lived in the most disadvantaged circumstances (age and CSC interactions), but this was not found among men (Supplementary figure S4).
Table 3.
Associations between childhood socioeconomic circumstances and trajectories of disability at older age, IADL
| Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| OR (95%CI) | OR (95%CI) | OR (95%CI) | OR (95%CI) | OR (95%CI) | ||||||
| Women | ||||||||||
| Age | 1.15 (1.14–1.16) | *** | 1.15 (1.13–1.16) | *** | 1.15 (1.12–1.17) | *** | 1.15 (1.12–1.18) | *** | 1.14 (1.12–1.17) | *** |
| Childhood socioeconomic circumstances (ref. Most disadvantaged) | ||||||||||
| Disadvantaged | 0.67 (0.57–0.79) | *** | 0.70 (0.60–0.82) | *** | 0.71 (0.61–0.83) | *** | 0.76 (0.65–0.89) | ** | 0.80 (0.69–0.93) | ** |
| Middle | 0.48 (0.41–0.57) | *** | 0.55 (0.46–0.65) | *** | 0.56 (0.47–0.67) | *** | 0.63 (0.53–0.74) | *** | 0.66 (0.56–0.78) | *** |
| Advantaged | 0.41 (0.33–0.49) | *** | 0.50 (0.41–0.61) | *** | 0.52 (0.42–0.64) | *** | 0.58 (0.48–0.71) | *** | 0.65 (0.53–0.79) | *** |
| Most advantaged | 0.28 (0.21–0.38) | *** | 0.40 (0.29–0.54) | *** | 0.43 (0.31–0.59) | *** | 0.50 (0.37–0.68) | *** | 0.58 (0.43–0.78) | *** |
| Age × childhood socioeconomic circumstances (ref. Most disadvantaged) | ||||||||||
| Age × disadvantaged | 0.98 (0.96–0.99) | ** | 0.98 (0.96–0.99) | ** | 0.98 (0.96–0.99) | ** | 0.97 (0.96–0.99) | *** | 0.98 (0.96–0.99) | ** |
| Age × middle | 0.97 (0.96–0.98) | *** | 0.97 (0.95–0.98) | *** | 0.97 (0.95–0.98) | *** | 0.96 (0.95–0.98) | *** | 0.96 (0.95–0.98) | *** |
| Age × advantaged | 0.97 (0.95–0.98) | *** | 0.96 (0.95–0.98) | *** | 0.96 (0.95–0.98) | *** | 0.96 (0.94–0.98) | *** | 0.96 (0.95–0.98) | *** |
| Age × most advantaged | 0.97 (0.94–0.99) | ** | 0.96 (0.94–0.99) | ** | 0.96 (0.94–0.99) | ** | 0.96 (0.93–0.98) | ** | 0.97 (0.94–0.99) | ** |
| Education (ref. Primary) | ||||||||||
| Secondary | – | 0.75 (0.65–0.85) | *** | 0.77 (0.68–0.88) | *** | 0.81 (0.71–0.92) | ** | 0.84 (0.74–0.95) | ** | |
| Tertiary | – | 0.45 (0.36–0.56) | *** | 0.56 (0.44–0.72) | *** | 0.64 (0.50–0.81) | *** | 0.70 (0.56–0.88) | ** | |
| Age × education (ref. Primary) | ||||||||||
| Age × secondary | – | 1.01 (1.00–1.02) | 1.01 (1.00–1.02) | 1.00 (0.99–1.02) | 1.00 (0.99–1.01) | |||||
| Age × tertiary | – | 1.01 (0.99–1.03) | 1.01 (0.99–1.03) | 1.01 (0.99–1.03) | 1.01 (0.99–1.02) | |||||
| Main occupational position (ref. High-skilled) | ||||||||||
| Low-skilled | – | – | 1.50 (1.23–1.85) | *** | 1.39 (1.14–1.70) | ** | 1.40 (1.15–1.70) | ** | ||
| Never worked | – | – | 1.99 (1.56–2.55) | *** | 1.75 (1.37–2.24) | *** | 1.79 (1.42–2.27) | *** | ||
| Age × main occupational position (ref. High-skilled) | ||||||||||
| Age × low-skilled | – | – | 1.00 (0.98–1.02) | 1.00 (0.98–1.02) | 1.00 (0.99–1.02) | |||||
| Age × never worked | – | – | 1.01 (0.99–1.03) | 1.01 (0.99–1.03) | 1.01 (0.99–1.03) | |||||
| Ability to make ends meet (ref. Easily) | ||||||||||
| Fairly easily | – | – | – | 1.35 (1.18–1.56) | *** | 1.23 (1.07–1.41) | ** | |||
| With some difficulty | – | – | – | 2.00 (1.70–2.35) | *** | 1.62 (1.38–1.89) | *** | |||
| With great difficulty | – | – | – | 4.43 (3.61–5.44) | *** | 3.12 (2.56–3.81) | *** | |||
| Age × ability to make ends meet (ref. Age × easily) | ||||||||||
| Age × fairly easily | – | – | – | 1.00 (0.99–1.02) | 1.00 (0.99–1.01) | |||||
| Age × with some difficulty | – | – | – | 1.00 (0.99–1.01) | 1.00 (0.99–1.01) | |||||
| Age × with great difficulty | – | – | – | 1.00 (0.98–1.01) | 1.00 (0.98–1.01) | |||||
| Men | ||||||||||
| Age | 1.23 (1.21–1.26) | *** | 1.22 (1.20–1.25) | *** | 1.25 (1.21–1.29) | *** | 1.25 (1.21–1.29) | *** | 1.24 (1.20–1.28) | *** |
| Childhood socioeconomic circumstances (ref. Most disadvantaged) | ||||||||||
| Disadvantaged | 0.81 (0.63–1.04) | 0.88 (0.69–1.13) | 0.95 (0.75–1.22) | 0.96 (0.76–1.22) | 0.99 (0.80–1.23) | |||||
| Middle | 0.53 (0.41–0.69) | *** | 0.64 (0.49–0.84) | ** | 0.70 (0.54–0.91) | ** | 0.76 (0.59–0.97) | * | 0.80 (0.64–1.01) | |
| Advantaged | 0.46 (0.34–0.63) | *** | 0.62 (0.45–0.85) | ** | 0.69 (0.50–0.94) | * | 0.75 (0.55–1.02) | 0.78 (0.58–1.03) | ||
| Most advantaged | 0.45 (0.29–0.70) | *** | 0.69 (0.44–1.08) | 0.81 (0.52–1.27) | 0.85 (0.55–1.31) | 0.86 (0.57–1.29) | ||||
| Age × childhood socioeconomic circumstances (ref. Most disadvantaged) | ||||||||||
| Age × disadvantaged | 1.00 (0.97–1.02) | 1.00 (0.98–1.02) | 0.99 (0.97–1.02) | 0.99 (0.97–1.01) | 1.00 (0.98–1.02) | |||||
| Age × middle | 0.98 (0.96–1.00) | 0.98 (0.96–1.01) | 0.98 (0.95–1.00) | 0.98 (0.95–1.00) | 0.98 (0.96–1.00) | |||||
| Age × advantaged | 0.98 (0.96–1.01) | 0.98 (0.96–1.01) | 0.97 (0.95–1.00) | 0.97 (0.95–1.00) | 0.98 (0.95–1.00) | |||||
| Age × most advantaged | 1.02 (0.98–1.06) | 1.02 (0.98–1.06) | 1.01 (0.97–1.05) | 1.00 (0.96–1.05) | 1.00 (0.96–1.04) | |||||
| Education (ref. Primary) | ||||||||||
| Secondary | – | 0.64 (0.52–0.80) | *** | 0.66 (0.53–0.82) | *** | 0.73 (0.60–0.90) | ** | 0.77 (0.64–0.94) | ** | |
| Tertiary | – | 0.40 (0.30–0.53) | *** | 0.46 (0.34–0.62) | *** | 0.59 (0.44–0.79) | *** | 0.64 (0.48–0.84) | ** | |
| Age × education (ref. Primary) | ||||||||||
| Age × secondary | – | 1.01 (0.99–1.03) | 1.01 (0.99–1.03) | 1.01 (0.99–1.03) | 1.01 (0.99–1.02) | |||||
| Age × tertiary | – | 1.01 (0.98–1.04) | 1.00 (0.97–1.03) | 1.00 (0.97–1.03) | 1.00 (0.97–1.03) | |||||
| Main occupational position (ref. High-skilled) | ||||||||||
| Low-skilled | – | – | 1.37 (1.10–1.70) | ** | 1.27 (1.02–1.57) | * | 1.22 (1.00–1.49) | * | ||
| Never worked | – | – | 4.45 (1.85–10.71) | ** | 3.79 (1.64–8.73) | ** | 2.73 (1.27–5.84) | * | ||
| Age × main occupational position (ref. High-skilled) | ||||||||||
| Age × low-skilled | – | – | 0.98 (0.96–1.00) | * | 0.98 (0.97–1.00) | 0.98 (0.97–1.00) | ||||
| Age × never worked | – | – | 0.94 (0.87–1.00) | 0.94 (0.87–1.00) | 0.92 (0.87–0.98) | * | ||||
| Ability to make ends meet (ref. Easily) | ||||||||||
| Fairly easily | – | – | – | 1.62 (1.32–1.99) | *** | 1.52 (1.26–1.84) | *** | |||
| With some difficulty | – | – | – | 2.90 (2.26–3.72) | *** | 2.46 (1.96–3.10) | *** | |||
| With great difficulty | – | – | – | 5.49 (3.94–7.65) | *** | 4.51 (3.32–6.12) | *** | |||
| Age × ability to make ends meet (ref. Age × easily) | ||||||||||
| Age × fairly easily | – | – | – | 1.00 (0.98–1.02) | 0.99 (0.97–1.01) | |||||
| Age × with some difficulty | – | – | – | 1.00 (0.98–1.02) | 0.99 (0.97–1.01) | |||||
| Age × with great difficulty | – | – | – | 0.99 (0.96–1.02) | 0.99 (0.96–1.01) |
Abbreviations: IADL, instrumental activity of daily living; OR, odds ratio; CI, confidence interval.
Model 1: adjusted for confounders and controls: birth cohorts, countries, attrition and living with biological parents.
Model 2: model 1 adjusted for young adulthood socioeconomic circumstances (education).
Model 3: model 2 adjusted for adulthood socioeconomic circumstances (main occupation).
Model 4: model 3 adjusted for old age socioeconomic circumstances (current ability to make ends meet with the household income).
Model 5: model 4 adjusted for covariates: partnership status, number of chronic diseases and body mass index.
P-values:
P<0.001,
P<0.01,
P<0.05.
Among women, a gradient across strata of CSC was observable across all 5 models, suggesting that the higher the socioeconomic advantage in childhood, the lower the risk of IADL limitations in older age. This gradient was barely mediated when including education, main occupational position and ability to make ends meet. All three adult socioeconomic indicators remained associated with the risk of IADL limitations in the fully adjusted model (Model 5).
Among men, three strata of CSC (middle, advantaged and most advantaged) were associated with a lower risk of IADL limitations at older age. These associations were slightly attenuated when adjusting for education (Model 2), particularly for the most advantaged group, and more strongly when adjusting for main occupation (Model 3) and ability to make ends meet (Model 4). Men with a low-skilled main occupational position or who never worked, as well as those reporting difficulties to make ends meet, were at a higher risk of IADL limitations, whereas the risk of IADL limitations was reduced for men with a secondary or tertiary degree (Models 4 and 5).
Sensitivity analyses
Adjusting the final model (Model 5) for attrition by death (Supplementary table S1) and by dropout (Supplementary table S2) yielded similar results for ADL and IADL limitations. Similar changes to the fully adjusted model were observed when excluding respondents with visual impairment (Supplementary table S3) and when including an index of health behaviours (Supplementary table S4). Among women, the main occupational position was no longer associated with the risk of ADL limitations. While the ability to make ends meet remained significantly associated, the association of CSC and ADL was only preserved when adjusting for health behaviours (Supplementary table S4). For men, the association of education and IADL was attenuated in both sensitivity analyses (Supplementary tables S3 and S4). When excluding men with visual impairment, the risk of IADL limitations became associated with middle and advantaged CSC, but the association of the main occupational position was no longer significant.
Discussion
Main findings
Overall, ADL and IADL limitations increased slowly over the process of ageing, among all participants, but the risk for ADL and IADL limitations differed by gender.
For women, results showed a clear social gradient: more advantaged CSC was associated with a lower risk of ADL and IADL limitations in later life. These associations were attenuated when including adult socioeconomic indicators and covariates but remained significant. Differences in the risk of ADL and IADL limitations by CSC increased with the process of ageing.
For men, disadvantaged CSC was associated with greater risk IADL limitations in later life. This association was mediated by education, the main occupational position and the ability to make ends meet. No association between CSC and risk of ADL limitations was observed.
Results did not change when adjusting the final model for attrition by death and by dropout. Including health behaviours yielded similar results, except for women whose main occupational position was no longer associated with ADL. When excluding women with visual impairment, only the association of the ability to make ends meet and ADL remained significant. When excluding men with visual impairment, IADL became associated with middle and advantaged CSC, but main occupational position was no longer significant and the association with education disappeared.
Lastly, among women, the significant growth over ageing of the risk of ADL and IADL limitations by CSC groups suggests that poor socioeconomic circumstances in childhood are related to an overall risk of disability in old age as well as to disability trajectories of individuals in the ageing process. For men however, we cannot find evidence of growing differences over ageing in the risk of IADL limitations by CSC groups. It means that we observed an association of CSC on the overall level of IADL, but not on ADL, and not on IADL trajectories.
Comparison with previous studies
In this study, at baseline, women had a higher probability of experiencing disability in old age than men in Europe, which confirms previous findings.2,11,35 Cohort variations in disability trajectories29 can influence gender differences. Although this difference in gender still needs clarification,36 our analysis showed that women’s adult socioeconomic attainments cannot compensate for disadvantaged CSC. However, this compensation was observed among men, whose high-skilled occupation and ability to make ends meet lowered the risk of IADL limitations by CSC.
This result is novel to the field and questions previous studies indicating only a partial mediation by adult socioeconomic indicators, relying on analyses not stratified by gender.12,13 The observed mediating effect among men was robust to sensitivity analyses including health behaviours, which is in line with previous findings.37 A first explanation for these findings is the cumulative advantage and disadvantage hypothesis:38 when having a bad start in life, women seemed to be at higher risk of accumulating disadvantaged socioeconomic opportunities over the life course, while men were not, because of gender inequalities for these respondents who lived a great part of their life in the 20th century (i.e. respondents were born between 1919 and 1954), and this despite better access to education for younger cohorts.14 Given the social norms (responsibility for the family and restricted access to education), social mobility—via education and occupation—was limited and therefore mainly driven by changes at the household level.
Strengths and limitations
The sample size of 24 440 participants was sufficient to draw robust conclusions. To reduce health selection bias, we included the maximum number of respondents, i.e. those who completed the retrospective module (wave 3) and at least one of the other five waves (waves 1, 2, 4, 5 and 6). Mixed-effects models allowed using incomplete and unbalanced data. The stratification by gender provided a more detailed picture about the differences in disability trajectories according to CSC and the uncertain role of adult socioeconomic indicators. However, this study has six limitations. First, self-reported retrospective information on childhood and adult socioeconomic indicators was used, which may be subject to recall bias. Nevertheless, the reliability of recall of simple measures of socioeconomic circumstances in a survey of older adults is good.39 Second, cross-national comparison on the prevalence of ADL and IADL is known to be limited by contextual, social and cultural differences across countries,40 for which we controlled by adjusting for country of residence. Third, we did not have information on all disability covariates, i.e. cognitive impairment. However, this lack of information should not bias our results since cognitive impairment is probably not a confounder of the association between CSC and ADL. Fourth, a limited number of SHARE participants answered the SHARELIFE module in wave 3, resulting in a significant reduction of the analytical sample. This may cause information bias on the exposure. Fifth, attrition is a significant issue in longitudinal studies with old age participants. To limit this bias, all models were controlled for mediator-outcome confounding. Such a strategy did not solve the issue of missing data due to attrition. Sixth, 2% of women and 1.8% of men in our sample grew up without their biological parents, raising the question of the relevance of the CSC score. For respondents who were adopted by their family (grand-parents, uncle or aunt) or by a host family, we consider that the score may apply to them. However, for respondents who grew up in a children’s home (N = 308), the score of CSC may be less reliable.
Conclusion
The more disadvantaged the socioeconomic circumstances during childhood are, the higher the risk of disability in old age. Men and women are not able to compensate for an unfavourable start in life equally well. For men, the association between CSC and disability is mediated by adult socioeconomic indicators, such as their main occupation during active life and their ability to make ends meet with their household income in older age. Conversely, for women, this association is only partially mediated. Our findings suggest that men born in the first half of the 20th century are more likely to make up for disadvantaged CSC than women over the life course, resulting in less disabling conditions at old age.
Supplementary Material
Acknowledgements
The authors are grateful to the Swiss National Science Foundation for its financial assistance. This paper uses data from SHARE Waves 1, 2, 3 (SHARELIFE), 4, 5 and 6 (DOIs: 10.6103/SHARE.w1.600, 10.6103/SHARE.w2.600, 10.6103/SHARE.w3.600, 10.6103/SHARE.w4.600, 10.6103/SHARE.w5.600, 10.6103/SHARE.w6.600), the SHARE data collection was primarily funded by the European Commission through FP5 (QLK6-CT-2001-00360), FP6 (SHARE-I3: RII-CT-2006-062193, COMPARE: CIT5-CT-2005-028857, SHARELIFE: CIT4-CT-2006-028812) and FP7 (SHARE-PREP: N°211909, SHARE-LEAP: N°227822, SHARE M4: N°261982). The authors gratefully acknowledge additional funding from the German Ministry of Education and Research, the Max Planck Society for the Advancement of Science, the U.S. National Institute on Aging (U01_AG09740-13S2, P01_AG005842, P01_AG08291, P30_AG12815, R21_AG025169, Y1-AG-4553-01, IAG_BSR06-11, OGHA_04-064, HHSN271201300071C) and various national funding sources (see www.share-project.org).
Funding
This publication benefited from the support of the Swiss National Centre of Competence in Research ‘LIVES – Overcoming vulnerability: Life course perspectives’, which is financed by the Swiss National Science Foundation (grant number: 51NF40-160590). Bernadette W. A. van der Linden is supported by the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No. 676060. Matthieu P. Boisgontier is supported by the Research Foundation—Flanders (FWO).
Conflicts of interest: None declared.
Key points
The more advantaged the socioeconomic circumstances during childhood are, the lower the risk for disability in older age.
Women, in contrast to men, could not compensate their risk for disability associated with childhood socioeconomic circumstances by adult socioeconomic attainments.
Findings underline childhood as a sensitive period of the life course to which public health policies should pay attention.
References
- 1. Wahrendorf M, Reinhardt JD, Siegrist J. Relationships of disability with age among adults aged 50 to 85: evidence from the United States, England and continental Europe. PLoS One 2013;8:e71893. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Crimmins EM, Kim JK, Solé-Auró A. Gender differences in health: results from SHARE, ELSA and HRS. Eur J Public Health 2011;21:81–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Balzi D, Lauretani F, Barchielli A, et al. Risk factors for disability in older persons over 3-year follow-up. Age Ageing 2010;39:92–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Millan-Calenti JC, Tubío J, Pita-Fernández 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;50:306–10. [DOI] [PubMed] [Google Scholar]
- 5. Stuck AE, Walthert JM, Nikolaus T, et al. Risk factors for functional status decline in community-living elderly people: a systematic literature review. Soc Sci Med 1999;48:445–69. [DOI] [PubMed] [Google Scholar]
- 6. Matthews RJ, Smith LK, Hancock RM, et al. Socioeconomic factors associated with the onset of disability in older age: a longitudinal study of people aged 75 years and over. Soc Sci Med 2005;61:1567–75. [DOI] [PubMed] [Google Scholar]
- 7. Foverskov E, Holm A. Socioeconomic inequality in health in the British household panel: tests of the social causation, health selection and the indirect selection hypothesis using dynamic fixed effects panel models. Soc Sci Med 2016;150:172–83. [DOI] [PubMed] [Google Scholar]
- 8. Zhong Y, Wang J, Nicholas S. Gender, childhood and adult socioeconomic inequalities in functional disability among Chinese older adults. Int J Equity Health 2017;16:165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Fujiwara T, Kondo K, Shirai K, et al. Associations of childhood socioeconomic status and adulthood height with functional limitations among Japanese older people: results from the JAGES 2010 Project. J Gerontol A Biol Sci Med Sci 2014;69:852–9. [DOI] [PubMed] [Google Scholar]
- 10. Freedman VA, Martin LG, Schoeni RF, Cornman JC. Declines in late-life disability: the role of early-and mid-life factors. Soc Sci Med 2008;66:1588–602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Turrell G, Lynch JW, Leite C, et al. Socioeconomic disadvantage in childhood and across the life course and all-cause mortality and physical function in adulthood: evidence from the Alameda County Study. J Epidemiol Community Health 2007;61:723–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Bowen ME, González HM. Childhood socioeconomic position and disability in later life: results of the health and retirement study. Am J Public Health 2010;100:S197–203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Haas S. Trajectories of functional health: the ‘long arm’ of childhood health and socioeconomic factors. Soc Sci Med 2008;66:849–61. [DOI] [PubMed] [Google Scholar]
- 14. Breen R, Luijkx R, Müller W, Pollak R. Long-term Trends in Educational Inequality in Europe: class Inequalities and Gender Differences. Eur Sociol Rev 2010;26:31–48. [Google Scholar]
- 15. Domenico DM, Jones KH. Career aspirations of women in the 20th century. JCTE 2007;22:27. [Google Scholar]
- 16. Börsch-Supan A, Brandt M, Hunkler C, et al. Data resource profile: the survey of health, ageing and retirement in Europe (SHARE). Int J Epidemiol 2013;42:992–1001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Brugiavini A, Cavapozzi D, Pasini G, Trevisan E. Working life histories from SHARELIFE: a retrospective panel. SHARE WP Series 2013;11–3. [Google Scholar]
- 18. Katz S, Ford AB, Moskowitz RW, et al. Studies of illness in the aged: the index of ADL: a standardized measure of biological and psychosocial function. JAMA 1963;185:914–9. [DOI] [PubMed] [Google Scholar]
- 19. Taylor R, Conway L, Calderwood L, et al. Health, wealth and lifestyles of the older population in England: the 2002 English Longitudinal Study of Ageing. Technical Report. National Center for Social Research, 2007. Report No.: P2058.
- 20. Lawton MP, Brody EM. Assessment of older people: self-maintaining and instrumental activities of daily living. Gerontologist 1969;9: 179–86. [PubMed] [Google Scholar]
- 21. Chittleborough CR, Baum F, Taylor A, et al. J Epidemiol Community Health 2006;60:981–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Dedman D, Gunnell D, Smith GD, Frankel S. Childhood housing conditions and later mortality in the Boyd Orr cohort. J Epidemiol Community Health 2001;55:10–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Evans MD, Kelley J, Sikora J, Treiman DJ. Family scholarly culture and educational success: books and schooling in 27 nations. Res Soc Stratif Mobil 2010;28:171–97. [Google Scholar]
- 24. Marsh A, Gordon D. Home Sweet Home? The Impact of Poor Housing on Health. Bristol: Policy Press, 1999. [Google Scholar]
- 25. Wahrendorf M, Blane D. Does labour market disadvantage help to explain why childhood circumstances are related to quality of life at older ages? Results from SHARE. Aging Ment Health 2015;19:584–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Wahrendorf M, Blane D, Bartley M, et al. Working conditions in mid-life and mental health in older ages. Adv Life Course Res 2013;18:16–25. [DOI] [PubMed] [Google Scholar]
- 27. International Labour Office ILO. International Standard Classification of Occupations. ISCO-08 Geneva, 2012. Available at: http://www.ilo.org/public/english/bureau/stat/isco/.
- 28. Eikemo TA, Bambra C, Judge K, Ringdal K. Welfare state regimes and differences in self-perceived health in Europe: a multilevel analysis. Soc Sci Med 2008;66:2281–95. [DOI] [PubMed] [Google Scholar]
- 29. Reynolds SL, Crimmins EM, Saito Y. Cohort differences in disability and disease presence. Gerontologist 1998;38:578–90. [DOI] [PubMed] [Google Scholar]
- 30. Yang Y, Lee LC. Sex and race disparities in health: cohort variations in life course patterns. Soc Forces 2009;87:2093–124. [Google Scholar]
- 31. Amemiya A, Fujiwara T, Murayama H, et al. Adverse childhood experiences and higher-level functional limitations among older Japanese people: results from the JAGES study. J Gerontol A Biol Sci Med Sci 2018;73:261–6. [DOI] [PubMed] [Google Scholar]
- 32. Bierman A. Functional limitations and psychological distress: marital status as moderator. Soc Ment Health 2012;2:35–52. [Google Scholar]
- 33. Yip JLY, Khawaja AP, Broadway D, et al. Visual acuity, self-reported vision and falls in the EPIC-Norfolk Eye study. Br J Ophthalmol 2014;98:377–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. LaCroix AZ, Guralnik JM, Berkman LF, et al. Maintaining mobility in late life. II. Smoking, alcohol consumption, physical activity, and body mass index. Am J Epidemiol 1993; 37:858–69. [DOI] [PubMed] [Google Scholar]
- 35. Murtagh KN, Hubert HB. Gender differences in physical disability among an elderly cohort. Am J Public Health 2004;94:1406–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Friedmann JM, Elasy T, Jensen GL. The relationship between body mass index and self-reported functional limitation among older adults: a gender difference. J Am Geriatr Soc 2001;49:398–403. [DOI] [PubMed] [Google Scholar]
- 37. Van de Mheen H, Stronks K, Looman C, Mackenbach J. Does childhood socioeconomic status influence adult health through behavioural factors? Int J Epidemiol 1998;27:431–7. [DOI] [PubMed] [Google Scholar]
- 38. Dannefer D. Cumulative advantage/disadvantage and the life course: cross-fertilizing age and social science theory. J Gerontol B Psychol Sci Soc Sci 2003;58:S327–37. [DOI] [PubMed] [Google Scholar]
- 39. Lacey RJ, Belcher J, Croft PR. Validity of two simple measures for estimating life-course socio-economic position in cross-sectional postal survey data in an older population: results from the North Staffordshire Osteoarthritis Project (NorStOP). BMC Med Res Methodol 2012;12:88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Chan KS, Kasper JD, Brandt J, Pezzin LE. Measurement equivalence in ADL and IADL difficulty across International Surveys of Aging: findings from the HRS, SHARE, and ELSA. J Gerontol B Psychol Sci Soc Sci 2012;67B:121–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.