Abstract
Background.
We examined physical functioning (PF) trajectories (maintaining, slowly declining, and rapidly declining) spanning 15 years in older women aged 65–80 and protective factors that predicted better current levels and less decline in functional independence outcomes after age 80.
Methods.
Women’s Health Initiative extension participants who met criteria (enrolled in either the clinical trial or observational study cohort, >80 years at the data release cutoff, PF survey data from initial enrollment to age 80, and functional independence survey data after age 80) were included in these analyses (mean [ SD ] age = 84.0 [1.4] years; N = 10,478). PF was measured with the SF-36 (mean = 4.9 occasions). Functional independence was measured by self-reported level of dependence in basic and instrumental activities of daily living (ADLs and IADLs) (mean = 3.4 and 3.3 occasions).
Results.
Maintaining consistent PF in older adulthood extends functional independence in ADL and IADL in late-life. Protective factors shared by ADL and IADL include maintaining PF over time, self-reported excellent or very good health, no history of hip fracture after age 55, and no history of cardiovascular disease. Better IADL function is uniquely predicted by a body mass index less than 25 and no depression. Less ADL and IADL decline is predicted by better self-reported health, and less IADL decline is uniquely predicted by having no history of hip fracture after age 55.
Conclusions.
Maintaining or improving PF and preventing injury and disease in older adulthood (ages 65–80) has far-reaching implications for improving late-life (after age 80) functional independence.
Keywords: ADL, Aging, IADL, Physical function, Successful aging
Almost 40% of community-dwelling older adults currently experience some type of functional limitation or disability, and an increasing number of middle-aged and older adults report difficulty with physical function ( 1 ). Overall disability prevalence rises sharply with age, averaging 25% among U.S. adults aged 45–64 and 35% among adults aged 65 years and older ( 2 ). The increase in the aging population’s disability burden is a strong predictor of nursing home admissions and increased expenses for long-term care services. Given the aging demographic of the U.S. population, that the majority of older adults (>75 years) are women, and that incident disability in this age and gender group is higher, women are living longer with functional limitations that decrease their independence ( 3 , 4 ). As a consequence, difficulty performing basic and instrumental activities of daily living (ADLs and IADLs) may increase the need for personal assistance or relocation to a residential care facility ( 5–7 ), with multiple adverse effects on health, safety, and mortality ( 8 , 9 ).
The ability to perform ADLs and IADLs is one measure of age-related disability ( 10 ). Basic ADLs such as bathing and dressing are self-care activities frequently associated with functional mobility, whereas IADLs such as managing finances and preparing a meal represent complex everyday tasks that can be delegated to others and are thought to rely on cognitive abilities ( 11 ). Physical functioning (PF) ( 12 , 13 ) is a construct that measures health-related mobility limitations in typical daily activities. Neuromusculoskeletal and movement-related functions are assumed by disability frameworks such as the International Classification of Impairments, Diseases, and Handicaps (ICIDH) ( 14 ) and the American Occupational Therapy Association (AOTA) Framework ( 15 ) to be a logical prerequisite of the ability to perform both ADLs and IADLs. Although a few longitudinal studies have shown that trajectories of change in constructs such as physical activity or physical function are associated with outcomes such as ADL ( 16–18 ), mortality ( 19 , 20 ), and falls ( 21 , 22 ), and that there is heterogeneity in longitudinal trajectories of physical function in older adults ( 23 ), existing longitudinal studies have not emphasized daily life functional independence outcomes. The majority of prior studies on related topics have been limited by small sample sizes, cross-sectional designs, non-community-dwelling samples, and cohorts with a single clinical diagnosis. Given the projected substantial increase in the number of women aged 65 and older in the coming decades, further understanding characteristics associated with preserved functional independence has important implications for successful aging and for public health.
In this study, we examined the influence of longitudinal trajectories of PF on functional independence outcomes in community-dwelling women who are currently older than 80 years of age, beginning at age 65, in order to identify those who consistently maintained or exhibited only small declines in their PF over time. In-line with theories postulating that leading a healthy lifestyle and being actively engaged can extend functional independence into late-life, and by extension that lifestyle changes made in older adulthood can maintain or improve daily life function even in late-life ( 24–26 ), we tested the hypothesis that women with a history of consistently maintaining their PF from age 65 to age 80 would have fewer limitations in ADL and IADL after age 80, and thus less disability burden. In addition, we also aimed to determine which factors, based on the literature, were most protective of women’s late-life functional independence.
Methods
Participants
The recruitment, enrollment, and primary objectives of the Women’s Health Initiative (WHI) Study are described in detail in this issue. A woman was included in the 80+ analytic sample if she was: (a) enrolled in either the WHI clinical trial or observational study cohort, (b) was an active participant in the WHI second extension, (c) was 80 years of age or greater at the time of the September 17, 2012 data release cutoff, (d) had PF data collected from initial enrollment to the most recent assessment before age 80, and (e) had functional independence (ADL and IADL) survey data collected after age 80. For the purposes of trajectory analyses, only the women with at least three measurements of PF before age 80 and at least three measurements of ADL and IADL after age 80 were included.
Measures
Questionnaires were administered to participants at baseline that included the SF-36 (PF subscale), a well-validated measure of self-reported physical function ( 12 , 27 ), as well as self-reported level of functional independence derived from four items that assessed basic ADLs and two items that assessed more complex IADLs. Values on the PF subscale ranged from 0 to 100 with a higher score indicating higher self-reported performance. Four items describing basic ADL (eg, “Can you … feed yourself, dress and undress yourself, get in and out of bed yourself, and take a bath or shower yourself.”) each of which had three possible values (1 = By myself without help, 2 = With some help, 3 = Completely unable to do this by myself) were summed for each participant. Two items describing IADL (eg, “Can you … do own grocery shopping, keep track of and take your own medications.”) each of which had three possible values (1 = By myself without help, 2 = With some help, 3 = Completely unable to do this by myself) were summed for each participant. A higher score indicates more functional limitation. Data were missing if any of the items were missing (<5% at each measurement occasion for ADL and IADL).
Participant characteristics included the most recent measurements of: age at visit with PF; age at baseline; education; race/ethnicity; income less than $35,000; body mass index (BMI; baseline and most recent) (kg/m 2 ); self-reported health; history of hip fracture at age 55 and above; history of cardiovascular disease (CVD) defined as having been diagnosed with coronary heart disease, adjudicated stroke (ischemic, hemorrhagic), venous thromboembolism, congestive heart failure, and peripheral artery disease among women aged 80 and older in the WHI cohort; smoking status; alcohol intake; depression; last PF measurement before turning 80 years of age; and PF at the most recent assessment. Depression was assessed at baseline using the Burnam 8-item depression screening instrument ( 28 ). It comprises six items from the Center for Epidemiologic Studies Depression Scale (CES-D) about the frequency of depressive symptoms during the past week and two items from the National Institute of Mental Health’s Diagnostic Interview Schedule (DIS) about the duration of symptoms. The scoring algorithm, using a prediction equation developed by Burnam and coworkers, gives a composite score between 0 and 1 which represents the probability of having depression, using a standard cut point of more than or equal to 0.06 to dichotomize the scale. Note in addition that severe depression at baseline was an exclusion for the WHI observational study and clinical trials.
Statistical Analyses
Individual trajectories of PF before turning 80 years of age were fitted using linear regression (regressing PF against time when the measure was taken). Thus, each trajectory can be summarized as the intercept and slope of the fitted regression line. We expected these developmental trajectories to be linear and an overall test showed the quadratic term to be nonsignificant ( p > .10). Similarly, individual trajectories of ADL and IADL were also summarized as intercept and slope. K -means clustering analysis was used to classify PF trajectories into different classes. We predefined the maximum number of clusters as four based on relevant literature ( 20 ), and three clusters provided the best fit to the data based on the results and prior knowledge. Different classes of PF trajectories were then used as predictors in modeling ADL and IADL outcomes. For ADL and IADL, we used two measures: the most recent measurement after age 80 and slope (change rate) after age 80. Analysis of covariance was used to adjust the models. Both unadjusted (only including PF classes) and adjusted models (also including all the proposed covariates) were used to evaluate differences in functional independence between different PF classes. Least squares means and corresponding SE s were calculated from the models for each class. Since we were also interested in other predictors of function independence, least squares mean was also calculated for each covariate. All analyses were performed using SAS 9.3 (SAS Institute, Cary, NC). All tests were two-sided and significance levels were set at .05.
Results
PF Trajectories
The PF trajectories for each participant aged 80 years or older spanned the time from their initial enrollment extending back from age 65 to their most recent assessment. The mean number of PF assessments used for the trajectory analysis was 4.9. The mean number of assessments used for the slope decrease in functional independence outcomes after age 80 was 3.4 for ADL and 3.3 for IADL. Three PF trajectory groups were identified: (a) maintaining or increasing physical function, (b) slowly declining physical function, and (c) rapidly declining physical function (see Figure 1).
Table 1 displays the baseline characteristics of the women in our analytic sample ( N = 10,478), by PF trajectory group. Women with slowly declining physical function comprised almost half of the total sample ( N = 5,148; 49.0%), followed by maintaining or increasing physical function ( N = 3,341; 32.0%), and rapidly declining physical function ( N = 1,989; 19%).
Table 1.
Characteristics of WHI Women Older Than 80 by PF Trajectory Group
| Characteristics | Category | Total N (%) | Rapidly Declining PF | Maintaining PF | Slowly Declining PF | p Value |
|---|---|---|---|---|---|---|
| 10,478 (100%) | 1,989 (19%) | 3,341 (32%) | 5,148 (49%) | |||
| Age at baseline | Mean ( SD ) | 68.5 (1.7) | 68.8 (1.8) | 68.5 (1.7) | 68.3 (1.7) | <.0001 |
| Age at visit with PF | Mean ( SD ) | 84.0 (1.4) | 84.3 (1.5) | 84.0 (1.4) | 83.9 (1.4) | <.0001 |
| Age at visit with PF | 80–84 | 7,723 (73.7) | 1,352 (68.0) | 2,471 (74.0) | 3,900 (75.8) | <.0001 |
| 85–89 | 2,724 (26.0) | 626 (31.5) | 858 (25.7) | 1,240 (24.1) | — | |
| 90+ | 31 (0.3) | 11 (0.6) | 12 (0.4) | 8 (0.2) | — | |
| >HS education | Yes | 8,312 (79.3) | 1,585 (79.7) | 2,635 (78.9) | 4,092 (79.5) | — |
| Race/ethnicity | American Indian or Alaskan native | 33 (0.3) | 7 (0.4) | 13 (0.4) | 13 (0.3) | .0393 |
| Asian or Pacific Islander | 203 (1.9) | 41 (2.1) | 82 (2.5) | 80 (1.6) | — | |
| Black or African American | 431 (4.1) | 82 (4.1) | 159 (4.8) | 190 (3.7) | — | |
| Hispanic/Latino | 166 (1.6) | 31 (1.6) | 59 (1.8) | 76 (1.5) | — | |
| White (not of Hispanic origin) | 9,534 (91.0) | 1,808 (90.9) | 2,988 (89.4) | 4,738 (92.0) | — | |
| Other | 111 (1.1) | 20 (1.0) | 40 (1.2) | 51 (1.0) | — | |
| Income > $35,000 | Yes | 5,790 (55.3) | 990 (49.8) | 1,913 (57.3) | 2,887 (56.1) | — |
| Most recent BMI | Mean ( SD ) | 26.9 (5.2) | 28.7 (5.7) | 25.9 (4.7) | 26.8 (5.2) | <.0001 |
| BMI (kg/m 2 ) | Less than 25 | 4,292 (41.0) | 513 (25.8) | 1,639 (49.1) | 2,140 (41.6) | <.0001 |
| 25–29 | 3,783 (36.1) | 754 (37.9) | 1,145 (34.3) | 1,884 (36.6) | — | |
| At least 30 | 2,403 (22.9) | 722 (36.3) | 557 (16.7) | 1,124 (21.8) | — | |
| Self-reported health | Excellent | 787 (7.5) | 52 (2.6) | 394 (11.8) | 341 (6.6) | <.0001 |
| Very good | 3,926 (37.5) | 436 (21.9) | 1,504 (45.0) | 1,986 (38.6) | — | |
| Good | 4,348 (41.5) | 958 (48.2) | 1,190 (35.6) | 2,200 (42.7) | — | |
| Fair | 1,294 (12.3) | 495 (24.9) | 235 (7.0) | 564 (11.0) | — | |
| Poor | 123 (1.2) | 48 (2.4) | 18 (0.5) | 57 (1.1) | — | |
| History of hip fracture ≥ age 55 | No | 9,939 (94.9) | 1,822 (91.6) | 3,202 (95.8) | 4,915 (95.5) | <.0001 |
| Yes | 539 (5.1) | 167 (8.4) | 139 (4.2) | 233 (4.5) | — | |
| Hx CVD | No | 8,247 (78.7) | 1,373 (69.0) | 2,793 (83.6) | 4,081 (79.3) | <.0001 |
| Yes | 2,231 (21.3) | 616 (31.0) | 548 (16.4) | 1,067 (20.7) | — | |
| Smoking status | Never | 5,689 (54.3) | 1,059 (53.2) | 1,867 (55.9) | 2,763 (53.7) | .1393 |
| Past | 4,661 (44.5) | 900 (45.2) | 1,440 (43.1) | 2,321 (45.1) | — | |
| Current | 128 (1.2) | 30 (1.5) | 34 (1.0) | 64 (1.2) | — | |
| Alcohol use (drinks/d) | None | 3,324 (31.7) | 772 (38.8) | 995 (29.8) | 1,557 (30.2) | <.0001 |
| <1 | 6,641 (63.4) | 1,131 (56.9) | 2,197 (65.8) | 3,313 (64.4) | — | |
| 1–2 | 475 (4.5) | 82 (4.1) | 139 (4.2) | 254 (4.9) | — | |
| ≥3 | 38 (0.4) | 4 (0.2) | 10 (0.3) | 24 (0.5) | — | |
| Depression | Yes | 714 (6.8) | 184 (9.3) | 206 (6.2) | 324 (6.3) | — |
| Most recent PF | Mean ( SD ) | 58.6 (26.8) | 37.1 (22.2) | 70.4 (24.1) | 59.3 (24.9) | <.0001 |
| PF prior to age 80 | Mean ( SD ) | 71.5 (23.7) | 42.4 (19.4) | 86.3 (15.3) | 73.1 (19.3) | <.0001 |
Note: PF is based on SF-36 scores ranging from 0 to 100. BMI = body mass index; CVD = cardiovascular disease; HS = higher secondary; PF = physical functioning; WHI = Women’s Health Initiative Study.
Population Characteristics of Women Who Maintained Their PF
The majority of women maintaining PF were 80–84 years of age ( N = 7,723; 73.7%), 85–89 ( N = 2,724; 26.0%), 90+ ( N = 31; 0.3%), and had a high school education or less ( N = 2,166; 20.7%). Most of the women were White (not of Hispanic origin) ( N = 2,988; 89.4%), followed by Black or African American ( N = 159; 4.8%), Asian or Pacific Islander ( N = 82; 2.5%), Hispanic or Latino ( N = 59; 1.8%), other ( N = 40; 1.2%), and American Indian or Alaskan Native ( N = 13; 0.4%). A little over half had an income more than or equal to $35,000 ( N = 1,913; 57.3%). A little less than half had a BMI less than 25 ( N = 1,639; 49.1%); 25–29 ( N = 1,145; 34.3%); or at least 30 ( N = 557; 16.7%). Most of the women reported very good ( N = 1,504; 45%) or good ( N = 1,190; 35.6%), versus excellent ( N = 394; 11.8%); fair ( N = 235; 7.0%); or poor ( N = 18; 0.5%) health.
Health outcomes for the women included a history of hip fracture at age 55 and above and a history of CVD. The majority of women maintaining PF had no history of hip fracture at age 55 and above ( N = 3,202; 95.8%) and no history of CVD ( N = 2,793; 83.6%). Less than half the women were never smokers ( N = 1,440; 43.1%), and a very small percentage were current smokers ( N = 34; 1.0%). Most women reported an average of less than one drink per day ( N = 2,197; 65.8%), followed by no alcohol ( N = 995; 29.8%); one to two drinks per day ( N = 139; 4.2%); and three or more drinks per day ( N = 10; 0.3%). A large majority of these women reported no depression ( N = 3,135; 93.8%). Participants’ last PF score before turning 80 was mean ( SD ) = 86.3 (15.3) and most recent PF score was 70.4 (24.1). Participants’ most recent BMI was 25.9 (4.7).
Functional Independence in Women Older Than 80
Table 2 describes the current level as well as slope increases in self-reported ADL and IADL limitations for unadjusted and adjusted models. Unless otherwise noted, we report statistics from models adjusted for all of the covariates included in the analyses in the text. Current levels of ADL limitations (mean [ SD ]: 4.11 [0.64]) were increased across the three trajectory groups with the greatest limitations seen in women with rapidly declining PF, as hypothesized. Least squares mean ( SE ) ADL scores were maintaining = 4.06 (0.01), slowly declining = 4.10 (0.01), and rapidly declining = 4.24 (0.01) (overall p < .0001). Current levels of IADL limitations (mean [ SD ]: 2.28 [0.74]) were also increased across the three trajectory groups, again with the greatest limitations seen in women with rapidly declining PF (maintaining = 2.23 [0.01], slowly declining = 2.25 [0.01], rapidly declining = 2.46 [0.02]) (overall p < .0001). Increases in IADL slope were significant across groups (maintaining = mean [ SE ] = 0.04 [0.003], slowly declining = 0.04 [0.003], rapidly declining = 0.06 [0.004]) (overall p = .002), whereas increases in ADL slope were not (maintaining = mean [ SE ] = 0.02 [0.003], slowly declining = 0.02 [0.003], rapidly declining: 0.03 [0.005]) (overall p = .10; unadjusted overall p ≤ .0001).
Table 2.
Current Functional Independence in Women Older Than 80 and Change in Functional Independence After Age 80 by Physical Functioning Trajectory Group
| Physical Functioning Trajectory Group | ||||||
|---|---|---|---|---|---|---|
| Functional Independence | Model | Total Sample N (%) 10,478 (100%) | Rapidly Declining Physical Functioning 1,989 (19%) | Slowly Declining Physical Functioning 3,341 (32%) | Maintaining Physical Functioning 5,148 (49%) | Overall p Value |
| Mean ( SD ) | LSM ( SE ) | LSM ( SE ) | LSM ( SE ) | |||
| ADL | Unadjusted | 4.11 (0.64) | 4.24 (0.01)* | 4.10 (0.01) † | 4.06 (0.01) † | <.0001 |
| Adjusted | 4.18 (0.01) | 4.10 (0.01) † | 4.08 (0.01) † | <.0001 | ||
| IADL | Unadjusted | 2.28 (0.74) | 2.58 (0.02)* | 2.24 (0.01) † | 2.18 (0.01) † | <.0001 |
| Adjusted | 2.46 (0.02) | 2.25 (0.01) † | 2.23 (0.01) † | <.0001 | ||
| ADL slope (score/y) | Unadjusted | 0.020 (0.20) | 0.040 (0.004) | 0.016 (0.0003) † | 0.014 (0.003) † | <.0001 |
| Adjusted | 0.028 (0.005) | 0.017 (0.003) | 0.019 (0.003) | .0959 | ||
| IADL slope (score/y) | Unadjusted | 0.044 (0.19) | 0.077 (0.004) | 0.038 (0.003) † | 0.034 (0.003) † | <.0001 |
| Adjusted | 0.058 (0.004) | 0.040 (0.003) † | 0.043 (0.003) † | .0015 | ||
Notes: Adjusted models include all covariates: age at visit with physical functioning; education; race/ethnicity; income <$35,000; baseline body mass index (kg/m 2 ); self-reported health; history of hip fracture ≥55 y; history of cardiovascular disease defined as having been diagnosed with coronary heart disease, adjudicated stroke (ischemic, hemorrhagic), venous thromboembolism, congestive heart failure, and peripheral artery disease among women aged 80 and older in the Women’s Health Initiative Study cohort; smoking status; alcohol intake; and depression. ADL range = 2–6; IADL range = 4–12. ADL = activity of daily living; IADL = instrumental activity of daily living; LSM = least square mean.
*Comparison between maintaining physical function trajectory with slowly declining physical function trajectory, p < .01.
† Comparison between maintaining or slowly declining physical function trajectories with rapidly declining physical function trajectory, p < .01.
In comparisons of maintaining versus rapidly declining PF trajectory groups ( Table 2 ), there were fewer limitations in ADL (4.08 [0.01] vs 4.18 [0.01]) and IADL (2.23 [0.01] vs 2.46 [0.02]), and less slope increase in IADL limitations (0.04 [0.003] vs 0.06 [0.004]) (all p < .01). In the maintaining PF trajectory group, there were fewer limitations in all the functional independence outcomes, but no difference in the slope increase in ADL limitations.
In comparisons of slowly declining versus rapidly declining PF trajectory groups ( Table 2 ), there were fewer limitations in ADL (4.10 [0.01] vs 4.18 [0.01]) and IADL (2.25 [0.01] vs 2.46 [0.02]), and less slope increase in IADL limitations (0.04 [0.003] vs 0.06 [0.004]) (all p < .01). In the slowly declining PF trajectory group, there were fewer limitations in all the functional independence outcomes, but no difference in the slope increase in ADL limitations.
In comparisons between maintaining and slowly declining PF trajectory groups ( Table 2 ), there were no significant differences in ADL, IADL ADL slope, or IADL slope (all p > .25). In unadjusted models, there were fewer current ADL limitations (4.06 [0.01] vs 4.10 [0.01]) (unadjusted p < .01) and IADL limitations (2.18 [0.01] vs 2.24 [0.01]) (unadjusted p < .01).
Characteristics of Women Older Than 80 That Predict Functional Independence
When we examined characteristics of women thought to predict functional independence and change in functional independence outcomes ( Table 3 ), women with BMI less than 25 had significantly fewer IADL limitations than women with BMI more than 30 (2.27 [0.01] vs 2.32 [0.01]) ( p < .05). Women with excellent self-reported health had significantly fewer ADL limitations (4.04 [0.02] vs 4.26 [0.02]) ( p < .01) and IADL limitations (2.13 [0.03] vs 2.65 [0.02]) ( p < .01) across all self-reported health categories. Women with no history of hip fracture had fewer limitations in ADL (4.11 [0.01] vs 4.21 [0.03]) ( p < .001) and IADL (2.27 [0.01] vs 2.52 [0.03]) ( p < .01), and less increase in IADL slope (0.04 [0.002] vs 0.08 [0.01]) ( p < .01). Women with no history of CVD had fewer limitations in ADL (4.10 [0.01] vs 4.16 [0.01]) ( p < .01) and IADL (2.27 [0.01] vs 2.35 [0.02]) ( p < .01), but not less of an increase in ADL slope and IADL slope (both p > .10). Women who reported little or no alcohol intake had significantly more IADL limitations then women who had three or more drinks per day (2.33 [0.012] vs 2.08 [0.11]) ( p < .01) and greater slope increase in IADL limitations (0.05 [0.003] vs 0.005 [0.03]) ( p < .001). In the absence of detailed information about alcohol use and abuse in this cohort, we chose not to interpret this counterintuitive finding. Women who reported no depression had significantly fewer limitations in IADL (2.28 [0.01] vs 2.36 [0.03]) ( p < .05), but not ADL, ADL slope, or IADL slope (all p > .5).
Table 3.
Risk and Protective Factors by Functional Independence Outcome (Adjusted and Unadjusted Models)
| Unadjusted | Adjusted | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| ADL | ADL Slope | IADL | IADL Slope | ADL | ADL Slope | IADL | IADL Slope | ||
| Predictor | Category | LSM ( SE ) | LSM ( SE ) | LSM ( SE ) | LSM ( SE ) | LSM ( SE ) | LSM ( SE ) | LSM ( SE ) | LSM ( SE ) |
| Physical functioning trajectory group | Maintaining | 4.057 (0.011)* | 0.014 (0.003)* | 2.177 (0.012)* | 0.034 (0.003)* | 4.085 (0.011)* | 0.019 (0.003) | 2.233 (0.012)* | 0.043 (0.003) † |
| Slowly declining | 4.096 (0.009) | 0.016 (0.003) | 2.237 (0.010) | 0.038 (0.003) | 4.101 (0.009) | 0.017 (0.003) | 2.248 (0.010) | 0.040 (0.003) | |
| Rapidly declining | 4.244 (0.014) | 0.041 (0.004) | 2.580 (0.016) | 0.077 (0.004) | 4.184 (0.014) | 0.028 (0.005) | 2.457 (0.016) | 0.058 (0.004) | |
| Age at visit with physical function | 80–84 | 4.106 (0.007) | 0.019 (0.002) | 2.267 (0.008) † | 0.043 (0.002) | 4.109 (0.007) | 0.020 (0.002) | 2.274 (0.008) | 0.044 (0.002) |
| 85–89 | 4.127 (0.012) | 0.021 (0.004) | 2.327 (0.014) | 0.046 (0.004) | 4.117 (0.012) | 0.019 (0.004) | 2.306 (0.013) | 0.044 (0.004) | |
| 90+ | 4.226 (0.115) | 0.047 (0.035) | 2.484 (0.132) | 0.060 (0.034) | 4.212 (0.112) | 0.048 (0.035) | 2.428 (0.123) | 0.054 (0.033) | |
|
Education:
high school or less |
Yes | 4.116 (0.014) | 0.019 (0.004) | 2.303 (0.016) | 0.042 (0.004) | 4.116 (0.013) | 0.019 (0.004) | 2.303 (0.015) | 0.042 (0.004) |
| No | 4.110 (0.007) | 0.020 (0.002) | 2.278 (0.008) | 0.045 (0.002) | 4.110 (0.007) | 0.020 (0.002) | 2.278 (0.007) | 0.045 (0.002) | |
| Race/ethnicity | American Indian or Alaskan native | 4.212 (0.112) | 0.034 (0.034) | 2.455 (0.128) | 0.086 (0.033) | 4.189 (0.108) | 0.030 (0.034) | 2.380 (0.119) | 0.077 (0.032) |
| Asian or Pacific Islander | 4.059 (0.045) | 0.023 (0.014) | 2.291 (0.052) | 0.033 (0.013) | 4.058 (0.044) | 0.024 (0.014) | 2.273 (0.048) | 0.028 (0.013) | |
| Black or African American | 4.107 (0.031) | 0.013 (0.010) | 2.341 (0.035) | 0.044 (0.009) | 4.075 (0.030) | 0.007 (0.010) | 2.270 (0.033) | 0.034 (0.009) | |
| Hispanic/Latino | 4.133 (0.050) | 0.032 (0.015) | 2.313 (0.057) | 0.051 (0.015) | 4.095 (0.048) | 0.024 (0.015) | 2.246 (0.053) | 0.041 (0.014) | |
| White (not of Hispanic origin) | 4.112 (0.007) | 0.020 (0.002) | 2.278 (0.008) | 0.044 (0.002) | 4.114 (0.006) | 0.020 (0.002) | 2.283 (0.007) | 0.045 (0.002) | |
| Other | 4.117 (0.061) | 0.015 (0.019) | 2.387 (0.070) | 0.062 (0.018) | 4.120 (0.059) | 0.016 (0.018) | 2.383 (0.065) | 0.061 (0.018) | |
| Income <$35,000 | Yes | 4.131 (0.009) † | 0.021 (0.003) | 2.328 (0.011)* | 0.050 (0.003) † | 4.114 (0.009) | 0.018 (0.003) | 2.289 (0.010) | 0.045 (0.003) |
| No | 4.096 (0.008) | 0.019 (0.003) | 2.247 (0.010) | 0.040 (0.003) | 4.110 (0.008) | 0.021 (0.003) | 2.278 (0.009) | 0.044 (0.002) | |
| BMI (kg/m 2 ) | <25 | 4.090 (0.010)* | 0.017 (0.003) † | 2.224 (0.011)* | 0.041 (0.003) ‡ | 4.109 (0.010) | 0.020 (0.003) | 2.268 (0.011) ‡ | 0.046 (0.003) |
| 25–29 | 4.096 (0.010) | 0.015 (0.003) | 2.277 (0.012) | 0.041 (0.003) | 4.098 (0.010) | 0.016 (0.003) | 2.279 (0.011) | 0.042 (0.003) | |
| At least 30 | 4.175 (0.013) | 0.032 (0.004) | 2.398 (0.015) | 0.054 (0.004) | 4.136 (0.013) | 0.026 (0.004) | 2.315 (0.014) | 0.043 (0.004) | |
| In general self-reported health is | Excellent | 4.019 (0.022)* | 0.007 (0.007)* | 2.075 (0.025)* | 0.018 (0.007)* | 4.044 (0.022)* | 0.009 (0.007)* | 2.130 (0.025)* | 0.020 (0.007)* |
| Very good | 4.044 (0.010) | 0.003 (0.003) | 2.117 (0.011) | 0.016 (0.003) | 4.058 (0.010) | 0.005 (0.003) | 2.150 (0.011) | 0.018 (0.003) | |
| Good | 4.101 (0.009) | 0.018 (0.003) | 2.299 (0.010) | 0.045 (0.003) | 4.094 (0.009) | 0.017 (0.003) | 2.285 (0.010) | 0.044 (0.003) | |
| Fair | 4.295 (0.017) | 0.059 (0.005) | 2.724 (0.019) | 0.118 (0.005) | 4.262 (0.018) | 0.056 (0.006) | 2.646 (0.020) | 0.113 (0.005) | |
| Poor | 5.317 (0.056) | 0.295 (0.018) | 3.724 (0.062) | 0.298 (0.017) | 5.282 (0.056) | 0.291 (0.018) | 3.640 (0.062) | 0.293 (0.017) | |
| History of hip fracture ≥ 55 y | No | 4.105 (0.006)* | 0.019 (0.002) ‡ | 2.267 (0.007)* | 0.042 (0.002)* | 4.106 (0.006) † | 0.019 (0.002) | 2.270 (0.007)* | 0.042 (0.002)* |
| Yes | 4.239 (0.028) | 0.041 (0.008) | 2.584 (0.032) | 0.089 (0.008) | 4.208 (0.027) | 0.035 (0.008) | 2.517 (0.030) | 0.079 (0.008) | |
| History of CVD | No | 4.088 (0.007)* | 0.016 (0.002) † | 2.243 (0.008)* | 0.040 (0.002)* | 4.099 (0.007)* | 0.018 (0.002) | 2.266 (0.008)* | 0.043 (0.002) |
| Yes | 4.198 (0.014) | 0.033 (0.004) | 2.432 (0.015) | 0.059 (0.004) | 4.160 (0.013) | 0.026 (0.004) | 2.346 (0.015) | 0.047 (0.004) | |
| Smoking status | Never | 4.124 (0.009) | 0.021 (0.003) | 2.292 (0.010) | 0.044 (0.003) | 4.125 (0.008) | 0.022 (0.003) | 2.293 (0.009) | 0.045 (0.002) |
| Past | 4.097 (0.009) | 0.019 (0.003) | 2.271 (0.011) | 0.044 (0.003) | 4.095 (0.009) | 0.018 (0.003) | 2.269 (0.010) | 0.043 (0.003) | |
| Current | 4.094 (0.057) | 0.016 (0.017) | 2.328 (0.065) | 0.056 (0.017) | 4.104 (0.055) | 0.018 (0.017) | 2.322 (0.060) | 0.055 (0.016) | |
| Alcohol intake (drinks/d) | None | 4.159 (0.011)* | 0.024 (0.003) | 2.386 (0.013)* | 0.056 (0.003)* | 4.135 (0.011) | 0.020 (0.003) | 2.334 (0.012)* | 0.050 (0.003) ‡ |
| <1 | 4.089 (0.008) | 0.017 (0.002) | 2.236 (0.009) | 0.038 (0.002) | 4.100 (0.008) | 0.019 (0.002) | 2.259 (0.008) | 0.041 (0.002) | |
| 1–2 | 4.105 (0.029) | 0.025 (0.009) | 2.246 (0.034) | 0.050 (0.009) | 4.114 (0.029) | 0.027 (0.009) | 2.270 (0.032) | 0.053 (0.009) | |
| ≥3 | 4.000 (0.104) | −0.002 (0.032) | 2.026 (0.119) | −0.005 (0.031) | 4.031 (0.101) | 0.005 (0.032) | 2.081 (0.111) | 0.005 (0.030) | |
| Depression | No | 4.105 (0.006) † | 0.018 (0.002) † | 2.266 (0.007)* | 0.042 (0.002)* | 4.112 (0.006) | 0.020 (0.002) | 2.278 (0.007) † | 0.044 (0.002) |
| Yes | 4.195 (0.024) | 0.040 (0.007) | 2.508 (0.027) | 0.073 (0.007) | 4.110 (0.024) | 0.022 (0.007) | 2.357 (0.026) | 0.048 (0.007) | |
Notes: Adjusted models include all covariates: age at visit with physical functioning; education; race/ethnicity; income <$35,000; baseline BMI (kg/m 2 ); self-reported health; history of hip fracture ≥55 y; history of CVD defined as having been diagnosed with coronary heart disease, adjudicated stroke (ischemic, hemorrhagic), venous thromboembolism, congestive heart failure, and peripheral artery disease among women aged 80 and older in the Women’s Health Initiative Study cohort; smoking status; alcohol intake; and depression. ADL = activity of daily living; BMI = body mass index; CVD = cardiovascular disease; IADL = instrumental activity of daily living; LSM = least square mean.
* p < .0001. †p < .01. ‡p < .05.
Overall, in fully adjusted models, better self-reported health, no history of hip fracture at age 55 and above, and no history of CVD predicted better ADL performance. In addition, better self-reported health, no history of hip fracture at age 55 and above, no history of CVD, BMI less than 25, and no depression were significant predictors of better IADL performance. Only women with a history of hip fracture at age 55 and above had a slope increase in IADL impairment (in addition to worse ADL and IADL performance). The only significant predictor of slope increase in ADL impairment in the fully adjusted model was excellent or very good self-reported health, although a number of variables including BMI, history of CVD including stroke, and depression were significant in the unadjusted models.
Discussion
About a third of the women older than 80 in the WHI maintained their PF during older adulthood, and an additional 49% had slowly declining PF over time. The majority of women who maintained their PF were in the youngest (80–84) age category, had no history of hip fracture after age 55, no history of CVD, and no depression. A little more than half had an income greater than $35,000, had never smoked, and drank less than one drink per day. Slightly less than half had a BMI less than 25, and self-reported very good but not excellent health. Women who had maintained their PF in older adulthood had significantly better self-reported performance on all of the functional independence outcomes than women with rapidly declining PF, in fully adjusted models. These women had better current ADL and IADL function and less decline in ADL and IADL function after age 80. When compared to women who had slowly declining PF over time, those who maintained their PF had better current ADL and IADL performance in unadjusted models, but fully adjusted models did not discriminate between the latter two groups on any of the functional independence outcomes.
ADL and ΔADL
Greater independence in ADL after age 80 was predicted by having maintained or slowly declining PF, self-reported excellent or very good health, no history of hip fracture after age 55, and no history of CVD. Less ADL decline was predicted only by having excellent or very good self-reported health. In-line with disability frameworks ( 14 , 15 , 29 , 30 ) that would predict a “sequential progression” of decreased PF, followed by decreased independence in ADL, our findings demonstrate that maintaining PF across the older adult lifespan extends performance of the basic self-care skills necessary to function independently in late-life. Our results are also in-line with Hirsch and colleagues who found that change in physical performance measures (PPM; stride length and grip strength) predicted self-reported ADL disability in 4,182 Cardiovascular Health Study participants, mean age 79.4 years ( 31 ). Notably, women’s own insight into their health status was the only predictor of the rate at which decline in ADL took place.
IADL and ΔIADL
Greater independence in IADL was predicted by having maintained or slowly declining PF, a BMI less than 25, excellent or very good self-reported health, no history of hip fracture after age 55, no history of CVD, and no depression. Additionally, less IADL decline was predicted by having maintaining or slowly declining PF, excellent or very good self-reported health, and no history of hip fracture after age 55. The IADL construct represents the cognitive component of task performance ( 10 ). Our finding that a BMI less than 25 and no depression support IADL but not ADL function in women older than 80 is a logical extension of the finding that normal BMI and no depression support cognitive performance ( 32 ). Another important finding is the association between hip fracture and IADL and hip fracture and IADL decline. We found that 96% of the women who maintained their PF had no history of hip fracture after age 55, and that hip fracture after age 55 was the only factor that predicted a slope increase in IADL dependence.
Although research on the relationship between hip fracture and IADL is sparse, there is a healthy debate surrounding the causal direction of the relationship between physical performance and cognition, of which IADL is an indirect measure. For example, Atkinson and colleagues ( 32 ) examined the direction of the relationship between physical performance measures (PPM) and global global cognition in a series of models (global cognition → Δ PPM; PPM → Δ global cognition; Δ global cognition → Δ PPM), concluding that cognitive decline (on average) precedes or co-occurs with physical decline, but did not report the directional association between change in physical performance and change in global cognition.
Summary and Conclusions
Viewed together, greater independence in ADL and IADL are predicted by excellent or very good self-reported health, no history of hip fracture after age 55, and no history of CVD. In addition to these variables, greater independence in IADL is predicted by BMI less than 25 and no depression. Less decline in ADL and IADL function are also predicted by self-reported excellent or very good health, but in addition to these variables, less decline in IADL function is predicted by no history of hip fracture after age 55.
Measures of daily life as well as PF are limited by scaling and other issues such as type of report ( 33 , 34 ). In all age groups, it is fairly well documented that performance-based measures of physical function are more reliable than self-report, however, this may or may not be exacerbated with age. Further research may identify subgroups of older adults whose self-report is more or less reliable for various reasons, such as cognitive impairment (eg, Albert et al. ( 33 )). However, in support of our finding that maintaining PF in older adulthood predicts functional independence in women in late-life, we also found that women’s own insight into their health was a strong predictor of functional independence.
A somewhat unexpected finding was the strong relationship between PF and change in IADL, a purported indirect measure of cognitive function. There were fewer items on the IADL than the ADL scale, but more change in IADL than ADL in the adjusted models, thus arguing against an explanation that the measure of IADL is less sensitive because it has fewer items. Alternatively, more change in IADL could be attributed to the task impurity of the measures, for example, PF to some degree relies on motor planning (the ability to plan and execute task-related movement), which involves cognitive abilities such as memory and executive function, as does IADL ( 11 , 34 ). However, protective factors associated with ADL and IADL were dissociable, that is, there were unique predictors of IADL and IADL decline. Declines in IADL may be more accelerated in women older than 80, and this is an area for further research. In contrast, if there are no major declines in ADL performance due to an acute injury or disease process, ADL decline may occur more slowly than IADL decline, or not be well-captured by a dependence scale. Finally, we did not examine the bidirectional nature of the physical function/functional independence relationship in this paper, although improvements in PF are known to be achieved routinely by improving performance on daily life functional outcomes ( 15 ). Strengths of this study include the long history of women’s PF and lifestyle factors in a population-based cohort such as the WHI that we were able to examine in conjunction with with functional independence outcomes. Future research that addresses measurement issues surrounding research on daily life activity is needed in order to clarify these and other outstanding issues.
Results of this study provide support for the hypothesis that maintaining consistent PF in older adulthood extends functional independence in ADLs and IADLs in late-life. Predictors of better ADL and IADL function include maintaining PF over time, self-reported excellent or very good health, no history of hip fracture after age 55, and no history of CVD. Better IADL function is uniquely predicted by having a BMI less than 25 and no depression. Less ADL and IADL decline is predicted by better self-reported health, but less IADL decline is uniquely predicted by having no history of hip fracture after age 55. Remarkably, WHI women currently older than 80 years report that they have mostly independent ADL and IADL function, but that they “need a little help” with IADL function. This is true especially for women who are “rapidly” declining in physical function at a rate that is 1.35 times greater per year than women who are maintaining their physical function. This has clinical implications for providing support for women older than 80 years with rapidly declining physical function who are experiencing greater decreases in their IADL function than peers who are maintaining their physical function. Our findings highlight the developmental importance of maintaining or improving PF and preventing injury and disease in older adulthood (65–80), and the far-reaching implications this has for improving late-life (after age 80) functional independence in older adults.
Supplementary Material
Supplementary Appendix can be found at: http://biomedgerontology.oxfordjournals.org/
Funding
The WHI program is funded by the National Heart, Lung, and Blood Institute, NIH, U.S. Department of Health and Human Services, through contracts HSN268201100046C, HHSN268201100001C, HHSN 268201100002C, HSN268201100003C, HHSN268201100004C, and HHSN271201100004C.
References
- 1. Martin LG, Freedman VA, Schoeni RF, et al. Trends in disability and related chronic conditions among people ages 50 to 64 . Health Aff . 2010. ; 29 : 725 – 731 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. The Prevalence of Disability in Adults by Age Group, Behavioral Risk Factor Surveillance System (BRFSS) . Disability Tip Sheet. Centers for Disease Control and Prevention; ; 2009. . http://www.cdc.gov/search.do?subset=disabilityandhealth&subset=disabilityandhealth&queryText=disability+tip+sheet . Accessed February 9, 2015. [Google Scholar]
- 3. Gill TM, Kurland B . The burden and patterns of disability in activities of daily living among community-living older persons . J Gerontol A Biol Sci Med Sci . 2003. ; 58 : 70 – 75 . [DOI] [PubMed] [Google Scholar]
- 4. Lin S, Beck AN, Finch BK, Hummer RA, Master RK . Trends in US older adult disability: exploring age, period, and cohort effects . Am J Pub Health . 2012. ; 102 : 2157 – 2163 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Gill TM, Allore H, Guo Z . Restricted activity and functional decline among community-living older persons . Arch Intern Med . 2003. ; 163 : 1317 – 1322 . [DOI] [PubMed] [Google Scholar]
- 6. Jette AM, Branch LG, Sleeper LA, Feldman H, Sullivan LM . High-risk profiles for nursing home admission . Gerontologist . 1992. ; 32 : 634 – 640 . [DOI] [PubMed] [Google Scholar]
- 7. Naik AD, Concato J, Gill TM . Bathing disability in community-living older persons: common, consequential, and complex . J Am Geriatr Soc . 2004. ; 52 : 1805 – 1810 . [DOI] [PubMed] [Google Scholar]
- 8. Gitlin LN, Winter L, Dennis MP, Corcoran M, Schinfeld S, Hauck WW . A randomized trial of a multicomponent home intervention to reduce functional difficulties in older adults . J Am Geriatr Soc . 2006. ; 54 : 809 – 816 . [DOI] [PubMed] [Google Scholar]
- 9. Keeler E, Guralnik JM, Tian H, Wallace RB, Reuben DB . The impact of functional status on life expectancy in older persons . J Gerontol A Biol Sci Med Sci . 2010. ; 65 : 727 – 733 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Lawton MP, Brody EM . Assessment of older people: self-maintaining and instrumental activities of daily living . Gerontologist . 1969. ; 9 : 179 – 186 . [PubMed] [Google Scholar]
- 11. Vaughan L, Giovanello K . Executive function in daily life: age-related influences of executive processes on instrumental activities of daily living . Psychol Aging . 2010. ; 25 : 343 – 355 . [DOI] [PubMed] [Google Scholar]
- 12. Hays RD, Sherbourne CD, Mazel RM . The RAND 36-Item Health Survey 1.0 . Health Economics . 1993. ; 2 : 217 – 227 . [DOI] [PubMed] [Google Scholar]
- 13. Syddall HE, Martin HJ, Harwood RH, Cooper C, Aihie Sayer A . The SF-36: a simple, effective measure of mobility-disability for epidemiological studies . J Nutr Health Aging . 2009. ; 13 : 57 – 62 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. World Health Organization (WHO) . International Classification of Functioning and Disability: ICIDH-2 . Geneva, Switzerland: : World Health Organization; ; 2001. . [Google Scholar]
- 15. American Occupational Therapy Association (AOTA) . Occupational therapy practice framework: domain and process (3rd edition) . Am J Occup Ther . 2014. ; 68 : 139 . [Google Scholar]
- 16. Gill TM, Williams CS, Mendes de Leon CF, Tinetti ME . The role of change in physical performance in determining risk for dependence in activities of daily living among nondisabled community-living elderly persons . J Clin Epidemiol . 1997. ; 50 : 765 – 772 . [DOI] [PubMed] [Google Scholar]
- 17. Kelley-Hayes M, Jette AM, Wolf PA, D’Agostino R, Odell PM . Functional limitations and disability among elders in the Framingham Study . Am J Pub Health . 1992. ; 8 : 841 – 845 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Han L, Allore H, Murphy T, Gill T, Peduzzi P, Lin H . Dynamics of functional aging based on latent-class trajectories of activities of daily living . Ann Epidemiol . 2013. ; 23 : 87 – 92 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Cesari M, Onder M, Zamboni V, et al. Physical function and self-rated health status as predictors of mortality: results from longitudinal analysis in the ilSIRENTE study . BMC Geriatrics . 2008. ; 8 : 34 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Xue QL, Bandeen-Roche K, Mielenz TJ, et al. Patterns of 12-year change in physical activity levels in community-dwelling older women: can modest levels of physical activity help older women live longer? Am J Epidemiol . 2012. ; 176 : 534 – 543 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Stevens JA, Mack KA, Paulozzi LJ, Ballesteros MF . Self-reported falls and fall-related injuries among persons aged > or = 65 years in the United States . J Safety Res . 2006. ; 38 : 345 – 349 . [DOI] [PubMed] [Google Scholar]
- 22. Chase CA, Mann K, Wasek S, Arbesman M . Systematic review of the effect of home modification and fall prevention programs on falls and the performance of community-dwelling older adults . Am J Occup Ther . 2012. ; 66 : 284 – 291 . [DOI] [PubMed] [Google Scholar]
- 23. Hsu HC, Jones BL . Multiple trajectories of successful aging of older and younger cohorts . Gerontologist . 2012. ; 52 : 843 – 856 . [DOI] [PubMed] [Google Scholar]
- 24. Hultsch DF, Hertzog C, Small BJ, Dixon RA . Use it or lose it: engaged lifestyle as a buffer of cognitive decline in aging? Psychol Aging . 1999. ; 14 : 245 – 263 . [DOI] [PubMed] [Google Scholar]
- 25. Rowe JW, Kahn RL . Successful aging . The Gerontologist . 1997. ; 37 : 433 – 440 . [DOI] [PubMed] [Google Scholar]
- 26. Schaie KW . The optimization of cognitive functioning in old age: predictions based on cohort-sequential and longitudinal data . In Baltes PB, Baltes MM , eds. Successful Aging: Perspectives From the Behavioral Sciences (pp. 94 – 117 ). Cambridge: : Cambridge University Press; ; 1990. . [Google Scholar]
- 27. Ware JE, Jr, Sherbourne CD . The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection . Med Care . 1992. ; 30 : 473 – 483 . [PubMed] [Google Scholar]
- 28. Burnam MA, Wells KB, Leake B, Landsverk J. Development of a brief screening instrument for detecting depressive disorders. Med Care . 1988;8:775–789. [DOI] [PubMed] [Google Scholar]
- 29. Nagi SZ . An epidemiology of disability among adults in the United States . Milbank Mem Fund Q Health Soc . 1976. ; 54 : 439 – 467 . [PubMed] [Google Scholar]
- 30. Verbrugge LM, Jette AM . The disablement process . Soc Sci Med . 1994. ; 38 : 1 – 14 . [DOI] [PubMed] [Google Scholar]
- 31. Hirsch CH, Buzková P, Robbins JA, Patel KV, Newman AB . Predicting late-life disability and death by the rate of decline in physical performance measures . Age Ageing . 2012. ; 41 : 155 – 161 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Atkinson HH, Rapp SR, Williamson JD, et al. The relationship between cognitive function and physical performance in older women: results from the Women’s Health Initiative Memory Study . J Gerontol A Biol Sci Med Sci . 2010. ; 65A : 300 – 306 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Albert SM, Bear-Lehman J, Burkhardt A, Merete-Roa B, Noboa-Lemonier R, Teresi J . Variation in sources of clinician-rated and self-rated instrumental activities of daily living disability . J Gerontol A Biol Sci Med Sci . 2006. ; 61 : 826 – 831 . [DOI] [PubMed] [Google Scholar]
- 34. Gill TM, Robison JT, Tinetti ME . Difficulty and dependence: two components of the disability continuum among community-living older persons . Ann Intern Med . 1998. ; 128 : 96 – 101 . [DOI] [PubMed] [Google Scholar]