Abstract
Background
Chronic medical conditions (CMCs) often emerge and accumulate during the transition from mid- to late-life, and the resulting multimorbidity can greatly impact physical function. We assessed the association of CMC presence and incidence on trajectories of physical function from mid- to early late-life in the Study of Women’s Health Across the Nation.
Methods
Physical function was assessed at eight clinic visits (average 14 years follow-up) using the physical function subscale of the Short Form-36. CMCs included osteoarthritis, diabetes, stroke, hypertension, heart disease, cancer, osteoporosis, and depressive symptomatology, and were considered cumulatively. Repeated-measures Poisson models estimated longitudinal change (expressed as percent difference) in physical function by chronic CMCs. Change-points assessed physical function change coincident with the development of a new condition.
Results
Women (N = 2,283) followed from age 50.0 ± 2.7 to 64.0 ± 3.7 years; 7.3% had zero CMCs through follow-up, 22.5% (N = 513) had no baseline CMCs but developed ≥1, 22.7% women had ≥1 baseline CMC but never developed another, and 47.6% had ≥1 baseline CMC and developed ≥1 more. Each additional baseline CMC was associated with 4.0% worse baseline physical function and annual decline of 0.20%/year. Women with more baseline CMCs had greater decline in physical function with a new CMC (−1.90% per condition); and annual decline when developing a new condition accelerated by −0.33%/year per condition.
Conclusions
Self-reported physical function changes are evident from mid- to early late-life with the development of CMCs. Preventing or delaying CMCs may delay declines in physical function, and these potential pathways to disability warrant further research.
Keywords: Multimorbidities, Physical function, Disablement process
Chronic medical conditions (CMCs) are highly related to physical function limitations and higher likelihood of disability in late life, with multimorbidity—two or more CMCs (1)—increasing this risk (2–5). Furthermore, the deleterious effects of multimorbidity on physical function compound with age (2,6). Women have higher life expectancy compared to men, but women are more likely to have functional limitations and disability (7). Differences in multimorbidity are thought to play a role in sex disparities in the disablement process (8). The effect of multimorbidity on physical function is stronger in women compared to men, and some chronic conditions that greatly contribute to disability in women (like arthritis, cardiovascular disease, and hearing difficulties) do not have as large of an effect on men (2,8).
Much of the research regarding multimorbidity and physical function has focused on late-life. However, midlife represents a critical period for incidence and accumulation of CMCs (9), as the resulting multimorbidity can potentially impact late-life health. Physical function limitations are becoming more common in midlife, and parallel increases in chronic conditions likely contribute to this trend (10). Sex differences in physical function are apparent during midlife; women have worse balance, steeper decline in muscle strength, and are more likely to report physical function difficulties (11–14).
Though a clear link exists between CMCs and physical function, studies have not examined the longitudinal effect of the development of new CMCs on change in physical function from mid- to late-life, specifically among women. We hypothesize that physical function will decline faster among women with more CMCs, and will decline further with new and multiple conditions (multimorbidity). Given prior work suggesting that the onset of physical function limitations varies by race/ethnicity, economic strain, and body mass index (BMI) (15), we also assessed moderating effects of these factors.
Methods
Participants
Participants were from the Study of Women’s Health Across the Nation (SWAN), a longitudinal, multiethnic cohort study of health during the menopausal transition. Detailed descriptions of recruitment and design have been reported (16). Briefly, SWAN recruited 3,302 women between the ages of 42–52 years at baseline in 1996/1997. Eligible women had an intact uterus and at least one ovary, were premenopausal or early peri-menopausal (ie, had at least one menstrual period in the past 3 months), had not used reproductive hormones in the previous 3 months, and were not pregnant, lactating or breastfeeding. Women were recruited from seven field centers across the United States, including Boston, MA, Chicago, IL, southeastern Michigan, MI, Los Angeles, CA, Newark NJ, Oakland, CA, and Pittsburgh, PA. All sites recruited non-Hispanic white women plus women of a designated race/ethnic group (African American, Chinese, Japanese, or Hispanic) based upon location. Visit 4 (2000/2001) served as the analytic baseline as it began the most comprehensive assessment of CMCs. Eligible women had complete data on physical function and CMCs at Visit 4 and ≥2 additional physical function assessment visits through Visit 15 (2016/2017). In total, 2,566 women completed Visit 4, of which 2,283 were included in analyses. A total of 283 women were excluded from analyses; n = 46 because they were missing physical function scores at analytic baseline and n = 237 because they had less than two additional assessments of physical function. Compared to women included in analyses, excluded women were slightly younger (49.7 ± 2.7 vs 50.0 ± 2.7 years, p = .03), had higher BMI (29.6 ± 7.0 vs 28.8 ± 7.4 kg/m2, p < .001), had lower levels of education (p < .001), and higher financial strain (p < .001). Marital and menopausal status did not differ.
Physical Function Assessment
Self-reported physical function was assessed using the 10-item physical function subscale of the Medical Outcomes Study Short Form 36 (SF-36) (17). The physical function subscale was assessed at annual follow-up Visits 4, 6, 8, 10, 11, 12, 13, and 15. This scale assesses perceived limitation (“limited a lot,” “limited a little,” or “not limited at all”) in the following activities: vigorous activities; lifting or carrying groceries; climbing several flights of stairs; climbing one flight of stairs; bending, kneeling, or stooping; walking more than one mile; walking several blocks; walking one block; and bathing or dressing. Raw scores were transformed using norm-based scoring methods; possible scores range from 0 to 100 with higher scores indicating better physical function.
Chronic Medical Condition (CMC) Assessment
Eight CMCs were included: osteoarthritis, hypertension, diabetes, heart disease (myocardial infarction or angina), stroke, cancer (except skin), osteoporosis, and depressive symptoms. CMCs were chosen based on prior relationships to physical function and disability in studies of midlife or older women (3,15,18–20). CMCs were assessed via self-report of a health care provider diagnosis except depressive symptomatology, which was assessed using the Center for Epidemiologic Studies-Depression (CES-D) scale. Clinically relevant depressive symptomatology was defined as greater than or equal to 16 on the CES-D (21). The baseline number of CMCs was the total number of conditions reported inclusively from Visits 0 to 4.
CMCs were assessed at all annual follow-up visits between 4 and 15, inclusively. Because CMSs were assessed at additional visits where physical function was not assessed, CMCs identified in interim visits were carried forward as incident conditions the next visit physical function was assessed. Using these additional years ensured that we did not miss any new CMCs during nonoverlapping time periods, as the questionnaire asks about new CMCs since the last visit.
Covariate Assessment
Sociodemographic factors included age, self-identified race/ethnicity (non-Hispanic white, African American, Chinese, Hispanic, Japanese), financial strain (“How hard is it for you to pay for the very basics like food, housing, medical care, and heating?” Responses included “not hard at all,” “somewhat hard,” and “very hard,” with somewhat/very hard combined for analyses), marital status (combined single/never married, married/living as married, separated/widowed/divorced for analyses), and education (high school or less, some college, college degree, or post college education). Self-rated health status was assessed via responses “excellent,” “very good,” “good,” “fair,” and “poor.” Behavioral factors included current smoking and physical activity (assessed using the total score of Kaiser Physical Activity Survey, KPAS) (22). Cognitive function (perceptual speed) was assessed using the Symbol Digit Modalities Test (23). Other risk factors known to influence physical function included BMI (calculated using measured height and weight), hormone use, bodily pain (assessed using the SF-36), and self-reported fracture since last visit.
Statistical Methods
Baseline participant characteristics were compared by baseline and follow-up CMC status. Four categories of baseline and CMC development were created and used for descriptive purposes only, including women who had (i) no CMCs at baseline and remained CMC-free during follow-up, (ii) no baseline CMCs and developed at least one during follow-up, (iii) at least one baseline CMC but did not develop, and (iv) at least one baseline CMC and developed at least one more CMC. Continuous variables were compared across categories using ANOVA or Kruskal–Wallis tests. Categorical variables were compared using chi-squared tests. CMCs were summed to create a CMC count that increased over time as women developed new conditions. CMCs were considered cumulatively, such that once a woman had a CMC, it was not removed from her CMC count.
Physical function scores followed a Poisson-like distribution. Because possible scores only included integers, we used Poisson regression models to elucidate the relationships between CMCs and longitudinal physical function. Beta values from these models were exponentiated to be expressed as incident rate ratios (IRR) and converted to percents. Percent difference associated with each model parameter was calculated by subtracting the IRR from 1 and multiplying by 100. To assess the effect of baseline CMCs on longitudinal physical function, we used Poisson-mixed models with only the count of baseline CMC (eight possible CMCs; expressed continuously), time, and a CMC × time interaction. To determine whether there was a difference in the rate of change in physical function after the development of a new CMC, we used piecewise Poisson models with the change-point fixed at the time of a new CMC. We created a new indicator variable to signify visits on or after the development of the first new CMC. Finally, we tested whether women who started with more CMCs had a steeper decline in physical function with the addition of a new CMC by adding a three way interaction between baseline CMCs, time, and the CMC change-point. Final models were adjusted for covariates listed earlier. To understand the association between CMCs and physical function over the course of a decade, percent differences associated with annual changes were multiplied by 10. For illustrative purposes, we also calculated the percent change associated with developing a new CMC based upon having zero, one, two, or three baseline conditions.
In sensitivity analyses we stratified models by obesity (BMI <30 kg/m2 vs ≥30 kg/m2, race/ethnicity, and financial strain (any vs no difficulty) to determine whether physical function changed differently between these groups. Additionally, because we did not have information on clinical diagnoses of depression, we re-ran analyses excluding depressive symptomatology as one of the CMCs. To determine whether the rates of change in physical function by CMC development were driven by women with multiple CMCs at baseline, we stratified models by baseline CMC status (0–1 CMC vs 2+ CMCs).
Results
In this cohort of 2,283 midlife women (baseline age 50.0 ± 2.7 years), 166 (7.3%) women had no CMCs at baseline and remained CMC-free during the average 14 years follow-up, 513 (22.5%) had no baseline CMCs and developed at least one during follow-up, 517 (22.7%) had at least one baseline CMC but never developed another, and 1,087 (47.6%) had at least one baseline CMC and developed at least one more CMC. Participant characteristics significantly differed by CMC status, with the exception of fracture history (Table 1). Baseline physical function varied by CMC status, but was high overall. Women without CMCs through follow-up had the best physical function score (median 100, interquartile range: 95–100), whereas women with at least one baseline CMC who developed at least one new CMC during follow-up had the worst scores (median 85, interquartile range: 70–95) (p < .001). Of the women with at least one baseline CMC, nearly half (48.1%, N = 772) had two or more. At baseline the most common CMCs were clinically relevant depressive symptomatology (44.6%), osteoarthritis (30.5%), and hypertension (30.4%). Prevalence of all CMCs rose through follow-up, though these remained most common (Figure 1).
Table 1.
Baseline Participant Characteristics by Chronic Medical Condition (CMC) Development
| Participant Characteristics | No CMCs Through Follow-up N = 166 |
No Baseline CMCs, Developed ≥1 N = 513 |
≥1 Baseline CMC, No New N = 517 |
≥1 Baseline CMC, Developed ≥1 N = 1,087 |
p Value |
|---|---|---|---|---|---|
| Age, years, mean ± SDa | 49.5 ± 2.6 | 49.8 ± 2.7 | 50.0 ± 2.6 | 50.2 ± 2.7 | .002 |
| Race/ethnicity, n (%) | <.0001 | ||||
| Non-Hispanic white | 91 (54.8) | 273 (53.2) | 263 (50.9) | 495 (45.5) | |
| African American | 24 (14.5) | 92 (17.9) | 125 (24.2) | 352 (32.4) | |
| Hispanic | 5 (3.0) | 13 (2.5) | 32 (6.2) | 67 (6.2) | |
| Chinese | 26 (15.7) | 62 (12.1) | 48 (9.3) | 73 (6.7) | |
| Japanese | 20 (12.1) | 73 (14.2) | 49 (9.5) | 100 (9.2) | |
| Education, n (%) | <.0001 | ||||
| High school or less | 26 (15.8) | 85 (16.7) | 100 (19.5) | 259 (24.0) | |
| Some college | 43 (26.1) | 149 (29.3) | 187 (36.4) | 357 (33.1) | |
| College degree | 44 (26.7) | 119 (23.4) | 107 (20.8) | 221 (20.5) | |
| Post college | 52 (31.5) | 156 (30.7) | 120 (23.4) | 241 (22.4) | |
| Financial strain, any, n (%) | 32 (19.4) | 124 (24.8) | 190 (37.6) | 439 (41.3) | <.0001 |
| Marital status, n (%) | <.0001 | ||||
| Single/never married | 23 (13.9) | 54 (10.6) | 65 (12.7) | 165 (15.2) | |
| Married/living as married | 128 (77.1) | 370 (72.6) | 323 (62.8) | 659 (60.9) | |
| Separated/widowed/divorced | 15 (9.0) | 86 (16.9) | 126 (24.5) | 259 (23.9) | |
| Menopausal status, n (%) | .0002 | ||||
| Premenopausal | 19 (11.5) | 45 (8.8) | 41 (8.0) | 57 (5.3) | |
| Early peri-menopausal | 82 (49.4) | 258 (50.6) | 230 (44.8) | 447 (41.3) | |
| Late peri-menopausal | 14 (8.4) | 42 (8.2) | 53 (10.3) | 122 (11.3) | |
| Postmenopausal | 34 (20.5) | 78 (15.3) | 113 (22.0) | 240 (22.2) | |
| Surgical menopause | 1 (0.6) | 13 (2.6) | 15 (2.9) | 33 (3.1) | |
| Unknown | 16 (9.6) | 74 (14.5) | 62 (12.1) | 184 (17.0) | |
| Self-rated health, n (%) | <.0001 | ||||
| Excellent | 65 (39.2) | 142 (28.1) | 70 (13.7) | 90 (8.4) | |
| Very good | 72 (43.4) | 231 (45.7) | 192 (37.5) | 363 (33.8) | |
| Good | 26 (15.7) | 113 (22.3) | 175 (34.2) | 410 (38.2) | |
| Fair | 3 (1.8) | 20 (4.0) | 69 (13.5) | 178 (16.6) | |
| Poor | 0 (0.0) | 0 (0.0) | 6 (1.2) | 33 (3.1) | |
| BMIb (kg/m2), mean ± SD | 24.9 ± 4.2 | 26.2 ± 5.8 | 28.9 ± 7.3 | 30.5 ± 7.9 | <.0001 |
| Current smoker, n (%) | 9 (5.4) | 57 (11.2) | 58 (11.2) | 182 (16.8) | <.0001 |
| Bodily painc, median (interquartile range) | 84 (74, 100) | 84 (62, 84) | 72 (51, 84) | 62 (51, 84) | <.0001 |
| Fracture history, n (%) | 2 (1.2) | 7 (1.4) | 13 (2.5) | 17 (1.6) | .43 |
| KPASd score, mean ± SD | 8.2 ± 1.7 | 8.0 ± 1.7 | 7.7 ± 1.7 | 7.5 ± 1.8 | <.0001 |
| Physical functione, median (interquartile range) | 100 (95–100) | 95 (90–100) | 90 (75–100) | 85 (70–95) | <.0001 |
| Symbol Digit Modalities Test, mean ± SD | 57.2 ± 9.6 | 57.6 ± 10.5 | 54.9 ± 11.8 | 53.3 ± 11.9 | <.0001 |
Note: Percentages are reflective of participants with complete data for the measure. Missing data <5%. aStandard deviation; bBody mass index; cSF-36 Bodily Pain subscale, dKaiser Physical Activity Survey; eSF-36 Physical Function subscale.
Figure 1.
Prevalence of chronic medical conditions (CMCs) at baseline and through follow-up. Note: CMCs considered cumulatively.
In covariate-adjusted Poisson models examining the effect of baseline CMCs on longitudinal physical function, each baseline CMC was associated with 3.80% worse baseline physical function (p < .001) compared to having no baseline CMC, and an additional 0.32% annual worsening (p < .001) (Table 2, Model 1). This annual worsening is in addition to the average annual physical function decline of 0.20%. For example, holding demographic, lifestyle, socioeconomic, and other health factors constant, a woman without CMCs at baseline would have 6.6%, 12.8%, and 18.6% better physical function 10 years later compared to a woman who had one, two, or three baseline CMCs, respectively.
Table 2.
Association of Physical Functioning with Baseline Chronic Medical Conditions (CMCs), Incident CMCs, and Interactions with Timea
| Model Parameters | Model 1: Baseline CMCs Only |
Model 2: Change-Point Model (two-way interaction) |
Model 3: Change-Point Model (three-way interaction) |
|||
|---|---|---|---|---|---|---|
| IRRb (95% CIc) |
Percent Differenced | IRR (95% CI) |
Percent Difference | IRR (95% CI) |
Percent Difference | |
| Baseline CMCs | 0.962 (0.952, 0.972) | −3.80 | 0.944 (0.934, 0.954) | −5.60 | 0.960 (0.950, 0.970) | −4.00 |
| Time (years) | 0.998 (0.997, 0.999) | −0.23 | 0.996 (0.995, 0.997) | −0.40 | 0.998 (0.997, 0.999) | −0.20 |
| Baseline CMCs × time | 0.997 (0.997, 0.997) | −0.32 | 0.998 (0.997, 0.999) | −0.20 | ||
| New CMC | 1.012 (1.002, 1.022) | 1.20 | 0.995 (0.981, 1.009) | N/A | ||
| New CMC × time | 0.997 (0.996, 0.998) | −0.30 | 0.999 (0.998, 1.00) | N/A | ||
| Baseline CMCs × new CMC | 1.019 (1.009, 1.029) | 1.90 | ||||
| Baseline CMCs × new CMC × time | 0.997 (0.996, 0.998) | −0.30 |
Note: aModels adjusted for age, race/ethnicity, financial strain, marital status, education, hormone use, smoking, body mass index, health status, bodily pain, fracture history, physical activity, and perceptual speed; bIncident rate ratio, calculated by exponentiating beta values; cConfidence interval; dCalculated by subtracting the IRR from 1 and multiplying by 100.
In covariate-adjusted change-point models examining the association between incident CMCs on longitudinal physical function (Table 2, Model 2), the annual rate of decline in physical function was 0.40% (p < .001). Each baseline CMC was associated with 5.50% worse initial physical function. Physical function decreased by 1.20% in association with the development of a new CMC during follow-up, and declined an additional 0.30% per year. Thus, a woman who had no baseline CMCs and did not develop any CMCs during follow-up would have 5.3%, 10.3%, and 15.0% better physical function 10 years later compared to a woman who had one, two, or three baseline CMCs and developed a new CMC one year into follow-up.
Finally, we examined covariate-adjusted change-point models of the association between incident CMCs as well as baseline CMCs on longitudinal physical function (Table 2, Model 3). Each baseline CMC was associated with 4.0% worse baseline physical function, and an annual decline of 0.20% per year. There was a statistically significant interaction between number of baseline CMCs and number of incident CMCs; women with more baseline CMCs had a greater decrement in physical function (1.90% lower for each baseline CMC) with the development of a new CMC and a faster annual decline in physical function with the development of a new CMC (p < .001). Based upon the results from this three-way interaction, the model estimates that a woman who had no baseline CMCs and did not develop any incident CMCs during follow-up would have 5.7%, 11.0%, and 16.1% better physical function after a decade compared to a woman with one, two, or three baseline CMCs who also developed a new CMC a year into follow-up, respectively.
Figure 2 visually depicts the predicted longitudinal change in physical function by CMC baseline and follow-up status from Models 2 and 3, with change-points of new CMC development set at the first year of follow-up for illustrative purposes. For example, in the final, three-way interaction model (Figure 2B), women with no CMCs at baseline who never developed any during follow-up, had an average score of 96 at baseline and a score of 91 ten years later. Women who started with at least one CMC and developed at least one more had a score of 83 at baseline and a score of 76 ten years later. In the three-way interaction model (Figure 2B), a woman with no CMCs at baseline, who never developed any at follow-up, has a physical function score of 94, and 10 years later, will have a score of 92, compared to a woman who developed a new condition whose score will decrease to 89.
Figure 2.
Longitudinal changes in physical function by chronic medical condition (CMC) development. Note: For groups that developed a new CMC (dashed lines), the change-point was set at Year 1. Panel A represents the interaction between time and new CMCs; Panel B represents the three-way interaction between baseline CMCs, time, and new CMCs.
In sensitivity analyses, results remained similar when stratifying by obesity status, race/ethnicity, and financial hardship. Results remained consistent when restricting analyses to only seven possible CMCs, excluding clinically relevant depressive symptomatology. In stratified analyses, the rate of change in physical function attributed to CMCs did not differ between women who started with 0 or 1 CMC compared to women with 2+. (Sensitivity analysis results not shown.)
Discussion
This study demonstrated the longitudinal effects of incident CMCs on perceived changes in physical function that were self-reported from mid- to early late-life. Less than 10% of the women remained free of CMCs throughout the duration of follow-up from mid- to early late-life, and approximately half of women already had at least one CMC at baseline and developed at least one more. Among the women with baseline CMCs, approximately 48% already had more than one. Women were generally high functioning at baseline. However, women with baseline CMCs who developed more during follow-up had median baseline function scores of 85—indicative of moderate limitations when considering this scale categorically (24). Self-reported physical function declined concurrently with the development of a new CMC. Moreover, women with more baseline CMCs began with worse physical function, and they had accelerated physical function decline as they developed new CMCs. Although annual changes in physical function were small, these initial changes led to larger cumulative changes over the course of a decade and into late-life, and to a widening gap in perceived physical function between women with no CMCs and women who had and/or developed CMCs. Because women were initially in their early 50s—with more than three decades of life expectancy remaining—these anticipated future changes are substantial. Importantly, early declines in physical function during midlife likely increase the risk of objective physical function limitations and disability in late-life.
The prevalence of CMCs and multimorbidity is increasing in the midlife population in the United States (25); however, estimates vary greatly depending on CMCs considered and the study population (26). Furthermore, no current consensus exists as to which conditions should be included when studying multiple CMCs or how multimorbidity should be defined (27). For example, a nationally representative survey focusing on six CMCs clinically relevant for midlife women (joint/muscular pain, urinary incontinence, depression, osteoporosis risk, moderate/severe vasomotor symptoms, and vulvar/vaginal atrophy) found that 34% of women surveyed had one CMC, and 31% had two or more (28). Similarly, research using the Medical Expenditure Panel Survey, weighted to represent the U.S. population, showed that 24% of people age 45–64 had one CMC, and 39% had two or more. CMCs were based upon medical record review (included 111 medical codes) (25). From a biologic perspective, Nagi’s disablement model states that active pathology at the cellular level due to disease, injury, or developmental conditions leads to impairments at the organ or system level, further leading to functional limitations and disability (29,30), Thus, as the prevalence of CMCs and multimorbidity rises at younger ages, parallel increases in the prevalence of physical function limitations at younger ages are likely, as is the potential for greater morbidity in late-life among those who develop CMCs at younger ages due to longer duration of CMCs.
We investigated the longitudinal changes in perceived physical function in women on average from the age of 50.0 ± 2.7 to 64.0 ± 3.7 years. Still, further work is needed in order to determine whether the decline in physical function associated with CMCs accelerates with age, and how objective physical function changes in association with CMC development. In NHANES, for women age 65–74, each additional CMC was associated with 1.65 times more functional limitations, as assessed using a questionnaire assessing limitations in 19 activities over 5 domains (2). Notably, this effect was stronger than the effect seen in men in this same age group (1.35 times more limitations per CMC). By age 75+, each additional CMC was associated with more than twice as many functional limitations in older women compared to 1.71 times as many limitations in older men. By very late-life, (age 80+) the association between multimorbidity and physical function is especially evident, as demonstrated by the Women’s Health Initiative, where women with multiple CMCs had SF-36 physical function scores approximately 12 points lower (on the 100 point scale) than women with zero or one CMC (5). Conversely, a recent study examining change in limitations of instrumental activities of daily living (IADL) by multimorbidity status among older adults showed that the effect of age on IADL limitation development was strongest in those without chronic conditions (31). It was suggested that those with multimorbidity in early late-life may use different coping strategies to adapt to challenges with daily activities compared to those without CMCs.
Strengths of this study include the long follow-up period of a well-characterized cohort of women followed from midlife to early old age, and we used a well-studied metric for assessing physical function. It is important to note; however, that this physical function measure assesses perceived physical function. The longitudinal association between CMCs and physical function could differ when examining objective vs self-reported physical function, which is a future direction of this work. Additionally, change-point models allowed us to detect changes in physical function with the addition of a new CMC (incident CMCs), rather than only uniform change. Change-point models have been used in studies of cognitive decline, particularly for determining the onset of accelerated decline in preclinical Alzheimer’s disease (32,33). To the best of our knowledge, no published studies have used change-point analyses to examine accelerated change in physical function with CMCs, and this is a novel approach in multimorbidity research overall. A limitation is that all CMCs are weighted equally, when some may have a greater effect than others. A future direction is to examine the differential change in physical function associated with each type of CMC, rather than considering them together.
Further limitations include that once a woman developed a new CMC, we were unable to account for it resolving or worsening, which could affect physical function over time. We selected CMCs based upon their association with physical function and disability in midlife and older women (2,3,18–20). Still, some CMCs known to influence physical function were not included, like chronic kidney disease, chronic pulmonary diseases, peripheral arterial disease, and neuropathy as they were not regularly assessed in SWAN. Finally, though we included depressive symptoms, diagnosed depression was not assessed. Many women with clinically relevant depressive symptomatology may not have had diagnosed depression, and these symptoms may have resolved over time. For consistency, once a woman had a CES-D score of greater than 16 points, she was considered to have clinically relevant depressive symptoms. However, excluding depressive symptomatology as a CMC did not change results.
Prior research has documented associations between the burden of CMCs and physical function in late life, when CMCs are already prevalent (2,5). The current findings extend this literature by quantifying changes in physical function concomitant with new CMCs arising in midlife. Efforts to preserve physical function, either by preventing the development of CMCs or promoting uptake of rehabilitative programs, should focus on midlife as the accumulation of CMCs during this period sets the stage for subsequent declines in physical function.
Funding
The Study of Women’s Health Across the Nation (SWAN) has grant support from the National Institutes of Health (NIH), DHHS, through the National Institute on Aging (NIA), the National Institute of Nursing Research (NINR), and the NIH Office of Research on Women’s Health (ORWH) (grants U01NR004061; U01AG012505, U01AG012535, U01AG012531, U01AG012539, U01AG012546, U01AG012553, U01AG012554, and U01AG012495). The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the NIA, NINR, ORWH, or the NIH.
Authors’ Contributions
B.L.M. developed the study idea. E.A., I.J., and K.K. developed the analytic plan. K.K. performed analyses. H.M.K. obtained funding for the study. All coauthors provided critical input on the interpretation of results. B.L.M drafted the initial manuscript. All coauthors critically revised the manuscript and approved the final version.
Conflict of Interest
None reported.
Acknowledgments
Clinical Centers: University of Michigan, Ann Arbor—Siobán Harlow, PI 2011–present, MaryFran Sowers, PI 1994–2011; Massachusetts General Hospital, Boston, MA—Joel Finkelstein, PI 1999–present; Robert Neer, PI 1994–1999; Rush University, Rush University Medical Center, Chicago, IL—Howard Kravitz, PI 2009–present; Lynda Powell, PI 1994–2009; University of California, Davis/Kaiser—Ellen Gold, PI; University of California, Los Angeles—Gail Greendale, PI; Albert Einstein College of Medicine, Bronx, NY—Carol Derby, PI 2011–present, Rachel Wildman, PI 2010–2011; Nanette Santoro, PI 2004–2010; University of Medicine and Dentistry—New Jersey Medical School, Newark—Gerson Weiss, PI 1994–2004; and the University of Pittsburgh, Pittsburgh, PA—Karen Matthews, PI. NIH Program Office: National Institute on Aging, Bethesda, MD—Chhanda Dutta 2016–present; Winifred Rossi 2012–2016; Sherry Sherman 1994–2012; Marcia Ory 1994–2001; National Institute of Nursing Research, Bethesda, MD—Program Officers. Central Laboratory: University of Michigan, Ann Arbor—Daniel McConnell (Central Ligand Assay Satellite Services). Coordinating Center: University of Pittsburgh, Pittsburgh, PA—Maria Mori Brooks, PI 2012–present; Kim Sutton-Tyrrell, PI 2001–2012; New England Research Institutes, Watertown, MA—Sonja McKinlay, PI 1995–2001. Steering Committee: Susan Johnson, Current Chair; Chris Gallagher, Former Chair. We thank the study staff at each site and all the women who participated in SWAN.
References
- 1. Multimorbidity: Technical Series on Safer Primary Care. Geneva, Switzerland: World Health Organization; 2016. [Google Scholar]
- 2. Jindai K, Nielson CM, Vorderstrasse BA, Quinones AR. Multimorbidity and functional limitations among adults 65 or older, NHANES 2005–2012. Prev Chronic Dis. 2016;13:E151. doi:10.5888/pcd13.160174 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Fried LP, Bandeen-Roche K, Kasper JD, Guralnik JM. Association of comorbidity with disability in older women: the Women’s Health and Aging Study. J Clin Epidemiol. 1999;52:27–37. doi:10.1016/s0895-4356(98)00124-3 [DOI] [PubMed] [Google Scholar]
- 4. Dunlop DD, Manheim LM, Sohn MW, Liu X, Chang RW. Incidence of functional limitation in older adults: the impact of gender, race, and chronic conditions. Arch Phys Med Rehabil. 2002;83:964–971. doi:10.1053/apmr.2002.32817 [DOI] [PubMed] [Google Scholar]
- 5. Rillamas-Sun E, LaCroix AZ, Bell CL, Ryckman K, Ockene JK, Wallace RB. The impact of multimorbidity and coronary disease comorbidity on physical function in women aged 80 years and older: the women’s health initiative. J Gerontol A Biol Sci Med Sci. 2016;71(Suppl 1):S54–S61. doi:10.1093/gerona/glv059 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Stenholm S, Westerlund H, Head J, et al. Comorbidity and functional trajectories from midlife to old age: The Health and Retirement Study. J Gerontol A Biol Sci Med Sci. 2015;70:332–338. doi:10.1093/gerona/glu113 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Freedman VA, Wolf DA, Spillman BC. Disability-free life expectancy over 30 years: a growing female disadvantage in the US population. Am J Public Health. 2016;106:1079–1085. doi:10.2105/AJPH.2016.303089 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Whitson HE, Landerman LR, Newman AB, Fried LP, Pieper CF, Cohen HJ. Chronic medical conditions and the sex-based disparity in disability: the Cardiovascular Health Study. J Gerontol A Biol Sci Med Sci. 2010;65:1325–1331. doi:10.1093/gerona/glq139 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Merrill SS, Verbrugge LM. Chapter 4: Health and disease in midlife. In: Willis SL, Reid JD, eds. Life in the Middle: Psychological and Social Development in Middle Age. Amsterdam, the Netherlands: Elsevier; 1999:77–103. [Google Scholar]
- 10. Martin LG, Schoeni RF. Trends in disability and related chronic conditions among the forty-and-over population: 1997–2010. Disabil Health J. 2014;7(1 Suppl):S4–14. doi:10.1016/j.dhjo.2013.06.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Phillips SK, Rook KM, Siddle NC, Bruce SA, Woledge RC. Muscle weakness in women occurs at an earlier age than in men, but strength is preserved by hormone replacement therapy. Clin Sci (Lond). 1993;84:95–98. doi:10.1042/cs0840095 [DOI] [PubMed] [Google Scholar]
- 12. Samson MM, Meeuwsen IB, Crowe A, Dessens JA, Duursma SA, Verhaar HJ. Relationships between physical performance measures, age, height and body weight in healthy adults. Age Ageing. 2000;29:235–242. doi:10.1093/ageing/29.3.235 [DOI] [PubMed] [Google Scholar]
- 13. Kuh D, Bassey EJ, Butterworth S, Hardy R, Wadsworth ME; Musculoskeletal Study Team Grip strength, postural control, and functional leg power in a representative cohort of British men and women: associations with physical activity, health status, and socioeconomic conditions. J Gerontol A Biol Sci Med Sci. 2005;60:224–231. doi:10.1093/gerona/60.2.224 [DOI] [PubMed] [Google Scholar]
- 14. Cooper R, Hardy R, Aihie Sayer A, et al. Age and gender differences in physical capability levels from mid-life onwards: the harmonisation and meta-analysis of data from eight UK cohort studies. PLoS One. 2011;6:e27899. doi:10.1371/journal.pone.0027899 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Ylitalo KR, Karvonen-Gutierrez CA, Fitzgerald N, et al. Relationship of race-ethnicity, body mass index, and economic strain with longitudinal self-report of physical functioning: the Study of Women’s Health Across the Nation. Ann Epidemiol. 2013;23:401–408. doi:10.1016/j.annepidem.2013.04.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Sowers MF, Crawford SL, Sternfeld B, Morganstein D, Gold EB, Greendale GA. SWAN: a multi-center, multi-ethnic, community-based cohort study of women and the menopausal transition. In: Lobo RA, Kelsey J, Marcus R, eds. Menopause: Biology and Pathobiology. San Diego, CA: Academic Press; 2000:175–188. [Google Scholar]
- 17. 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]
- 18. Tseng LA, El Khoudary SR, Young EA, et al. The association of menopause status with physical function: the Study of Women’s Health Across the Nation. Menopause. 2012;19:1186–1192. doi:10.1097/gme.0b013e3182565740 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Karvonen-Gutierrez CA, Ylitalo KR. Prevalence and correlates of disability in a late middle-aged population of women. J Aging Health. 2013;25:701–717. doi:10.1177/0898264313488165 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Pope SK, Sowers MF, Welch GW, Albrecht G. Functional limitations in women at midlife: the role of health conditions, behavioral and environmental factors. Womens Health Issues. 2001;11:494–502. [DOI] [PubMed] [Google Scholar]
- 21. Boyd JH, Weissman MM, Thompson WD, Myers JK. Screening for depression in a community sample. Understanding the discrepancies between depression symptom and diagnostic scales. Arch Gen Psychiatry. 1982;39:1195–1200. doi:10.1001/archpsyc.1982.04290100059010 [DOI] [PubMed] [Google Scholar]
- 22. Sternfeld B, Ainsworth BE, Quesenberry CP. Physical activity patterns in a diverse population of women. Prev Med. 1999;28:313–323. doi:10.1006/pmed.1998.0470 [DOI] [PubMed] [Google Scholar]
- 23. Smith A. Symbol Digit Modalities Test: Manual. Los Angeles, CA: Western Psychological Services; 1982. [Google Scholar]
- 24. Rose MS, Koshman ML, Spreng S, Sheldon R. Statistical issues encountered in the comparison of health-related quality of life in diseased patients to published general population norms: problems and solutions. J Clin Epidemiol. 1999;52:405–412. doi:10.1016/s0895-4356(99)00014-1 [DOI] [PubMed] [Google Scholar]
- 25. Paez KA, Zhao L, Hwang W. Rising out-of-pocket spending for chronic conditions: a ten-year trend. Health Aff (Millwood). 2009;28:15–25. doi:10.1377/hlthaff.28.1.15 [DOI] [PubMed] [Google Scholar]
- 26. Fortin M, Stewart M, Poitras ME, Almirall J, Maddocks H. A systematic review of prevalence studies on multimorbidity: toward a more uniform methodology. Ann Fam Med. 2012;10:142–151. doi:10.1370/afm.1337 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Johnston MC, Crilly M, Black C, Prescott GJ, Mercer SW. Defining and measuring multimorbidity: a systematic review of systematic reviews. Eur J Public Health. 2019;29:182–189. doi:10.1093/eurpub/cky098 [DOI] [PubMed] [Google Scholar]
- 28. Lang K, Alexander IM, Simon J, et al. The impact of multimorbidity on quality of life among midlife women: findings from a U.S. nationally representative survey. J Womens Health (Larchmt). 2015;24:374–383. doi:10.1089/jwh.2014.4907 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. NAGI SZ. A study in the evaluation of disability and rehabilitation potential: concepts, methods, and procedures. Am J Public Health Nations Health. 1964;54:1568–1579. doi:10.2105/ajph.54.9.1568 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Verbrugge LM, Jette AM. The disablement process. Soc Sci Med. 1994;38:1–14. doi:10.1016/0277-9536(94)90294-1 [DOI] [PubMed] [Google Scholar]
- 31. Mueller-Schotte S, Zuithoff NPA, van der Schouw YT, Schuurmans MJ, Bleijenberg N. Trends in risk of limitations in instrumental activities of daily living over age in older persons with and without multiple chronic conditions. J Gerontol A Biol Sci Med Sci. 2019. doi:10.1093/gerona/glz049. https://www.ncbi.nlm.nih.gov/pubmed/30772903 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Yu L, Boyle P, Wilson RS, et al. A random change point model for cognitive decline in Alzheimer’s disease and mild cognitive impairment. Neuroepidemiology. 2012;39:73–83. doi:10.1159/000339365 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Hall CB, Lipton RB, Sliwinski M, Stewart WF. A change point model for estimating the onset of cognitive decline in preclinical Alzheimer’s disease. Stat Med. 2000;19:1555–1566. doi:10.1002/(sici)1097-0258(20000615/30)19:11/12<1555::aid-sim445>3.0.co;2-3 [DOI] [PubMed] [Google Scholar]