Association of Physical Activity and Physical Functioning Phenotypes With Fall Risk Among Women

Abstract

Objective: Physical activity (PA) may slow aging-related declines in physical functioning (PF), but the relationship of PA and falls is not well understood. This study examined the association of PA and PF with falls. Methods: The Study of Women’s Health Across the Nation participants (n = 1597; age: 65.1 years ± 2.7) reported PF and PA in 2012–2013 and falls in 2016–2017. Four phenotypes were identified: high PA–high PF, high PA–low PF, low PA–high PF, and low PA–low PF. Results: One-third (29.3%) reported ≥1 fall. Women with low PA–low PF (RR = 1.32; 95% CI: 1.06, 1.66) and with high PA–low PF (RR = 1.37; 95% CI: 1.07, 1.74) were more likely to fall than high PA–high PF. Over time, women with worsening PF had increased fall risk (RR = 1.43; 95% CI: 1.17, 1.74), but women who increased PA did not. Discussion: Poor PF increases the risk of falls, regardless of PA. However, increasing PA does not necessarily increase fall risk, reinforcing the importance of PA engagement.

Keywords

physical functioning physical activity falls

Introduction

Physical inactivity is a leading risk factor for obesity and related comorbidities, including type 2 diabetes, coronary heart disease, and stroke (Reiner et al., 2013). Current guidelines for physical activity (PA) recommend that adults accumulate at least 150–300 minutes of moderate intensity or 75–150 minutes of vigorous-PA per week, in addition to at least 2 days per week of muscle-strengthening activities, to achieve health benefits (U.S. Department of Health and Human Services (HHS) & Office of Disease Prevention and Health Promotion, 2018). Yet, many adults are insufficiently active. Surveillance reports from the 2018 National Health Interview Survey indicated that 25.4% of adults engaged in no leisure-time PA, and only 24.0% of adults met both aerobic and muscle-strengthening activity recommendations (Centers for Disease Control and Prevention Centers for, 2013).

Habitual PA is associated with a more favorable body composition throughout the life course. Aging-related changes in body composition including increased fat mass and decreased skeletal muscle mass may be prevented or delayed with consistent engagement in PA (Cruz-Jentoft et al., 2014; Steffl et al., 2017). Sarcopenia and related decreases in muscle mass and strength are associated with poor lower-extremity physical functioning (PF) and mobility (Sowers et al., 2005, 2006). Thus, PA engagement, including regular exercise, is particularly important for midlife and older adults because it delays or mitigates age-related functional declines through maintenance of muscle mass and strength, gait speed, and balance. Once functional impairments begin, individuals may have physical difficulty with engaging in PA and thus become less active.

Paradoxically, however, the literature on PA and falls is somewhat mixed. Some work suggests that PA increases the risk of falls because PA participation increases the opportunity to fall (Bea et al., 2017). Individuals with high levels of PA but poor PF may be at high fall risk due to an increased opportunity to fall during activity coupled with poor compensatory mechanisms like strength and balance to navigate an environmental challenge. Conversely, a recent report from the Women’s Health Initiative (WHI) demonstrated increased fall risk in older ambulatory women (mean age 79 years ± 6.7) with low PA and low PF but not an increased fall risk for women with high PA (Buchner et al., 2017). This work supports the importance of PF and suggests that unique phenotypes of PA engagement and functional status (e.g., low PA and low PF; high PA and low PF) may be important indicators of fall risk. This work should be replicated in younger women as findings from the Behavioral Risk Factor Surveillance System indicate that women aged 55–64 years report more falls and injurious falls than either younger or older women and all men (Ylitalo & Karvonen-Gutierrez, 2016).

The purpose of this study is to identify PA and PF characteristic groups (i.e., “phenotypes”) and describe the relationship of phenotypes and falls among women in late midlife (60–65 years) or early late adulthood (65–70 years) in the Study of Women’s Health Across the Nation (SWAN). We hypothesized that we would observe a trend in the association between phenotypes and falls such that women classified as having both high PA and high PF would report the lowest fall incidence, with increasingly higher fall incidence for the low PA and high PF group, the high PA and low PF group, and finally the low PA and low PF group.

Methods

Study Population

The Study of Women’s Health Across the Nation is a prospective, multicenter, multiethnic study of the menopause transition and aging. Sampling and recruitment methods have been described previously (Sowers et al., 2000). Briefly, 3302 women who at the 1996 baseline were aged 42–52 years, premenopausal, and self-identified with site-designated race–ethnic groups were recruited from seven locations (Boston, Chicago, Detroit, Los Angeles, New Jersey, Oakland, and Pittsburgh) for a longitudinal study (Sowers et al., 2000). Near-annual study visits provided clinical and behavioral data. For this analysis, participants with complete questionnaire data for PF and PA at both visit 12 (V12, 2012–2013) and visit 15 (V15, 2016–2017) and fall data at V15 were included. The final analytic sample for the present article was n = 1597 (48.4% of baseline) participants with complete data. Compared with the analytic sample, excluded women were similar in age but more likely to be Hispanic or African American (p < .001) and have a high school education or less (p < .001).

Measures

Falls were measured at visit 15 by self-report: “During the past year, have you fallen and landed on the floor or ground (or fallen and hit an object like a table or stair)?” If affirmative, two follow-up questions were asked: “How many times have you fallen in the past year?” and “During the past year, were you injured in any of these falls?” Women who reported an injury were additionally asked about the type of injury (broken or fractured bone, head injury, sprain or strain, bruises, or bleeding). Recurrent falls were defined as ≥2 falls during the previous year.

PF was measured by self-report from the 10-item PF subscale of the Medical Outcomes Survey Short Form (SF-36) (Ware & Sherbourne, 1992). The PF subscale demonstrates two-week test–retest reliability (intraclass correlation coefficient = .81) and substantial construct validity (Brazier et al., 1992). Participants reported whether they were limited a lot, a little, or not limited at all in vigorous or moderate activities, lifting/carrying groceries, climbing stairs, bending/kneeling/stooping, walking, bathing, and dressing. Responses were scored and transformed to a normalized range of 0–100. High scores indicate better PF. Previously used cut points were used to define no limitations (scores 86–100), some limitations (51–85), or substantial limitations (0–50) (Rose et al., 1999; Ylitalo et al., 2013).

PA was measured by self-report using the Kaiser Physical Activity Survey (KPAS), an adaptation of the Baecke questionnaire designed specifically for women (Ainsworth et al., 2000). For the sports and exercise domain, participants were asked to report up to two sports or exercise activities performed most frequently over the previous 12 months. The KPAS sports and exercise domain demonstrates 1-month test–retest reliability (intraclass correlation coefficient = .84) and reasonable accuracy compared to O₂ peak (Spearman correlation = .76; Ainsworth et al., 2000). Surveys obtained information about the activity type (open-ended), frequency (months in the previous 12 months), and duration (hours per week). Open-ended responses to primary and secondary sports/exercise activities were categorized into 65 groups and assigned a general metabolic equivalent of a task (MET) value using the Compendium of Physical Activities (Ainsworth et al., 2020). MET values were multiplied by frequency and duration to calculate average MET minutes per week. While the traditional KPAS scoring results in an index score ranging from 1 (low) to 5 (high), the sports and exercise domain was rescored and summarized as MET minutes per week in order to provide a more physiologically meaningful estimate.

High PA (≥450 MET minutes/week) and low PA (<450 MET minutes/week) were dichotomized based on the duration and intensity PA guidelines for adults in the United States (U.S. Department of Health and Human Services (HHS) & Office of Disease Prevention and Health Promotion, 2018). High PF (SF-12 scores ≥ 86, “no limitations”) and low PF (SF-12 scores < 86, “some or substantial limitations”) were dichotomized. We defined four “phenotypes,” or groups with similar characteristics, based on PF and PA clusters: (1) high PA and high PF, (2) high PA and low PF, (3) low PA and high PF, and (4) low PA and low PF.

Other covariates were intrinsic risk factors for falls (Institute of Medicine (US) Division of Health Promotion and Disease Prevention et al., 1992) and were selected a priori based on their known associations with falls. Covariates included age, race/ethnicity, educational attainment, marital status, socioeconomic status, body mass index (BMI), smoking status, alcohol consumption, depressive symptoms, osteoarthritis, and diabetes status. Height was measured without shoes in centimeters by a fixed stadiometer, and weight was measured in light clothing in kilograms by a balance beam scale. Body mass index was calculated as weight [kg]/height [m²] and categorized as underweight or normal weight (BMI ≤ 24.9 kg/m²), overweight (BMI 25–29.9 kg/m²), and obese (BMI ≥ 30 kg/m²) for non-Hispanic white, African American, and Hispanic women. We used Asian-specific cut points, categorized as underweight or normal weight (BMI < 23 kg/m²), overweight (BMI 23–24.9 kg/m²), and obese (BMI ≥ 25 kg/m²), for Japanese and Chinese women (Wee et al., 2010; WHO Expert Consultation, 2004; Woo et al., 2007). For all race–ethnic groups, the underweight/normal weight group was used as the reference. Diabetes status was defined using self-report of a physician diagnosis of diabetes, current use of diabetes medications, or, when available, a fasting blood glucose ≥126 mg/dL from visit 13. Smoking status was defined as current smoking (yes or no). Alcohol consumption was measured by self-report and categorized as none or ≤1 drink/month, >1 drink/month to <2 drinks/week, or ≥2 drinks/week. Depressive symptoms were ascertained using the Center for Epidemiological Studies Depression Scale (CES-D) and categorized as depressive symptoms (CES-D ≥16) or no depressive symptoms (CES-D <16) (Radloff, 1977).

Statistical Analysis

Data management and analyses were conducted using SAS v9.4 (SAS Institute, Inc, Cary, NC). Means (SD) or frequencies and proportions were reported for all variables of interest. Sociodemographic and health-related characteristics were compared by fall status using Student t-tests for continuous variables and chi-square tests for categorical variables. We identified four mutually exclusive PA–PF phenotypes at visit 12. Sociodemographic and health-related characteristics were also compared across phenotypes using generalized linear models for continuous variables and chi-square tests for categorical variables. Unadjusted and adjusted Poisson regression was performed using PROC GENMOD to generate risk ratios (RRs) and corresponding 95% confidence intervals (CIs) for binary fall outcomes. We evaluated the association of visit 12 phenotypes with visit 15 falls and visit 12 to visit 15 PA or PF change with visit 15 falls.

Results

Mean age was 65.1 (±2.7) years (range: 60–72). Almost half (49.4%) of the sample was non-Hispanic white, one quarter (24.3%) was non-Hispanic Black, 4.3% was Hispanic, 10.6% was Chinese, and 11.3% was Japanese. Approximately one-fifth (21.7%) reported that paying for basics like food, housing, and heat was very or somewhat difficult, half did not have a college degree, two-thirds were overweight or obese, and 16.3% had diabetes (Table 1).

Table 1.

Sociodemographic and health-related characteristics at visit 15 in total sample and by any fall status.

	Total	≥1 Fall	No Falls	p
	(n = 1597)	(n = 468)	(n = 1129)	p
Age, years (std)	65.1 (2.7)	65.2 (2.7)	65.1 (2.7)	0.38
Race/ethnicity, %
Non-Hispanic white	49.4	54.7	47.2	0.07
Non-Hispanic Black	24.3	21.6	25.4
Hispanic	4.3	4.7	4.2
Chinese	10.6	9.2	11.3
Japanese	11.3	9.8	12.0
Educational attainment, %
HS or less	19.9	18.5	20.5	0.04
Some college	30.4	28.0	31.4
College degree	23.0	21.9	23.4
Graduate degree	26.7	31.6	24.6
Marital status
Never married	12.4	11.5	12.8	0.11
Currently married	58.3	55.6	59.5
Formerly married	29.2	32.9	27.7
Socioeconomic status, %
Not difficult to pay for basics	78.3	73.7	80.3	0.004
Somewhat/very difficult	21.7	26.3	19.8
BMI categories, %
Underweight or normal	29.7	24.9	31.8	0.03
Overweight	27.9	30.8	26.6
Obese	42.4	44.4	41.6
Diabetes, %	16.3	21.2	14.4	<0.001
Current smoker, %	6.2	6.2	6.1	0.97
Alcohol consumption, %
None or ≤1 drink/month	51.6	50.0	52.3	0.53
>1 drink/month to <2 drinks/week	23.4	23.2	23.5
≥2 drinks/week	25.0	26.8	24.2
Depressive symptoms, %	12.8	19.7	10.0	<0.001
Osteoarthritis, %	36.8	45.3	33.2	<0.001
Physical functioning, SF-36 score (std)	76.0 (24.8)	71.4 (27.9)	77.9 (23.2)	<0.001
Physical activity, MET minutes/week (std)	496.1 (559.0)	480.3 (565.9)	502.7 (556.2)	0.47

At study visit 15, 29.3% (n = 468) of participants reported ≥1 fall in the previous year, 12.2% (n = 195) reported ≥2 falls (i.e., “recurrent falls”), and 8.0% (n = 128) reported recurrent falls with at least one injury. Among participants with recurrent falls, 17 reported a broken or fractured bone. Non-fracture fall injuries were common among the 128 women with recurrent injurious falls: 46 reported sprain/strain, 109 reported bruise, and 37 reported bleeding.

Participants who reported ≥1 fall were statistically significantly more likely to report difficulty paying for basics; be overweight or obese; and have diabetes, depressive symptoms, and poor PF. There were no statistically significant differences in fall status by age, race/ethnicity, marital status, smoking or alcohol consumption, or self-reported PA (Table 1).

At study visit 12, 22.7% (n = 362) reported high PA and high PF, 31.1% (n = 496) reported low PA and high PF, 18.8% (n = 301) reported high PA and low PF, and 27.4% (n = 438) reported low PA and low PF. We observed a statistically significant difference across these phenotypes in almost all sociodemographic and health variables. Non-Hispanic white women and Japanese women were more likely to be categorized in the high PA and high PF phenotype; non-Hispanic Black women were more likely to be categorized in the low PA and low PF phenotype. We observed a trend across phenotypes with respect to diabetes status, with 7.5% of high PA and high PF phenotypes having diabetes to 22.2% of low PA and low PF phenotypes having diabetes. Participants with low PF—regardless of activity level—were notably more likely to have depressive symptoms and osteoarthritis (Table 2).

Table 2.

Sociodemographic and health-related characteristics by phenotypes at visit 12.

	High PA–High PF	Low PA–High PF	High PA–Low PF	Low PA–Low PF	p
	n = 362	n = 496	n = 301	n = 438	p
Age, years (std)	59.6 (2.6)	59.7 (2.9)	59.7 (2.7)	59.7 (2.7)	0.83
Race/ethnicity, %
Non-Hispanic white	56.1	50.0	50.2	42.7	<0.001
Non-Hispanic Black	15.5	22.2	18.9	37.7
Hispanic	1.9	5.2	4.0	5.5
Chinese	8.8	11.7	14.6	8.2
Japanese	17.7	10.9	12.3	5.9
Educational attainment, %
HS or less	13.1	16.6	22.5	27.6	<0.001
Some college	26.7	30.8	28.9	34.0
College degree	27.6	24.8	20.1	19.1
Graduate degree	32.6	27.8	28.5	19.3
Marital status
Never married	9.7	9.9	13.6	15.8	0.002
Currently married	67.0	64.0	62.1	53.4
Formerly married	23.3	26.1	24.3	30.7
Socioeconomic status, %
Not difficult to pay for basics	82.5	78.5	66.4	56.7	<0.001
Somewhat/very difficult	17.6	21.5	33.6	43.3
BMI categories, %
Underweight or normal	35.5	37.3	20.1	19.7	<0.001
Overweight	34.6	33.9	25.5	21.7
Obese	29.9	28.8	54.5	58.6
Diabetes, %	7.5	10.1	17.9	22.2	<0.001
Current smoker, %	6.9	5.3	12.0	9.8	0.003
Alcohol consumption, %
None or ≤1 drink/month	46.0	46.7	54.0	57.2	0.002
>1 drink/month to <2 drinks/week	24.1	27.1	25.0	23.8
≥2 drinks/week	29.9	26.3	21.0	19.0
Depressive symptoms, %	10.3	6.9	25.7	20.7	<0.001
Osteoarthritis, %	16.9	15.7	40.5	48.4	<0.001
Physical functioning, SF-36 score (std)	95.6 (4.0)	96.2 (3.9)	64.7 (22.0)	60.3 (22.0)	<0.001
Physical activity, MET minutes/week (std)	1047.7 (488.9)	112.3 (129.1)	1046.6 (528.6)	110.6 (132.9)	<0.001
Any fall at V15, %	24.3	28.0	33.2	32.2	0.03
Recurrent falls at V15, %	9.7	11.7	14.3	13.5	0.24

Note. PA = physical activity; PF = physical functioning; MET = metabolic equivalent of a task.

Women who reported high PA and low PF at visit 12 were 37% more likely to report a fall at visit 15 than women with high PA and high PF (RR = 1.37; 95% CI: 1.07, 1.74; p = .01); women who reported low PA and low PF at visit 12 were 32% more likely to report a fall at visit 15 than women with high PA and high PF (RR = 1.32; 95% CI: 1.06, 1.66; p = .02). Neither association reached statistical significance at the α = .05 level after adjusting for SES, BMI, diabetes, and depressive symptoms (Table 3).

Table 3.

Relationship of physical activity and physical functioning phenotypes with fall outcomes at visit 15.

	Any Fall				Recurrent Fall
	Unadjusted		Adjusted		Unadjusted		Adjusted
V12 phenotypes	RR (95% CI)	p	RR (95% CI)	P	RR (95% CI)	p	RR (95% CI)	p
High PA–High PF (ref)	1.00		1.00		1.00		1.00
Low PA–High PF	1.15 (0.92, 1.45)	0.23	1.20 (0.95, 1.51)	0.13	1.21 (0.81, 1.80)	0.35	1.31 (0.85, 1.99)	0.22
High PA–Low PF	1.37 (1.07, 1.74)	0.01	1.19 (0.92, 1.55)	0.18	1.48 (0.97, 2.25)	0.07	1.22 (0.76, 1.95)	0.41
Low PA–Low PF	1.32 (1.06, 1.66)	0.02	1.22 (0.95, 1.55)	0.12	1.39 (0.94, 2.07)	0.10	1.21 (0.78, 1.88)	0.39

Note. PA = physical activity; PF = physical functioning.

Multivariate model adjusted for SES, BMI, diabetes, and depressive symptoms.

Finally, we evaluated change in PA and PF categories. From visit 12 to visit 15, 841 (52.7%) participants maintained the same phenotype, 273 (17.1%) participants changed from no PF limitations to some or substantial PF limitations, and 243 (15.2%) participants changed from not meeting PA guidelines to meeting PA guidelines (Supplemental Table 1). Compared to women who maintained or improved PF categories, women who developed some or substantial PF limitations—that is, moved to the high PA and low PF phenotype or to the low PA and low PF phenotype between visit 12 and visit 15— had a statistically significantly increased risk of falls. Even after adjusting for visit 12 PA MET minutes, change in PA category, SES, BMI, diabetes, depressive symptoms, and visit 12 PF score, women who experienced worsening PF from “no limitations” to “some or substantial limitations” were 43% more likely to report ≥1 fall (aRR = 1.43; 95% CI: 1.17, 1.74; p < .001) and 70% more likely to report recurrent falls (aRR = 1.70; 95% CI: 1.18, 2.43; p = .004). Compared to women who maintained PA or decreased PA categories, women who increased PA and began meeting national guidelines for PA—that is, moved to the high PA and high PF phenotype or to the high PA and low PF phenotype between visit 12 and visit 15— did not have a statistically significantly increased risk of either single or recurrent falls (Table 4).

Table 4.

Relationship of physical activity and physical functioning change (V12 to V15) with fall outcomes at visit 15.

Any Fall	Model 1		Model 2		Model 3
Any Fall	RR (95% CI)	p	RR (95% CI)	p	RR (95% CI)	p
Lower PF (vs. Higher PF)	1.25 (1.04, 1.50)	0.02	1.35 (1.12, 1.61)	0.001	1.43 (1.17, 1.74)	<0.001
Higher PA (vs. Lower PA)	0.87 (0.69, 1.09)	0.24	0.88 (0.70, 1.11)	0.28	0.88 (0.70, 1.11)	0.28

Recurrent Fall	Model 1		Model 2		Model 3
Recurrent Fall	RR (95% CI)	p	RR (95% CI)	p	RR (95% CI)	P
Lower PF (vs. Higher PF)	1.33 (0.97, 1.82)	0.08	1.40 (1.00, 1.96)	0.05	1.70 (1.18, 2.43)	0.004
Higher PA (vs. Lower PA)	0.94 (0.65, 1.37)	0.75	0.99 (0.68, 1.45)	0.97	1.01 (0.69, 1.49)	0.95

Note. PA = physical activity; PF = physical functioning.

Model 1: Multivariate model with lower PF and higher PA as independent variables.

Model 2: Model 1 + SES, BMI, diabetes, and depressive symptoms.

Model 3: Model 2 + V12 PF score and V12 PA MET minutes.

Discussion

The purpose of this study was to identify PA and PF phenotypes and describe the relationship of phenotypes and fall outcomes in late midlife and early old age women. Although we expected a large proportion of women to be classified as highly active without functional limitations or highly inactive with functional limitations, we also observed that approximately half of our sample had discordant levels of PA and PF. We hypothesized that discordant phenotypes (e.g., low PA high PF; high PA low PF) would be at greater fall risk than women with high PA and high PF. However, we observed that only the low PF groups were associated with falls, and this association was mitigated after adjustment for covariates. Nevertheless, the four PA and PF phenotypes appeared to identify important patterns in health profiles, particularly for cardiometabolic factors like BMI and diabetes status. Over time, women who experienced declines in PF phenotype categories had a significantly increased risk of falls, but women who increased PA phenotype categories did not have increased risk of falls. These findings suggest that women can engage in PA to the extent they are able without increasing fall risk, even for women with functional limitations.

Current guidelines for PA engagement recommend that adults accumulate at least 150–300 minutes a week of moderate intensity or 75–150 minutes of vigorous intensity aerobic activity each week; adults with chronic conditions or disabilities should meet these recommendations as they are able to “avoid inactivity” (U.S. Department of Health and Human Services (HHS) & Office of Disease Prevention and Health Promotion, 2018). Although current guidelines state that PA benefits for older adults include reduced risk of fall-related injuries, the literature suggests a somewhat more complicated relationship (U.S. Department of Health and Human Services (HHS) & Office of Disease Prevention and Health Promotion, 2018). In the Osteoporotic Fractures in Men Study, age modified the association between PA and falls (Cauley et al., 2013). In men <80 years, those with the lowest energy expenditure had a lower risk of falling; in men >80 years, those with the lowest energy expenditure had a higher risk of falling (Cauley et al., 2013). Absolute differences in PA by fall status, while statistically significant, were small (Cauley et al., 2013). In the Baltimore Longitudinal Study of Aging, there were no differences in total daily PA between fallers and nonfallers (Nastasi et al., 2018). “Risky” fallers, defined as participants who fell while riding a bike, hiking, doing yard work, etc., had higher PA, better balance, and better PF than “normal” fallers, which emphasizes the importance of functional limitations (Nastasi et al., 2018). In the WHI observational study, increases in PA over time were associated with increased fall risks among women aged 50–79 years (Bea et al., 2017). Likewise, in the Hispanic Established Populations for Epidemiologic Study of the Elderly study, high levels of PA with low levels of PF were associated with increased fall risk among older (mean age 82.3 years ± 4.6) Mexican American adults over time (Lewis et al., 2016). In our study of adult women (65.1 years ± 2.7), poor PF increased the risk of falls for women. Women who increased PA participation over time did not have increased fall risk, even when adjusting for PF. Our work is consistent with more recent work from the WHI that suggests falls are not more common for older adult women who participate in higher levels of moderate- to vigorous-intensity PA (Buchner et al., 2017). These findings underscore the importance of PA engagement throughout the life course, even for women with PF limitations.

PA has numerous health benefits, including the slowing of aging-related declines in PF (Dugan et al., 2018; Manini & Pahor, 2008). Women report more disability than men at every age, but women who maintain at least moderate PA engagement demonstrate better PF than those who maintain low PA (Freedman et al., 2016; Gabriel et al., 2017). PF maintenance throughout the life course is important, particularly for women. In our study, phenotypes with low PF, regardless of PA level, had poorer health than phenotypes with high PF. Women in both the high PA and low PF and the low PA and low PF groups had disproportionately higher obesity, diabetes, depressive symptoms, and osteoarthritis at both time points.

PF limitations are dynamic, and the critical midlife period for women is characterized by changes in functional limitations, suggesting that adults can experience both improvement and worsening of functional status (Ylitalo et al., 2013). PF has been posited as the “sixth vital sign” and easy-to-use assessments like the self-reported SF-12 may be useful in the clinical setting to identify older adults at high fall risk (Bierman, 2001; Richardson et al., 2012). Our findings suggest that declines in PF—and specifically, a categorical change from no limitations in PF to some or substantial limitations in PF on the SF-12 questionnaire—are significantly associated with fall risk, regardless of PA engagement. This finding provides further support for evaluation of PF to identify those at higher risk of falls. Based on our results, poor PF, even if coupled by high PA engagement, increases risk of falls. Thus, while PA is beneficial and should continue to be encouraged, consideration of both PA and PF is critical among aging populations. Furthermore, we observed that women who increase PA over time—for example, begin to meet PA guidelines—do not have a greater fall risk in early old age women, even after accounting for PF limitations. Our findings are consistent with Buchner et al. which demonstrated that women with high PA were not more likely to fall (Buchner et al., 2017). Thus, even in individuals with PF limitations, PA should be encouraged, consistent with recommendations to participate in activities according to abilities and to limit inactivity (U.S. Department of Health and Human Services (HHS) & Office of Disease Prevention and Health Promotion, 2018).

This study has several limitations. First, although short-term recall of fall events is moderately accurate, our use of a 12-month recall period may underestimate the incidence of falls. Future work should consider more frequent fall calendars to increase the recall of fall events. We focused our work on intrinsic risk factors for falls, yet a fall is the result of intrinsic and extrinsic factors (Institute of Medicine, 1992). Accordingly, future work should consider fear of falling and fall circumstances to measure situational and/or environmental hazards. We have a robust set of covariates, but we may have unmeasured confounding. Although we included several measures of fall outcomes including recurrent falls, we were limited in our statistical power to detect differences in recurrent falls across the four phenotypes. Our main analyses used self-reported PA because only self-reported data were available at both time points of interest. Although we used meaningful categorizations, we also conducted a sensitivity analysis with PA and PF as continuous variables and found similar results. Compared to the original baseline cohort, participants excluded from our analytic sample were more likely to be Hispanic or African American and have a high school education or less. Finally, future work should include a closer examination of device-based PA, specifically because devices can distinguish intensity level of activity.

Conclusion

PA, including regular exercise, delays or mitigates age-related functional declines. Even among women with poor PF, increasing PA does not appear to increase the risk of falls among late-midlife and early old-age women, suggesting that women can engage in PA to the extent they are able. Falls are a major public health concern, particularly for older adult women. Given the widely appreciated benefits of PA on many age-related health domains, including long-term benefits for physical function, women should be encouraged to engage in PA throughout the life course.

Supplemental Material

sj-pdf-1-jah-10.1177_0898264320988405 – Supplemental Material for Association of Physical Activity and Physical Functioning Phenotypes With Fall Risk Among Women

Supplemental Material, sj-pdf-1-jah-10.1177_0898264320988405 for Association of Physical Activity and Physical Functioning Phenotypes With Fall Risk Among Women by Kelly R. Ylitalo, Carrie A. Karvonen-Gutierrez, Barbara Sternfeld and Kelley Pettee Gabriel in Journal of Aging and Health

Footnotes

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.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: 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, U01AG012495). The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of the NIA, NINR, ORWH, or NIH. K.R.Y. is supported by NIA under Award Number K01AG058754. C.K.G. is supported by NIA under Award Number K01AG054615.

ORCID iD

Kelly R. Ylitalo

Supplemental Material

Supplemental material for this article is available online.

References

Ainsworth

B. E.

Sternfeld

Richardson

M. T.

Jackson

(2000). Evaluation of the kaiser physical activity survey in women. Medicine & Science in Sports & Exercise, 32(7), 1327-1338. doi:10.1097/00005768-200007000-00022

Ainsworth

B. E.

Haskell

W. L.

Herrmann

S. D.

Meckes

Bassett

D. R.

Tudor-Locke

Greer

J. L.

Vezina

Whitt-Glover

M. C.

Leon

A. S.

(2020). Compendium of physical activities tracking guide. Healthy Lifestyles Research Center, College of Nursing & Health Innovation, Arizona State University. https://sites.google.com/site/compendiumofphysical activities/

Bea

J. W.

Thomson

C. A.

Wallace

R. B.

Seguin

R. A.

Going

S. B.

LaCroix

Eaton

Ockene

J. K.

LaMonte

M. J.

Jackson

Jerry Mysiw

Wactawski-Wende

(2017). Changes in physical activity, sedentary time, and risk of falling: The Women’s health initiative observational study. Preventive Medicine, 95, 103-109. doi:10.1016/j.ypmed.2016.11.025

Bierman

A. S.

(2001). Functional status. Journal of General Internal Medicine, 16(11), 785-786. doi:10.1111/j.1525-1497.2001.10918.x

Brazier

J. E.

Harper

Jones

N. M.

O’Cathain

Thomas

K. J.

Usherwood

Westlake

(1992). Validating the SF-36 health survey questionnaire: New outcome measure for primary care. Bmj: British Medical Journal, 305(6846), 160-164. doi:10.1136/bmj.305.6846.160

Buchner

D. M.

Rillamas-Sun

Lamonte

M. J.

Marshall

S. W.

Hunt

Zhang

Rosenberg

D. E.

Lee

I. M.

Evenson

K. R.

Herring

A. H.

Lewis

C. E.

Stefanick

M. L.

LaCroix

A. Z.

(2017). Accelerometer-measured moderate to vigorous physical activity and incidence rates of falls in older women. Journal of the American Geriatrics Society, 65(11), 2480-2487. doi:10.1111/jgs.14960

Cauley

J. A.

Harrison

S. L.

Cawthon

P. M.

Ensrud

K. E.

Danielson

M. E.

Orwoll

Mackey

D. C.

(2013). Objective measures of physical activity, fractures and falls: The osteoporotic fractures in men study. Journal of the American Geriatrics Society, 61(7), 1080-1088. doi:10.1111/jgs.12326

Centers for Disease Control and Prevention . (2013). Trends in meeting the 2008 physical activity guidelines, 2008—2018 percentage (95% confidence interval). https://www.cdc.gov/physicalactivity/downloads/trends-in-the-prevalence-of-physical-activity-508.pdf

Cruz-Jentoft

A. J.

Landi

Schneider

S. M.

Zuniga

Arai

Boirie

Chen

L. K.

Fielding

R. A.

Martin

F. C.

Michel

J. P.

Sieber

Stout

J. R.

Studenski

S. A.

Vellas

Woo

Zamboni

Cederholm

(2014). Prevalence of and interventions for sarcopenia in ageing adults: A systematic review. Report of the international sarcopenia initiative (EWGSOP and IWGS). Age and Ageing, 43(6), 748-759. doi:10.1093/ageing/afu115

10.

Dugan

S. A.

Gabriel

K. P.

Lange-Maia

B. S.

Karvonen-Gutierrez

(2018). Physical activity and physical function. Obstetrics and Gynecology Clinics of North America, 45(4), 723-736. doi:10.1016/j.ogc.2018.07.009

11.

Freedman

V. A.

Wolf

D. A.

Spillman

B. C.

(2016). Disability-free life expectancy over 30 years: A growing female disadvantage in the US population. American Journal of Public Health, 106(6), 1079-1085. doi:10.2105/ajph.2016.303089

12.

Gabriel

K. P.

Sternfeld

Colvin

Stewart

Strotmeyer

E. S.

Cauley

J. A.

Dugan

Karvonen-Gutierrez

(2017). Physical activity trajectories during midlife and subsequent risk of physical functioning decline in late mid-life: The study of women’s health across the nation (SWAN). Preventive Medicine, 105, 287-294. doi:10.1016/j.ypmed.2017.10.005

13.

Institute of Medicine (US) Division of Health Promotion and Disease Prevention (1992). The Second Fifty Years: Promoting Health and Preventing Disability. Berg

R. L.

Cassells

J. S.

(Eds.). Washington, DC: National Academies Press (US). 15, Falls in Older Persons: Risk Factors and Prevention. Available from: https://www.ncbi.nlm.nih.gov/books/NBK235613/

14.

Lewis

Z. H.

Markides

K. S.

Ottenbacher

K. J.

Al Snih

(2016). The Role of physical activity and physical function on the risk of falls in older Mexican Americans. Journal of Aging and Physical Activity, 24(3), 342-349. doi:10.1123/japa.2015-0150

15.

Manini

T. M.

Pahor

(2008). Physical activity and maintaining physical function in older adults. British Journal of Sports Medicine, 43(1), 28-31. doi:10.1136/bjsm.2008.053736

16.

Nastasi

A. J.

Ahuja

Zipunnikov

Simonsick

E. M.

Ferrucci

Schrack

J. A.

(2018). Objectively measured physical activity and falls in well-functioning older adults. American Journal of Physical Medicine & Rehabilitation, 97(4), 255-260. doi:10.1097/phm.0000000000000830

17.

Radloff

L. S.

(1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385-401. doi:10.1177/014662167700100306

18.

Reiner

Niermann

Jekauc

Woll

(2013). Long-term health benefits of physical activity – a systematic review of longitudinal studies. BMC Public Health, 13(1). doi:10.1186/1471-2458-13-813

19.

Richardson

Letts

Chan

Officer

Wojkowski

Oliver

Moore

McCarthy

Price

Kinzie

(2012). Monitoring physical functioning as the sixth vital sign: Evaluating patient and practice engagement in chronic illness care in a primary care setting--a quasi-experimental design. BMC Family Practice, 13(1). doi:10.1186/1471-2296-13-29

20.

Rose

M. S.

Koshman

M. L.

Spreng

Sheldon

(1999). Statistical issues encountered in the comparison of health-related quality of life in diseased patients to published general population norms. Journal of Clinical Epidemiology, 52(5), 405-412. doi:10.1016/s0895-4356(99)00014-1

21.

Sowers

Jannausch

M. L.

Gross

Karvonen-Gutierrez

C. A.

Palmieri

R. M.

Crutchfield

Richards-Mccullough

(2006). Performance-based physical functioning in African-American and caucasian women at midlife: Considering body composition, quadriceps strength, and knee osteoarthritis. American Journal of Epidemiology, 163(10), 950-958. doi:10.1093/aje/kwj109

22.

Sowers

M. R.

Crutchfield

Richards

Wilkin

M. K.

Furniss

Jannausch

Zhang

Gross

(2005). Sarcopenia is related to physical functioning and leg strength in middle-aged women. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 60(4), 486-490. doi:10.1093/gerona/60.4.486

23.

Sowers

Crawford

Sternfeld

Morganstein

Gold

Greendale

Evans

Neer

Matthews

Sherman

Weiss

Kelsey

(2000). SWAN: A multicenter, multiethnic, community-based cohort study of women and the menopausal transition. In Lobo

R. A.

Kelsey

Marcus

(Eds.), Menopause: Biology and pathobiology (pp. 175-188). San Diego: Academic Press.

24.

Steffl

Bohannon

R. W.

Sontakova

Tufano

J. J.

Shiells

Holmerova

(2017). Relationship between sarcopenia and physical activity in older people: A systematic review and meta-analysis. Clinical Interventions in Aging, 12, 835-845. doi:10.2147/cia.s132940

25.

U.S. Department of Health and Human Services (HHS), & Office of Disease Prevention and Health Promotion (2018). Physical activity guidelines for Americans (2nd ed.). Washington, DC: HHS.

26.

Ware

J. E.

Jr Sherbourne

C. D.

(1992). The MOS 36-ltem short-form health survey (SF-36). Medical Care, 30(6), 473-483.

27.

Wee

H.-L.

Thumboo

Lee

Tai

E. S.

(2010). Association of body mass index with short-form 36 physical and mental component summary scores in a multiethnic Asian population. International Journal of Obesity, 34(6), 1034-1043. doi:10.1038/ijo.2010.24

28.

WHO Expert Consultation (2004). Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. The Lancet, 363(9403), 157-163. doi:10.1016/s0140-6736(03)15268-3

29.

Woo

Leung

Kwok

(2007). BMI, body composition, and physical functioning in older adults. Obesity, 15(7), 1886-1894. doi:10.1038/oby.2007.223

30.

Ylitalo

K. R.

Karvonen-Gutierrez

C. A.

(2016). Body mass index, falls, and injurious falls among U.S. adults: Findings from the 2014 behavioral risk factor surveillance system. Preventive Medicine, 91, 217-223. doi:10.1016/j.ypmed.2016.08.044

31.

Ylitalo

K. R.

Karvonen-Gutierrez

C. A.

Fitzgerald

Zheng

Sternfeld

El Khoudary

S. R.

Harlow

S. D.

(2013). 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. Annals of Epidemiology, 23(7), 401-408. doi:10.1016/j.annepidem.2013.04.008

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.31 MB