Examining the relationship between D₃Cr muscle mass, physical performance, and functional limitation in postmenopausal women

Abstract

Background

The D₃-creatine (D₃Cr) dilution method provides a non-invasive and accurate assessment of muscle mass. Our objective was to examine the relationship between D₃Cr muscle mass, objective measures of physical performance, and self-report physical function postmenopausal women.

Methods

We enrolled 2382 women from the Women’s Health Initiative cohort to complete the D₃Cr dilution protocol. The primary exposure variable is D₃Cr muscle mass, examined as an absolute value in kilograms (kg) and relative to total body weight (D₃Cr muscle mass/body mass in kg). Tertiles of absolute D₃Cr muscle mass and D₃Cr muscle mass/body mass were examined. The primary outcomes were objective measures of physical performance, including balance, gait speed, chair stands, and hand grip strength. The short physical performance battery (SPPB) provided an overall measure of physical performance. Physical function and functional limitation were assessed via self-report questionnaire. The association between D₃Cr muscle mass kg⁻¹ (in tertiles) and functional limitation was assessed in multivariate logistic regression analyses.

Results

The mean age of the study sample was 84.5 years. Average D₃Cr muscle mass in the study population was 17.6 kg (SD = 4.0) and 26.1% (SD = 5.8). Mean SPPB scores were higher across higher tertiles of D₃Cr muscle mass (7.0 in T1, 7.8 in T3). Women with low muscle mass were more likely to self-report prevalent limitation in physical function.

Conclusions

In this large cross-sectional study of older women, there was a strong relationship between low D₃Cr muscle mass, physical performance, and functional limitation.

Introduction

Decline in muscle mass is one of the most striking and consistent physical changes associated with aging, particularly in postmenopausal women.^1,2 In older adults, progressive loss of muscle mass may result in impaired physical performance, difficulty completing activities of daily living (ADL), sarcopenia, and frailty.^3–7 Clinical diagnosis of low muscle mass is a sentinel indicator of functional decline and is strongly associated with adverse outcomes, including falls, hospitalization, frailty, disability, and mortality.^6,8–10 Older women are at higher risk of falls, fracture, disability, and hospitalization than similarly aged men due to postmenopausal acceleration of muscle loss and other changes in body composition.¹¹

The introduction of the D₃-creatine (D₃Cr) dilution method has transformed our ability to measure muscle mass in large-scale research studies.^12–14 The D₃Cr dilution method provides a noninvasive measure of skeletal muscle mass using a single urine sample. Until recently, measurement of muscle mass required the use of clinical imaging modalities, such as computed tomography(CT) or magnetic resonance imaging (MRI).^15,16 Dual-energy X-ray absorptiometry and bioelectrical impedance analysis are commonly used to approximate muscle mass, but both rely on indirect estimates of lean tissue as a proxy for muscle mass rather than quantifying skeletal muscle mass directly.¹⁷ The D₃Cr method provides a direct estimate of total body skeletal muscle mass by using isotope dilution to quantify an individual’s creatine pool size using a validated algorithm.¹⁸

Prior research has demonstrated that D₃Cr muscle mass is strongly associated with clinically meaningful health outcomes in older adults.¹⁵ In the Osteoporotic Fractures in Men (MrOS) cohort, low D₃Cr muscle mass was associated with poor physical performance including slower gait speed, lower short physical performance battery (SPPB) scores, mobility limitation, injurious falls, disability, hip fracture, and all-cause mortality in older men (n = 1382; mean age = 84.2 years).^19–22 The Study of Muscle, Mobility, and Aging (SOMMA; n = 359 men and 520 women) reported an association between D₃Cr muscle mass and objective measures of muscle function, including leg strength, peak power, walking speed, and cardiorespiratory fitness.²³ Additionally, a narrative review by Pagano et al.¹⁵ highlighted the strong association between D₃Cr and clinical outcomes.

Collectively, this prior research demonstrates that D₃Cr muscle mass is a clinically meaningful biomarker of functional capacity in older adults. However, there may be important sex differences in the association between D₃Cr muscle mass and functional outcomes to explore. For example, in SOMMA, multivariable adjusted estimates of the association between D₃Cr muscle mass and self-reported mobility limitation and functional limitation were weaker in women than men.^24,25 Much of the current evidence base for D₃Cr muscle mass is derived from studies of older men, with comparatively limited data in women. There is a need for research on D₃Cr muscle mass in postmenopausal women, particularly in the old (>65) and oldest-old (>85 year) age groups. Older women are a high-risk group for age-related decline in skeletal muscle mass.^1,26 Despite this, women remain underrepresented in studies using D₃Cr muscle mass and the relationship between D₃Cr muscle mass and functional outcomes in older women is not well-characterized.¹⁵ Critical gaps persist in understanding how low muscle mass translates into clinically relevant impairment in women. Research is needed to inform sex-specific risk stratification and optimize strategies to preserve function and independence among aging women.

In this study, we used the D₃Cr dilution method to measure D₃Cr muscle mass in a large sample of postmenopausal women from the Women’s Health Initiative (WHI) study. The goal of this study was to examine the cross-sectional relationship between D₃Cr muscle mass, objective measures of physical performance, and self-report physical function and functional limitation in postmenopausal women. We hypothesize that D₃Cr muscle mass will be positively associated with higher scores on objective measures of physical performance (eg, greater handgrip strength and faster gait speed) and lower likelihood of self-reported functional limitations in postmenopausal women.

Methods

Study population

The WHI is a landmark study of women’s health that has been ongoing since the 1990s. Originally comprised of 161 808 postmenopausal women recruited from across the United States, participants have been followed continuously since study baseline.^27–29 Since 2005, participants have completed annual questionnaires. Between 2023 and 2025, the WHI Clinical Coordinating Center invited a sub-set of WHI participants to complete a home visit administered by trained examiners. A standardized data collection protocol was used for all home visit participants. Recruitment for the home visit cohort included WHI hormone trial participants, all Black/African American and Hispanic women, and cancer survivors from the LILAC cohort.³⁰ All WHI participants who completed a home visit were invited by mail to participate in this study of D₃Cr muscle mass.³⁰ The median number of days between the WHI home visit and D₃Cr study was 147 days (SD = 100). Participants provided informed consent to participate in the WHI home visit and D₃Cr muscle mass study.

Measurement of D₃Cr muscle mass

The D₃Cr dilution method can be used to estimate muscle mass in any setting, either in person or remotely.^19,30,31 Participants ingest a capsule with stable isotope-labeled creatine and provide a fasting urine sample 2-4 days later. Using high performance liquid chromatography and mass spectrometry, measurement of urinary creatinine enrichment provides a measure of whole-body creatine pool size.³² Whole-body creatine is proportional to skeletal muscle mass.³³ A validated algorithm is used to quantify total body skeletal muscle mass (D₃Cr muscle mass) using ratio of “labeled” D₃-creatinine to “unlabeled” creatinine in urine.^13,19 The D₃Cr dilution method is safe and there are no known risks associated with D₃-creatine. The dose of creatine used in the D₃-creatine dilution method is far less than the amount typically consumed from dietary sources or nutritional supplements.^34,35 Validity and reliability of the D₃Cr dilution method analytic protocol have been established in prior research. The coefficient of variation for D₃Cr muscle mass is reported as 3%-5%, demonstrating good reliability.¹⁸ Validity has been established through strong correlations with gold-standard imaging techniques, such as MRI and CT (r = 0.9).¹³

In this study, D₃Cr muscle mass was measured using a novel fully remote collection approach following the WHI in home visit.³⁰ Briefly, participants were sent a study kit including a 25 mg oral dose of D₃Cr, a urine collection cup and a retractable, filter-paper dipstick. Participants were instructed to take the D₃Cr capsule on a specific date and provide a fasting, morning urine sample 2-4 days later. They returned the dipstick to the study team via FedEx overnight shipping in a cold, temperature-controlled shipping envelope with ice packs. Upon receipt, the study team stored each dipstick in a −20 °C freezer. Dipsticks were sent in batches for liquid chromatography–mass spectrometry (LC/MS) analysis at the University of California, Berkeley.

Absolute D₃Cr muscle mass is reported in kilograms (kg). We also report D₃Cr muscle mass relative to total body weight (D₃Cr kg⁻¹). This allows for comparison of D₃Cr muscle mass across individuals of different body weight.^19,36 For interpretability and clinical utility, both absolute (kg) and relative (D₃Cr kg⁻¹) muscle mass will be presented in 3 categories (tertiles) in the main analysis.

Outcome measurements

Objective measures of physical performance

During the WHI home visit, physical performance was measured via the SPPB and grip strength test.^37–39 The SPPB consists of 3 assessments: balance, gait speed, and chair stand.³⁷ Gait speed is assessed through a timed 3- or 4-m walk. The chair stand test measures time to rise from a chair five times without using arms. The balance component includes assessment of static balance through 3 progressively challenging tests: (1) side by side, (2) semi-tandem, and (3) tandem. Each position must be held for a pre-specified amount of time, before moving on to the subsequent balance test. Participants who completed the tandem stand were also asked to do a 30-s one leg balance test. Each component is scored from 0 to 4, with a total score ranging from 0 (worst performance) to 12 (best performance).⁴⁰ Handgrip strength was assessed during home visits using a standardized protocol with a calibrated handheld dynamometer.^41–43 Two trials were recorded for each hand, with the highest value used for the analysis. Average grip strength results are presented in the results, consistent with prior research on D₃Cr in older men.¹⁹

WHI self-reported measures of physical function and functional limitation

Annual WHI questionnaires include questions related to physical function, ability to complete basic and instrumental ADL, and sedentary behavior.^40,44 Physical function was measured via the physical function subscale of the RAND-36 questionnaire.⁴⁵ Higher scores on this scale indicate better physical function.⁴⁶ Activities of daily livings include self-care activities and complex everyday tasks.⁴⁷ Sedentary behavior was ascertained from questions asking participants to indicate the amount of daily time spent sitting while watching television; at the computer; reading or doing crafts; talking on the phone or texting; and sitting in a car, bus, train, or other mode of transport.

Measurement of covariates

The WHI dataset includes measures of sociodemographic, lifestyle, and health-related characteristics. In this analysis, we examined covariates measured at WHI baseline (1993-1998), in self-report WHI follow-up assessments from baseline to present, and during the WHI home visit.

Statistical analysis

This is a cross-sectional analysis of the relationship between D₃Cr muscle mass, objective measures of physical performance from the WHI home visit, and self-report measures of physical function. The timeline of data collection was: WHI in-home visit to collect objective measures of physical function; immediately following the in-home visit, participants were invited to complete the remote assessment of D₃Cr muscle mass; and finally, self-report measures of physical function and functional limitation were collected from WHI questionnaires completed in the same year. In this analysis, we used the self-report measures obtained from participants in the annual WHI data collection cycle immediately following home visit and D₃Cr muscle mass assessment.

Initial descriptive analyses examined demographic characteristics, anthropometric variables, objective measures of physical performance, and self-report physical function in the overall study cohort and according to tertile of D₃Cr muscle mass. Complete case analysis was used due to the low proportion of missing data. Missing data were assumed to be missing at random. We computed age-adjusted p-values for trend across D₃Cr muscle mass tertiles for each variable in our descriptive analyses, adjusting for age at the time of D₃Cr muscle mass measurement. We estimated prevalence ratios (PRs) and 95% confidence intervals for binary outcomes using modified Poisson regression to examine the relationship between D₃Cr muscle mass and measures of physical limitation.⁴⁸ Multivariable adjustment was used to adjust for confounding by age, race/ethnicity, and cancer status. We compared two different models for each outcome: model 1 with absolute D₃Cr muscle mass (kg) as the exposure and model 2 with relative D₃Cr muscle mass (D₃Cr kg⁻¹) as the exposure. We also expressed the effect size in terms of per SD increment. All effects are expressed as PR along with their 95% confidence intervals. All analyses above were conducted using R Studio version 2022.07.1.

Results

D₃Cr muscle mass measurements were obtained from 2382 women. Participants ranged in age from 74.9 years to 101.5 years (mean 84.5, SD 4.6 years, Table 1). In the total sample, 1988 (83.5%) women reported their race as white and 6.8% reported being Hispanic or Latina. Overall, mean D₃Cr muscle mass in the study population was 17.6 kg (SD = 4.0) and 26.1% (SD = 5.8). We created D₃Cr tertiles from the data, representing low (T1), moderate (T2), and high muscle mass (T3) related to body weight, expressed as a percentage. Women in T1 were older than women in T3. There are notable differences in the distribution of muscle mass according to race, with a greater proportion of Black women in the highest versus lowest tertile of muscle mass. Women with low muscle mass had higher BMI (28.8 kg m⁻²) than women in with high muscle mass (23.3 kg m⁻²), and there were many more women with obesity in T1 versus T3 (36.6% vs 5.9%).

Table 1.

Characteristics of WHI participants overall and stratified by tertiles of D₃Cr muscle mass per body weight (D₃Cr kg⁻¹).

	Total (n = 2382)	T1 (<0.231, n = 794)	T2 (≥0.231-< 0.277, n = 794)	T3 (≥0.277, n = 794)	p-trend
Age at visit (years), M (SD)	84.5 (4.6)	85.3 (4.9)	84.3 (4.4)	83.9 (4.3)	<.001
Ethnicity, n (%)					.551
Not Hispanic/Latino	2219 (93.2)	740 (93.2)	735 (92.6)	744 (93.7)
Hispanic/Latino	162 (6.8)	54 (6.8)	59 (7.4)	49 (6.2)
Race, n (%)					.012
Black	291 (12.2)	78 (9.8)	87 (11.0)	126 (15.9)
White	1988 (83.5)	677 (85.3)	668 (84.1)	643 (81.0)
BMI (kg m−2), M (SD)	26.0 ± 5.5	28.8 (6.0)	25.8 (4.5)	23.3 (4.3)	<.001
Height (cm), M (SD)	162.9 (6.5)	162.5 (6.7)	162.9 (6.4)	163.3 (6.4)	.140
Weight (kg), M (SD)	68.8 (14.2)	76.0 (15.4)	68.3 (12.2)	62.0 (11.1)	<.001
Waist circumference (inches), M (SD); n = 2203	35.6 (5.3)	37.4 (5.4)	35.6 (4.8)	33.8 (5.0)	<.001
Alcohol servings per week, M (SD)	2.4 (4.4) (n = 1633)	2.3 (4.2) (n = 572)	2.4 (4.2) (n = 555)	2.7 (4.9) (n = 506)	.177
Current smoking, n (%)	47 (2.0)	12 (1.5)	21 (2.6)	14 (1.8)	.312
Use of assistive devices, n (%)					<.001
None	1447 (60.7)	391 (49.2)	506 (63.7)	550 (69.3)
Cane	270 (11.3)	118 (14.9)	87 (11.0)	65 (8.2)
Walker	180 (7.6)	103 (13.0)	42 (5.3)	35 (4.4)
Wheelchair	9 (0.4)	5 (0.6)	2 (0.3)	2 (0.3)
Kyphosis assessment, n (%)					<.001
None	823 (34.6)	241 (30.4)	265 (33.4)	317 (39.9)
Slight	1043 (43.8)	334 (42.1)	372 (46.9)	337 (42.4)
Moderate	325 (13.6)	144 (18.1)	95 (12.0)	86 (10.8)
Severe	35 (1.5)	17 (2.1)	16 (2.0)	2 (0.3)

Table 2 presents associations between measures of D₃Cr muscle mass and objective measures of physical performance from WHI home visits. Overall, women with high muscle mass performed better on measures of physical performance than women with low muscle mass, including measures of strength and balance. High muscle mass is associated with greater hand grip strength and lower extremity strength measured by timed walk and chair stand time. Women with high muscle mass were able to hold static balance tests (semi-tandem and tandem positions) longer than women with low muscle mass. There is evidence of a dose response relationship between SPPB score and relative muscle mass: average SPPB scores were 7.0, 7.8, and 8.2 for women in T1, T2, and T3, and an age adjusted p-for-trend test was <.001 across tertiles of the exposure. Figure S1A-C also illustrate these relationships.

Table 2.

Cross-sectional relationship between objective measures of physical performance and tertiles of D₃Cr muscle mass per body weight (D₃Cr kg⁻¹) in kilograms.

	Total (n = 2382)	T1 (n = 794)	T2 (n = 794)	T3 (n = 794)	p-trend
Average grip strength (kg), M (SD); n = 2194	13.8 (6.0)	12.7 (5.7)	14.1(5.8)	14.6 (6.3)	<.001
Side by side balance test time (s), M (SD); n = 2008	9.8 (1.0)	9.8 (0.9)	9.8 (1.0)	9.8 (1.0)	.985
Semi-tandem balance test time (s), M (SD); n = 1817	28.1 (5.8)	27.6 (6.4)	28.3 (5.2)	28.2 (5.6)	.220
Average tandem stand balance test time (s), M (SD); n = 301	13.1 (7.9)	12.3 (8.3)	13.4 (7.9)	13.7 (7.4)	.519
Average one leg balance test time (s), M (SD); n = 395	11.2 (7.3)	8.9 (6.9)	12.2 (7.6)	12.0 (7.1)	.001
Average timed walk test time (s), M (SD); n = 1962	7.6 (7.1)	8.3 (8.5)	7.6 (7.5)	7.0 (4.9)	.004
Repeated chair stand time (s), M (SD); n = 1867	15.8 (9.0)	16.6 (8.9)	15.9 (9.4)	14.9 (8.6)	.007
SPPB overall score (0-12), M (SD); n = 1800	7.9 (2.6)	7.2 (2.7)	8.0 (2.5)	8.4 (2.5)	<.001

Abbreviation: SPPB, short physical performance battery.

There was a positive association between muscle mass and self-report physical function (Table 3). Women with high muscle mass had higher mean scores than women with low muscle mass (RAND-36 score 72.6 vs 54.8). This pattern persisted across all self-report measures of physical function. Women with low muscle mass were more likely to report limitations in vigorous activities, moderate activities, carrying groceries, climbing several flights of stairs, climbing one flight of stairs, bending, walking a mile, walking several blocks, bathing, or dressing. All p-for-trend results were less than .001. Results examining the ability to complete instrumental ADL are reported in the Supplementary Material (Table S2). Sedentary behavior was inversely associated with muscle mass: spending >2 h daily engaging in sedentary behavior was associated with low muscle mass.

Table 3.

Cross-sectional description of tertiles of D₃Cr kg⁻¹ and self-report measures of physical function and sedentary behavior.

	Total (n = 2382)	T1 (<0.231, n = 794)	T2 (≥0.231-<0.277, n = 794)	T3 (≥0.277, n = 794)	p-trend
RAND-36 Physical functioning construct, M (SD)	64.4 (26.0)	54.8 (26.8)	65.8 (24.1)	72.6 (24.0)	<.001
Self-report limitations, n (%)
Vigorous activities	2187 (91.8)	741 (93.3)	744 (93.7)	702 (88.4)	.001
Moderate activities	1377 (57.8)	548 (69.0)	455 (57.3)	374 (47.1)	<.001
Carrying groceries	878 (36.9)	386 (48.6)	280 (35.3)	212 (26.7)	<.001
Several flights of stairs	1587 (66.6)	627 (79.0)	534 (67.3)	426 (53.7)	<.001
One flight of stairs	842 (35.3)	392 (49.4)	254 (32.0)	196 (24.7)	<.001
Bending	1462 (61.4)	558 (70.3)	492 (62.0)	412 (51.9)	<.001
Walking >1 mile	1535 (64.4)	618 (77.8)	508 (64.0)	409 (51.5)	<.001
Walking several blocks	1152 (48.4)	502 (63.2)	366 (46.1)	284 (35.8)	<.001
Bathing or dressing	208 (8.7)	105 (13.2)	56 (7.1)	47 (5.9)	<.001
Sedentary time (hours/day), n (%)
Watching TV					<.001
None	70 (2.9)	19 (2.4)	22 (2.8)	29 (3.7)
Up to 2 h	695 (29.2)	178 (22.4)	243 (30.6)	274 (34.5)
More than 2 h	1093 (45.9)	407 (51.3)	355 (44.7)	331 (41.7)
Sitting at computer					.870
None	339 (14.2)	132 (16.6)	110 (13.9)	97 (12.2)
Up to 2 h	1284 (53.9)	373 (47.0)	434 (54.7)	477 (60.1)
More than 2 h	229 (9.6)	96 (12.1)	77 (9.7)	56 (7.1)
Reading, listening to music					.044
None	134 (5.6)	44 (5.5)	48 (6.0)	42 (5.3)
Up to 2 h	1357 (57.0)	415 (52.3)	450 (56.7)	492 (62.0)
More than 2 h	351 (14.7)	142 (17.9)	116 (14.6)	93 (11.7)
Talking on phone/texting					.253
None	102 (4.3)	26 (3.3)	35 (4.4)	41 (5.2)
Up to 2 h	1719 (72.2)	562 (70.8)	576 (72.5)	581 (73.2)
More than 2 h	28 (1.2)	11 (1.4)	8 (1.0)	9 (1.1)
Sitting on car/bus/train					.092
None	335 (14.1)	128 (16.1)	112 (14.1)	95 (12.0)
Up to 2 h	1508 (63.3)	478 (60.2)	499 (62.8)	531 (66.9)
More than 2 h	23 (1.0)	5 (0.6)	11 (1.4)	7 (0.9)

Prevalence ratios from multivariable adjusted logistic regression models provide evidence of a monotonic relationship between relative D₃Cr muscle mass (D₃Cr kg⁻¹) and self-report functional limitation (Table 4). There is a consistent pattern in which lower D₃Cr muscle mass is associated with higher prevalence of self-report limitations. The association is stronger and more consistent when muscle mass is expressed relative to body weight. D₃Cr muscle mass/kg shows a clear dose-response pattern across nearly all outcomes, with T1 having substantially higher prevalence of limitations than T3. Results indicating per-SD increments reinforce this pattern: higher relative muscle mass is associated with lower prevalence of limitations across all outcomes. For each SD increment in relative muscle mass, women are less likely to report functional limitation. Additional models for absolute D₃Cr (kg) and self-report measures of physical function, adjusting for weight or BMI, are presented in Table S3. Adjusting for body weight or BMI increased the magnitude of effect estimates compared to those presented in Table 4 for absolute D₃Cr. Adjustment for additional covariates (Tables S4 and S5) resulted in no change to estimates for either relative or absolute muscle mass.

Table 4.

Adjusted^a PRs for the cross-sectional relationship between D₃Cr muscle mass measures and functional limitations.

	D3Cr muscle mass (kg)	D3Cr muscle mass/kg
Limited moderate activities
T1	1.19 (1.04, 1.36)	1.38 (1.21, 1.58)
T2	1.13 (0.99, 1.30)	1.20 (1.04, 1.37)
T3	1.00 (referent)	1.00 (referent)
Per SD increment	0.94 (0.89, 1.00)	0.86 (0.82, 0.92)
Limited lifting or carrying groceries
T1	1.30 (1.10, 1.55)	1.68 (1.42, 2.00)
T2	1.14 (0.96, 1.36)	1.29 (1.08, 1.55)
T3	1.00 (referent)	1.00 (referent)
Per SD increment	0.91 (0.84, 0.98)	0.79 (0.73, 0.85)
Limited climbing several flights of stairs
T1	1.07 (0.95, 1.22)	1.42 (1.26, 1.61)
T2	1.05 (0.93, 1.19)	1.24 (1.09, 1.41)
T3	1.00 (referent)	1.00 (referent)
Per SD increment	0.98 (0.92, 1.03)	0.85 (0.81, 0.90)
Limited bending, kneeling, stooping
T1	1.05 (0.93, 1.20)	1.32 (1.16, 1.50)
T2	1.05 (0.92, 1.19)	1.19 (1.04, 1.35)
T3	1.00 (referent)	1.00 (referent)
Per SD increment	0.99 (0.94, 1.05)	0.89 (0.84, 0.94)
Limited walking several blocks
T1	1.11 (0.96, 1.29)	1.65 (1.43, 1.92)
T2	1.06 (0.91, 1.22)	1.26 (1.08, 1.48)
T3	1.00 (referent)	1.00 (referent)
Per SD increment	0.98 (0.92, 1.04)	0.80 (0.74, 0.85)
Limited walking more than 1 mile
T1	1.13 (1.00, 1.29)	1.44 (1.27, 1.63)
T2	1.07 (0.94, 1.22)	1.23 (1.08, 1.40)
T3	1.00 (referent)	1.00 (referent)
Per SD increment	0.97 (0.92, 1.03)	0.85 (0.81, 0.90)

Abbreviation: PR, prevalence ratio.

^aMultivariable logistic regression models adjusted for age, race, and cancer status.

Discussion

In this large cross-sectional analysis of postmenopausal women, high D₃Cr muscle mass was positively associated with physical performance, including objective measures of strength, balance, walking speed, and self-report functional limitations.

Multivariable models demonstrated a monotonic association between D₃Cr muscle mass per kg body weight and outcomes. Relative D₃Cr muscle mass per kg body weight appears to be a stronger and more consistent correlate of functional limitation than absolute muscle mass in kilograms. Our results underscore the importance of interpreting muscle mass in relation to overall body weight. This issue is increasingly emphasized in contemporary discussions of sarcopenia and sarcopenic obesity. The Global Leadership Initiative in Sarcopenia (GLIS) conceptual definition identifies muscle mass, muscle strength, and muscle-specific strength as core elements of sarcopenia, with muscle-specific strength reflecting strength relative to muscle size rather than strength or mass in isolation.⁴⁹ This reinforces the idea that muscle quantity alone may be insufficient without consideration of body size and functional capacity.⁴⁹ Similarly, the Sarcopenic Obesity Global Leadership Initiative (SOGLI) consensus on sarcopenic obesity recommends evaluating low muscle mass in the context of excess adiposity.⁵⁰ Individuals with higher body weight may have appear to have adequate absolute muscle mass but insufficient muscle relative to the mechanical and metabolic demands imposed by excess body mass. Our results support the concept that relative muscle deficit may be more clinically informative for identifying impaired physical performance and function, particularly in populations with overweight or obesity. These findings align with emerging consensus that body-weight-adjusted muscle measures may better capture the functional consequences of low muscle mass than absolute muscle mass alone.

Our results provide a comprehensive examination of D₃Cr muscle mass in postmenopausal women. These results are important given that research examining the relationship between lean body mass, physical performance, and physical function in older adults has yielded mixed results. Prior inconsistencies in research findings may be due to measurement error in the measurement of muscle mass.^19,51,52 Dual-energy X-ray absorptiometry lean body mass often used as an approximation of muscle mass, but it does not directly quantify muscle mass. The D₃Cr dilution method directly quantifies skeletal muscle mass. Direct measurement of muscle mass using the D₃Cr dilution method reduces misclassification of muscle mass and stands to improve our understanding of the role of muscle mass in healthy aging. Our findings support the need for research on D₃Cr muscle mass as a clinically relevant biomarker; a target for interventions aimed at preserving physical performance and reducing the risk of physical limitation in older women.

We observed an inverse relationship between sedentary behavior and D₃Cr muscle mass. Sedentary behavior has received increasing attention in the geriatrics literature because accumulating evidence suggests it is independently associated with adverse health outcomes, even among individuals who engage in regular physical activity.⁵³ Prior WHI research has demonstrated that prolonged periods of sitting are associated with increased risk of cardiovascular disease, diabetes, disability, and mortality in postmenopausal women.^54–59 These findings raise the possibility that sedentary time may contribute to loss of skeletal muscle mass through reduced mechanical loading and inactivity, thereby accelerating functional decline. Conversely, low muscle mass may also predispose individuals to sedentary behavior, suggesting a potentially bidirectional relationship. Longitudinal, prospective studies are needed to clarify the temporal nature of these associations and to determine whether reducing sedentary time may help preserve muscle mass and functional independence in older adults.

Our findings are broadly consistent with prior studies of D₃Cr muscle mass in older men, including strong associations between low muscle mass, impaired physical performance, and self-report physical function in MrOS. WHI and MrOS have similar age distributions (mean age 85.3 years in WHI vs 84.2 years in MrOS) but absolute and relative muscle mass were substantially lower in women compared to men.²¹ On average, WHI women had 7 kg less absolute D₃Cr muscle mass (17vs 24 kg), equating to 5% less D₃Cr muscle mass overall. This finding is consistent with known sex differences in body composition. Prior MrOS analyses reported robust associations between D₃Cr muscle mass and multiple domains of function, with somewhat stronger effect sizes than those observed in WHI.^19,20,36 For example, in MrOS, each standard deviation increase in percent D₃Cr muscle mass was associated with lower odds of difficulty getting in or out of bed or chairs (odds ratio, 0.57 [95% CI, 0.47-0.69]), whereas in WHI, a more modest association was observed for a related functional task (bending, kneeling, or stooping; odds ratio, 0.89 [95% CI, 0.82-0.97]). These differences may reflect sex-specific relationships between muscle mass and functional capacity, differences in functional measures, or greater heterogeneity in functional limitation among older women.

There has been significant debate about the relative contributions of muscle quantity (mass) versus function (i.e. grip strength, gait speed) in predicting clinical outcomes in older adults. Recent guidelines for defining sarcopenia have prioritized muscle function over muscle mass with the rationale that function is the clinically relevant construct.^4,60–62 The present results add to literature base demonstrating that muscle mass measured via D₃Cr is also important to consider as a component of sarcopenia definitions, given its strong association with physical performance and functional limitation. Our study adds to prior findings from MrOS and SOMMA. Together, these studies demonstrate that D₃Cr muscle mass may represent an important component of healthy aging and functional status in older adults. Further research is required to investigate the relationship between D₃Cr muscle mass and clinically relevant outcomes, morbidity, and mortality in women.

Measuring muscle mass using D₃Cr: real-world implications

Our findings demonstrate the feasibility of measuring D₃Cr muscle mass remotely in a large cohort of older women. This approach has fundamentally changed our ability to obtain a direct estimate of muscle mass in clinical and population health research. The method is safe, noninvasive, accurate, and only requires a single, fasting urine sample. To date, the D₃Cr dilution method has been used in research settings. In the future, clinical trials could be conducted to examine whether D₃Cr muscle mass measures could be incorporated into routine clinical care in geriatrics. D₃Cr muscle mass could also be incorporated into future iterations of diagnostic criteria for sarcopenia and sarcopenic obesity. Current consensus efforts, including GLIS and SOGLI, emphasize muscle mass and strength as central components of sarcopenia, while recognizing that optimal measurement approaches for muscle mass remain under development. D₃Cr could be incorporated as a complement to current definitions used to quantify muscle mass, as part of risk-stratification approaches, or for the prediction of clinically relevant functional outcomes.

Strengths and limitations

Our study is the largest study of D₃Cr muscle mass in a sample of older women and the largest single-study of D₃Cr conducted to date. Other strengths of the present study include the well-characterized and extensively phenotyped cohort of postmenopausal women who have been followed continuously for ∼30 years; objective measures of physical performance in conjunction with self-report measures of physical function; anthropometric characteristics from the WHI home visit. This study is cross-sectional which precludes inference about temporality. This is a limitation of the present work. The length of time between the in-home visit and D₃Cr measurement may also represent a limitation, as muscle mass and/or physical function could have changed in this period. Additional potential limitations include possible survivor biases as the sample included only those who were still alive and who consented to the home visit and, separately, the D₃Cr dilution method. It is possible that the selection of the study sample introduced a form of selection (collider stratification) bias. The most common reasons for declining study participation were poor health and/or caregiving related responsibilities. There were also women that could not provide informed consent or follow the study protocol due to cognitive decline.

Conclusions

Our findings indicate that in a population of older women, D₃Cr muscle mass is associated with physical performance and functional limitation. This work extends the existing literature on D₃Cr muscle mass, making an important contribution to our understanding of D₃Cr muscle mass in older women. In postmenopausal women, maintaining and increasing muscle mass should be a priority to preserve function, prevent disability, and prolong independence.

Supplementary material

Supplementary material is available at The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences online.

Funding

The WHI program is funded by the National Heart, Lung, and Blood Institute, National Institutes of Health, U.S. Department of Health and Human Services through contracts 75N92021D00001, 75N92021D00002, 75N92021D00003, 75N92021D00004, 75N92021D00005. Ms. LILAC is funded by the National Cancer Institute, National Institutes of Health, U.S. Department of Health and Human Services (5R37CA258761; PI H.R.B).

Conflicts of interest

None declared.

Data availability

WHI data are publicly available on request and proposal approval from the clinical coordinating center.

Author contributions

Hailey R. Banack (Conceptualization [lead], Data curation [lead], Formal analysis [supporting], Funding acquisition [lead], Methodology [lead], Writing—original draft [lead]), Christopher Kim (Formal analysis [supporting], Visualization [supporting], Writing—original draft [supporting]), Jean Wactawski-Wende (Conceptualization [supporting], Funding acquisition [supporting], Investigation [supporting], Methodology [supporting], Writing—review & editing [supporting]), Elizabeth M. Cespedes Feliciano (Conceptualization [supporting], Funding acquisition [supporting], Methodology [supporting], Writing—review & editing [supporting]), Bette J. Caan (Conceptualization [supporting], Funding acquisition [supporting], Methodology [supporting], Writing—review & editing [supporting]), Garnet L. Anderson (Conceptualization [supporting], Data curation [supporting], Investigation [supporting], Methodology [supporting], Writing—review & editing [supporting]), Heather M. Ochs-Balcom (Conceptualization [supporting], Methodology [supporting], Writing—review & editing [supporting]), Catherine Lee (Conceptualization [supporting], Formal analysis [supporting], Writing—review & editing [supporting]), Mahalakshmi Shankaran (Data curation [lead], Investigation [supporting], Validation [supporting]), and William J. Evans (Conceptualization [supporting], Funding acquisition [supporting], Methodology [supporting], Writing—review & editing [supporting])