Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Feb 4.
Published in final edited form as: J Gerontol A Biol Sci Med Sci. 2026 Jan 2;81(1):glaf130. doi: 10.1093/gerona/glaf130

Associations of Accelerometer-Estimated Sleep with Cardiorespiratory Fitness and Energetic Efficiency among Middle-Aged and Older Adults

Daniel D Callow 1,*, Yiwei Yue 2,*, Idiatou Diallo 2, Jill A Rabinowitz 3, Yang An 4, Alfonso J Alfini 5, Mark N Wu 6, Sarah K Wanigatunga 2, Amal A Wanigatunga 7,8, Luigi Ferrucci 4, Eleanor M Simonsick 4, Jennifer A Schrack 7,8,**, Adam P Spira 1,2,8,**
PMCID: PMC12758963  NIHMSID: NIHMS2135381  PMID: 40577079

Abstract

Background:

Sleep disturbances and cardiovascular disease are common and often co-occur in middle-aged and older adults, but less is known about associations of sleep with cardiorespiratory fitness and energy efficiency in these populations. We examined cross-sectional associations of accelerometer-derived sleep metrics with cardiorespiratory fitness, walking energetics, and resting metabolic rate, and explored whether associations were moderated by age, sex, and race.

Methods:

We studied 263 participants from the Baltimore Longitudinal Study of Aging (mean age 72.7±10.1 years, 53.6% women). Predictors included total sleep time (TST), sleep efficiency (SE), sleep onset latency (SOL), wake after sleep onset (WASO), and average wake bout length (WBL). Outcomes included measures of cardiorespiratory fitness (i.e., maximal oxygen consumption (VO2peak)) and energetic efficiency (i.e., energetic cost of walking (ECW) and resting metabolic rate (RMR)).

Results:

After adjusting for demographics, comorbidities, and self-reported physical activity, longer WBL was associated with lower VO2peak (B=−1.01 ml/kg/min, p<0.01) and higher RMR (B=43.25 kcal, p<0.05), lower SE was associated with lower VO2peak (B=1.07 ml/kg/min, p<0.01), and shorter TST was associated with lower VO2peak (B=0.33 ml/kg/min, p<0.05). Higher SE was associated with lower RMR among middle-aged adults but not older adults (interaction p-value<0.05).

Conclusion:

Shorter TST, longer WBL, and lower SE are associated with poorer cardiorespiratory fitness and energetic efficiency among middle-aged and older adults. Longitudinal studies are needed to understand the temporality of these associaitons and potential targets for interventions in these populations.

Keywords: sleep health, physical activity, metabolism, resting metabolic rate, aging

Introduction

Sleep is a basic, vital function for optimal health involved in autonomic nervous system regulation, cardiac function, mitochondrial metabolism, and cellular and tissue restoration 1,2. Middle-aged and older adults face an elevated risk of sleep disturbances, including difficulties in sleep initiation and maintenance 3,4 that are tied to detrimental cardiometabolic risk factors, including chronic inflammation, hypertension, obesity, and diabetes mellitus 5–7. However, our understanding of the relationship between poor sleep and cardiovascular and metabolic health in the context of aging remains limited. Given the burden of sleep-related issues in this demographic and the expanding aging population, it is imperative to better understand associations between sleep and critical cardiometabolic health metrics, such as cardiorespiratory fitness (CRF) and efficiency of energy utilization (“energetic efficiency”).

Aging is linked to widespread changes in cardiovascular processes, including lower cardiorespiratory fitness (CRF) 8,9, which heightens the risk of cardiometabolic disease, cardiovascular disease, functional decline, and all-cause mortality 10,11. Concurrently, aging often entails changes in functional capacity and energetic efficiency, including a higher energetic cost of walking (ECW; i.e., energy expended during overground walking) and changes in resting metabolic rate (RMR; energy required to function at rest) 12,13. Such changes are associated with health and functioning. Notably, increases in the ECW are thought to represent metabolic inefficiencies and predict a decline in usual gait speed among older adults 14, and slower gait speeds are associated with adverse health outcomes 15, including cognitive decline 16, cardiovascular disease 17, and all-cause mortality 18. Additionally, while aging is generally associated with a decline in RMR and poorer metabolic health resulting from loss of lean mass, comorbid conditions may reverse this decline by necessitating additional energy expenditure for recovery and repair, and maintenance of homeostasis 13,19. However, the potential relationships of age-related sleep disturbances with CRF and energetic efficiency remain unclear.

Existing studies examining associations of sleep with CRF or energetic efficiency in middle-aged and older adults suggest that sleep disturbances are associated with elevated overnight RMR20, greater fatigue and higher blood pressure, as well as lower physical activity levels 21. 22Notably, studies linking sleep to metabolism or cardiovascular health have primarily focused on young and middle-aged adults and have found sleep disturbances linked to weight gain, glucose dysregulation, and increased diabetes and cardiovascular risk 7,23. Moreover, previous studies reporting associations between poor sleep and low CRF have predominantly relied on smaller sample sizes24–26 and subjective sleep measures, which are susceptible to recall and social desirability biases 27.

Considering age is a risk factor for sleep disturbances and diminished CRF and energetic efficiency, and recognizing the potential influence of sleep quality and quantity on cardiovascular and metabolic health, a better understanding is needed of the links among various aspects of sleep, CRF, and energy efficiency. Therefore, the primary objective of this study was to examine cross-sectional associations of objective measures of sleep quality and quantity with CRF and energetic efficiency in middle-aged and older adults. We hypothesize that better sleep quality and longer sleep duration will be associated with higher CRF and RMR, as well as lower ECW in our sample. Additionally, because age, sex, and race have been previously linked to sleep and health conditions that may result from poor CRF 28,29, we explored whether these associations differed by sociodemographic characteristics.

Methods

Participants

We studied participants from the Baltimore Longitudinal Study of Aging (BLSA), an observational study established in 1958 by the National Institute on Aging Intramural Research Program to understand health changes that occur with aging 30. The BLSA is an ongoing study that continuously enrolls healthy adults aged 20 years or older and follows them longitudinally; the length of time between follow-ups depends on participant age (<60: every 4 years, 60–79: every 2 years, ≥80: every year).

In this study, we restricted the sample to those aged ≥ 50 years who completed wrist actigraphy, fitness, ECW, and RMR assessments. We excluded 21 participants who had mild cognitive impairment or dementia, based on an adjudication process described in detail elsewhere 31,32. We analyzed the most recent visit at which participants had actigraphy, fitness, ECW, and RMR data, resulting in an analytic sample of 263 participants. The study protocol was approved by the National Institutes of Health Intramural Research Program Institutional Review Board, and each participant provided written informed consent at each visit.

Objective Sleep Measures

Each participant was instructed to wear an accelerometer (Actiwatch-2, Philips-Respironics, Bend, OR) on their non-dominant wrist for a period of seven consecutive days. The device records movement through activity counts via an accelerometer and registers ambient light levels utilizing an integrated photometer. While wearing the accelerometer, participants were asked to complete sleep diaries in both the morning and evening. These diaries facilitated documentation of sleep-related behaviors, such as the duration of time spent in bed in the evening with the intention of sleeping (‘lights out’), the time participants arose in the morning to commence their day, instances of napping, and any instances of accelerometer removal. Participants were instructed to press an event-marker button on the accelerometer device at ‘lights out’ and when getting up to start their day to permit identification of in-bed intervals. Using the Actiware software v. 6.0.9 (Philips-Respironics), a widely used algorithm 33 was applied to derive several conventional sleep parameters, including sleep onset latency (SOL; number of minutes it takes to fall asleep after lights out), sleep efficiency (SE; percentage of time in bed asleep), wake after sleep onset (WASO; number of minutes awake after sleep onset), total sleep time (TST; total number of minutes slept while in bed), and average wake bout length (WBL; total number of minutes awake divided by the number of wake bouts). Inclusion criteria for sleep data required a minimum of three valid nights of actigraphy monitoring, consistent with prior BLSA sleep research 34.

Cardiorespiratory Fitness and Energetic Efficiency Measures

Cardiorespiratory fitness (i.e., maximal oxygen consumption (VO2peak; ml/kg/min)), walking energy utilization (Energetic Cost of Walking (ECW; ml/kg/m), and resting metabolic rate (RMR; kcal/day)) variables were computed and treated as the main outcome variables in our analyses. Certified nurse practitioners and specialized technicians assessed all measures following standardized protocols at the National Institute on Aging Intramural Research Program Clinical Research Unit in Baltimore, Maryland.

VO2peak was recorded during a graded maximal treadmill test following a modified Balke protocol 35. Men walked at 3.5 miles/hour (~5.6 km/hour) and women at 3.0 miles/hour (~4.8 km/hour) on a motor-driven treadmill. The treadmill grade was increased by 3% every 2 minutes until volitional exhaustion. Expired O2 and CO2 were measured continuously using a metabolic cart (Cosmed K4b2, Cosmed, Rome, Italy), and oxygen consumption was calculated every 30 seconds throughout the exercise session, with the highest value used as VO2peak, and a peak respiratory exchange ratio (RER) of at least 1.0 being required 35.

ECW was measured as the energy expended during normal-paced overground walking. Participants were asked to walk continuously around a 20-m course for 2.5 minutes. During the walk, a portable indirect calorimeter (Cosmed K4b2, Cosmed, Rome, Italy) and 2-way nonrebreathing mask were used to collect oxygen consumption (V̇O2) and carbon dioxide production (V̇CO2) using breath-by-breath measurements averaged over 30-second intervals to reduce variability. Readings from the first 1.5 minutes of walking were excluded to allow for workload adjustment. ECW was recorded as the average V̇O2 (mL/kg/min) in the final minute and expressed per meter walked (V̇O2, mL/kg/m) to standardize for gait speed, providing a single measure of ECW.

RMR was evaluated first thing in the morning in a fasted, rested state for 16 minutes via indirect calorimetry after an overnight stay in BLSA clinic. Before testing, the analyzer device (Cosmed K4b2, Cosmed, Rome, Italy) was calibrated using a 3.0-L flow syringe and gases of known concentrations. To reduce variability, the device collected gas-exchange data on a breath-by-breath basis averaged over 30-s intervals. RMR in kilocalories per day was calculated from gas-exchange data using the Weir equation 36. The first 5 min of data were discarded to account for adaptation to the testing procedures, and the remaining 11 min were averaged to obtain a single measure of RMR.

Other Measures

BLSA participants self-reported their demographic information at enrollment, including their age, sex, race/ethnicity, and education level. Race/ethnicity was categorized as ‘White and Other’ or ‘Black’. Body composition (i.e., fatness and leanness) was assessed at each visit after obtaining total fat mass, bone mass, and lean mass via total body dual-energy x-ray absorptiometry (DXA) using the Prodigy Scanner (General Electric, Madison, WI) and analyzed with APEX version 10.51.006 software. We calculated percent fatness and leanness using the following formulas: Fatness = (fat mass / (fat mass + bone mass + lean mass))*100; Leanness = (lean mass / (fat mass + bone mass + lean mass))*100. A dichotomized sleep medication use variable (0 = never, 1 = use for <1/week or more) was calculated for each participant based on self-report. A summed comorbidity index variable was modeled continuously based on the number of the following conditions that participants reported they currently had or had a history of: heart attack, chronic obstructive pulmonary disease, Parkinson’s disease, depression, hypertension, diabetes, coronary heart disease, congestive heart failure, stroke, and peripheral artery disease. Depressive symptoms were measured using the 20-item Center for Epidemiological Studies Depression Scale (CES-D) 35; excluding the CES-D sleep item when computing total scores to avoid collinearity with accelerometer sleep predictors. Smoking history was assessed based on participant self-report. Participants were coded as non-smokers if they ‘never smoked’, ‘quit smoking at least 10 years prior’ or ‘quit within the past 10 years’ (coded as 0), and as smokers if they were a ‘current smoker’ (coded as 1). Better sleep may lead individuals to feel more rested and thus more motivated to engage in physical exercise37. Given regular participation in physical activity may be a potential pathway through which sleep could impact fitness and energetic efficiency, we included self-reported physical activity (i.e., amount of time spent engaging in moderate to vigorous physical activity per week) as a covariate in analyses. The physical activity variable included the following categories: ‘0–29 min of activity/week’, ‘30–74 min of activity/week’, ‘75–150 min of activity/week’, and ‘more than 150 min of activity/week’.

Statistical Analysis

We calculated descriptive statistics (e.g., means, standard deviations) of study variables for the whole sample, and stratified by sex and race. T-tests and Pearson’s chi-squared test were performed to examine sex and race differences in study constructs. Next, prior to model fitting, all continuous covariates and average wake bout length were mean-centered. Total sleep time was mean-centered and divided by 30 minutes. Sleep onset latency and wake after sleep onset were centered by the mean and divided by 10 minutes. Sleep efficiency was mean-centered and divided by 10%.

To examine cross-sectional associations of sleep parameters with fitness and energetic efficiency, we fit multiple linear regression models with a single sleep parameter as the primary predictor and a single fitness or metabolic parameter as the primary outcome measure. We conducted two models, one in which self-reported physical activity was not accounted for and the second in which self-reported physical activity was accounted for. Model 1 included age, sex, race, education, leanness, CES-D score (minus the sleep item), sleep medication use, the comorbidity index, and smoking history, and a single sleep parameter. Model 2 contained all variables in Model 2, plus self-reported physical activity.

We also investigated whether sex, race, and age moderated associations of sleep with fitness and metabolism by fitting two-way interactions between the potential moderator and the primary predictor. Only statistically significant (p < 0.05) interaction terms (sleep X moderator) were reported. Model-derived point estimates for significant interaction terms were calculated to determine the direction and magnitude of associations between sleep and outcomes by race, sex, and age. For significant age interactions, we obtained model-derived point estimates of associations between the primary predictors and outcomes at ages 50, 60, 70, and 80. All data analyses were performed using Stata software version SP 17.0 (StataCorp, College Station, TX).

Results

Participants (N = 263) had a mean ± standard deviation age of 72.7 ± 10.1 years; 141 (53.6%) participants were female and 202 (76.8%) identified as White (Table 1). Participants averaged total body fat of 34.1 ± 8.7%, and a mean CES-D score of 4.0 ± 4.5. The participants’ mean number of comorbidities was 1.0 ± 0.9, 26 (9.9%) participants reported sleep medication use, and 7 (2.7%) were current smokers or quit < 10 years ago. The mean values for age, comorbidity index, VO2peak, and resting metabolic rate were lower in women compared to men. Conversely, women had higher TST and SE relative to men.

Table 1.

Participant Characteristics, Mean (SD) or N (%)

All
(N = 263)		 Sex Race
Male
(n = 122)		 Female
(n = 141)		 P-value a White
(n = 202)		 Non-White
(n = 61)		 P-value a
Age, in years 72.7 (10.1) 74.8 (9.7) 70.9 (10.0) 0.002 73.8 (9.7) 69.1 (10.4) 0.001
Sex 0.12
 Male 122 (46.4%) -- -- 99 (49.0%) 23 (37.7%)
 Female 141 (53.6%) -- -- 103 (51.0%) 38 (62.3%)
Race 0.12
 White 202 (76.8%) 99 (81.1%) 103 (73.0%) -- --
 Non-White 61 (23.2%) 23 (18.9%) 38 (27.0%) -- --
Education 0.48 0.87
 Less than a high school education or high school education or some college 34 (12.9%) 14 (11.5%) 20 (14.2%) 26 (12.9%) 8 (13.1%)
 College education 54 (20.5%) 22 (18.0%) 32 (22.7%) 40 (19.8%) 14 (23.0%)
 Post college education 174 (66.2%) 85 (69.7%) 89 (63.1%) 135 (66.8%) 39 (63.9%)
 Missing 1 (0.4%) 1 (0.8%) 0 (0.0%) 1 (0.5%) 0 (0.0%)
Fatness (%) 34.1 (8.7) 28.5 (6.7) 38.8 (7.2) <0.001 33.0 (8.4) 37.5 (8.8) <0.001
Leanness (%) 62.2 (8.4) 67.5 (6.6) 57.8 (7.0) <0.001 63.3 (8.1) 58.7 (8.4) <0.001
CES-D Scoreb 4.0 (4.5) 4.1 (4.7) 3.9 (4.4) 0.72 3.9 (4.7) 4.2 (4.2) 0.76
Sleep Medication Use 0.73 0.11
 No 212 (80.6%) 98 (80.3%) 114 (80.9%) 157 (77.7%) 55 (90.2%)
 Yes 26 (9.9%) 13 (10.7%) 13 (9.2%) 23 (11.4%) 3 (4.9%)
 Missing 25 (9.5%) 11 (9.0%) 14 (9.9%) 22 (10.9%) 3 (4.9%)
Comorbidity Index 1.0 (0.9) 1.1 (1.0) 0.8 (0.8) 0.009 0.9 (0.9) 1.2 (0.9) 0.028
Smoking Status 0.086 0.57
 Never smoked or quit > 10 years ago 255 (97.0%) 120 (98.4%) 135 (95.7%) 195 (96.5%) 60 (98.4%)
 Current smoker or quit < 10 years ago 7 (2.7%) 1 (0.8%) 6 (4.3%) 6 (3.0%) 1 (1.6%)
 Missing 1 (0.4%) 1 (0.8%) 0 (0.0%) 1 (0.5%) 0 (0.0%)
Physical activity 0.096 0.76
 0–29 min/week 123 (46.8%) 50 (41.0%) 73 (51.8%) 92 (45.5%) 31 (50.8%)
 30–74 min/week 35 (13.3%) 22 (18.0%) 13 (9.2%) 26 (12.9%) 9 (14.8%)
 75–149 min/week 40 (15.2%) 21 (17.2%) 19 (13.5%) 33 (16.3%) 7 (11.5%)
 150+ min/week 63 (24.0%) 27 (22.1%) 36 (25.5%) 49 (24.3%) 14 (23.0%)
 Missing 2 (0.8%) 2 (1.6%) 0 (0.0%) 2 (1.0%) 0 (0.0%)
Fitness measures
 VO2peak 22.8 (5.9) 24.7 (6.3) 21.2 (5.0) <0.001 23.6 (5.8) 20.6 (5.5) 0.001
 Energetic cost of walking (ml/kg/m) 0.16 (0.03) 0.16 (0.03) 0.16 (0.03) 0.23 0.16 (0.03) 0.15 (0.03) <0.001
 Resting metabolic rate (kcal/day) 1136.6 (284.7) 1276.9 (293.3) 1016.2 (214.0) <0.001 1162.3 (288.6) 1049.9 (254.8) 0.011
Sleep parameters
 TST (minutes) 398.2 (62.7) 385.9 (63.5) 408.9 (60.1) 0.003 404.6 (59.4) 377.3 (69.0) 0.003
 WASO (minutes) 46.0 (21.7) 48.5 (23.3) 43.8 (20.1) 0.078 46.4 (22.8) 44.5 (17.8) 0.54
 WBL (minutes) 1.7 (0.8) 1.7 (0.8) 1.7 (0.7) 0.76 1.7 (0.8) 1.6 (0.7) 0.50
 SE (%) 83.6 (7.6) 81.9 (8.5) 85.0 (6.4) <0.001 84.1 (7.5) 81.8 (7.5) 0.037
 SOL (minutes) 13.5 (13.9) 14.8 (16.7) 12.3 (10.9) 0.15 13.1 (14.1) 14.6 (13.2) 0.45
Open in a new tab

Notes.

a

P-value: P-values following a t-test which indicates differences between population subgroups (i.e., sex and race)

b

CES-D: Center for Epidemiologic Studies Depression Scale score with the sleep item removed.

TST= Total Sleep Time; WASO= Wake After Sleep Onset; WBL= Average Wake Bout Length; SE= Sleep Efficiency; SOL= Sleep Onset Latency

Associations of Sleep with Cardiorespiratory Fitness (VO2peak)

Longer TST was associated with higher VO2peak across Models 1–2 (Model 2: B = 0.35 ml/kg/min, 95% CI = 0.05, 0.65) (Table 2). Shorter WBL (Model 2: B = −1.03 ml/kg/min, 95% CI = −1.73, −0.33) and greater SE (Model 2: B = 1.10 ml/kg/min, 95% CI = 0.31, 1.89) were associated with higher VO2peak across Models 1–2, see Table 2.

Table 2.

Cross-sectional Associations of Accelerometer Derived Sleep Metrics with Fitness and Metabolic Efficiency Measures, estimated beta (95% CI) (N = 263)

Model 1 Model 2
Cardiorespiratory Fitness
VO2peak (ml/kg/min)
TST 0.35* (0.05, 0.65) 0.35* (0.05, 0.65)
WASO −0.20 (−0.45, 0.05) −0.21 (−0.46, 0.04)
WBL −0.96* (−1.66, −0.27) −1.03* (−1.73, −0.33)
SE 1.06* (0.27, 1.85) 1.10* (0.31, 1.89)
SOL −0.12 (−0.54, 0.30) −0.13 (−0.55, 0.29)
Metabolic Efficiency
Energy Cost of Walking (ml/kg/m)
TST −0.0016 (−0.0038, 0.0005) −0.0016 (−0.0038, 0.0005)
WASO 0.0010 (−0.0009, 0.0028) 0.0009 (−0.0008, 0.0030)
WBL 0.0038 (−0.0014, 0.0091) 0.0027 (−0.0010, 0.0095)
SE −0.0025 (−0.0082, 0.0032) −0.0028 (−0.0085, 0.0030)
SOL 0.0004 (−0.0028, 0.0035) 0.0004 (−0.0027, 0.0035)
Resting Metabolic Rate (kcal/day)
TST −7.73 (−23.99, 8.52) −7.82 (−24.02, 8.38)
WASO 10.19 (−3.77, 24.15) 9.35 (−4.63, 23.33)
WBL 42.25* (3.18, 84.32) 39.02 (−0.35, 78.38)
SE −24.58 (−66.40, 17.24) −22.82 (−64.60, 18.96)
SOL −6.86 (29.50, 15.77) −6.73 (−29.30, 15.83)
Open in a new tab

Notes.

*

p ≤.05

**

p <.01

***

p <.001

Model 1: adjusted for age, sex, race, education, leanness, sleep medication use, depression symptoms (minus the sleep item), comorbidities, and smoking status

Model 2 adjusted for Model 1 covariates and physical activity

TST= Total Sleep Time; WASO= Wake After Sleep Onset; WBL= Average Wake Bout Length; SE= Sleep Efficiency; SOL= Sleep Onset Latency.

Associations of Sleep with Energetic Efficiency

Energy Cost of Walking (ECW).

No significant associations or moderating effects were observed for ECW.

Resting Metabolic Rate (RMR).

Longer WBL was associated with higher RMR in Model 1 (Model =1: B = 42.25 kcal/day, 95% CI = 3.18, 84.32), but not in Model 2 (Model =2: B = 39.02 kcal/day, 95% CI = −0.35, 78.38). There was an interaction between age and SE (interaction p-value = 0.014) such that greater SE was linked to lower RMR at ages 50 and 60 (Age 50: B = −131.73 kcal/day, 95% CI = −229.03, −34.44; Age 60: B = −82.90, 95% CI = −146.44, −19.37), but not at 70 or 80.

Discussion

The present study adds to the limited research identifying an association between sleep and fitness using objectively measured sleep and builds on previous work to include resting energy expenditure and energetic efficiency. Findings reveal that shorter total sleep time (TST), lower sleep efficiency (SE), and longer wake bout length (WBL) are associated with lower CRF and less efficient energy utilization in middle-aged and older adults. Notably, our findings suggest these associations may occur independent of self-reported physical activity levels. Further, we observed that some associations between sleep and energetic efficiency were more pronounced in middle- and younger-old-aged adults than those over the age of 70. Combined with previous evidence on sleep and cardiometabolic health, these results suggest that better sleep, particularly in middle-aged and younger-old-aged adults, is related to better fitness and energy utilization, which may help explain previously established associations between sleep and cardiometabolic health in late life.

Sleep and Cardiorespiratory Fitness

Our findings that lower SE and longer WBL are associated with poorer CRF are in line with a previous study showing that worse accelerometer derived SE and WBL were associated with poorer CRF in heart failure patients 38. Additionally, a large cross-sectional study of nearly 3500 adults in Norway found that self-reported insomnia symptoms were inversely related to CRF in models adjusted for subjective physical activity 39, while a small study in 28 men found that poorer self-reported sleep quality, measured by the Pittsburgh Sleep Quality Index, was related to lower CRF 26. Our findings are consistent with these previous reports, suggesting that poorer sleep efficiency and greater sleep disturbances are associated with lower CRF in middle-aged and older adults.

In adjusted models, we found that each additional 30 minutes of total sleep time was associated with an approximately 5% higher VO₂peak. This is partially consistent with previous reports of associations of shorter self-reported TST and more severe self-reported insomnia symptoms with higher BMI, waist circumference, and cardiovascular disease among adults 40,41. Shorter self-reported sleep duration and insomnia symptoms have also been linked to lower CRF in middle-aged adults 26,42. Mechanistically, poor sleep is associated with various factors that are negatively associated with cardiovascular health and could impact CRF, including autonomic nervous system imbalance, inflammation, endothelial dysfunction, and poor blood pressure regulation 43,44. Although a further understanding of these potential mechanisms is outside of the scope of our study, it is the first to assess associations between objectively measured sleep duration and VO2peak—measures that are less affected by recall or social desirability bias than self-reported measures 27. Therefore, while our results provide more objective evidence for an association between sleep duration and CRF in middle-aged and older adults, additional longitudinal studies are needed to confirm this link and identify underlying mechanisms.

Sleep and Energetic Efficiency

We also found that longer WBL was associated with higher RMR in our sample when not accounting for self-reported physical activity. Although prior research has linked self-reported shorter sleep duration with lower RMR, fat oxidation, and weight gain in younger and middle-aged adults 45,46, our findings raise the possibility that poor sleep may contribute to poor cardiovascular and metabolic outcomes in mid-to-late life. In healthy middle-aged and older adults, advancing age is typically associated with a decline in resting metabolic rate (RMR), independent of sex and body composition. However, when accounting for these factors, higher RMR among older adults has been linked to a greater prevalence of disease and comorbid conditions, as well as greater disease risk at follow-up 47. This suggests that the observed association between longer WBL and higher RMR in our sample may reflect increased homeostatic demand due to early changes in health status, rather than a direct effect of aging itself 13,47. This interpretation aligns with a previous cross-sectional study of older adults which found that age was positively associated with overnight RMR and that better sleep quality, as measured by polysomnography, was associated with lower overnight RMR 20.

Importantly, we found that age moderated the association of SE with RMR, such that higher SE was associated with lower RMR at ages 50 and 60 years, but not at ages 70 or 80. These results are partially consistent with previous work reporting self-reported sleep duration was more strongly associated with a composite metabolic risk score (blood pressure, cholesterol, fasting glucose, and triglyceride levels) in younger middle-aged adults than among older adults 48. Furthermore, in previous work in the BLSA, higher RMR predicted future multimorbidity 47, which may also contribute to sleep disruptions 49. Therefore, our findings suggest that better sleep is associated with a more efficient resting metabolism in middle-aged and young older adults, which could be partially related to previously established associations between poorer sleep and hormonal imbalances, metabolic dysregulation, and a greater number of mood and stress disorders 44. Importantly, our associations persisted even after adjusting for comorbid conditions. Notably, comorbidity was low in our sample, which may limit the generalizability of our findings. While we captured a broad measure of overall health burden, future studies in populations with a higher comorbidity burden may provide more nuanced insights into how specific comorbidities influence energetic measures such as RMR. Moreover, subclinical disease may contribute to both poorer sleep quality and higher RMR in mid-to-late life, warranting further research to explore these associations and their temporality.

A greater ECW has consistently been linked to adverse health outcomes in older adults 15, but we did not find any associations between accelerometer derived sleep parameters and ECW. This lack of association could be due to our sample being relatively healthy, as higher ECW often reflects biomechanical inefficiencies, as opposed to metabolic “diseases”, which are more strongly tied to poor sleep 50. To our knowledge, there are no previous reports of associations between sleep quantity and quality measures and ECW in middle-aged and older adults. Future studies in this domain are needed to further confirm our findings.

Limitations and Conclusions

As our analytic sample consisted of primarily White, highly educated middle-aged and older adults free of cognitive impairment, results may not extend to the general population. Moreover, although we used an objective measure to estimate sleep parameters, we did not use polysomnography—the gold standard sleep measure. This precluded assessing the role of sleep stages (i.e., sleep architecture) or sleep-disordered breathing with fitness, an important area for future research. We also did not preclude participants based on the presence of sleep disorders such as sleep apnea, insomnia, or narcolepsy. Although we controlled for leading risk factors for many of these disorders (i.e. age, sex, and body composition), it is possible their presence could further influence or confound these relationships. Additionally, previous reviews have indicated potential bidirectional associations of CRF and physical activity with sleep; however, we did not examine these potential associations in this manuscript due to the challenges of determining causal inference in observational studies and insufficient longitudinal sleep data. It is also possible that level of physical function and disability could partially confound associations between sleep problems and energetic efficiency, while we controlled for reported regular participation in physical activity in our analysis, future studies should further account for these factors. Future studies are needed to examine the potential bidirectional associations among sleep, physical activity, and CRF and energetic efficiency longitudinally, which may highlight opportunities for intervention.

In summary, we found that objective indices of poor sleep were associated with lower CRF and less efficient energy utilization among middle aged and older adults, independent of physical activity levels. Furthermore, age moderated associations between sleep quality and RMR, suggesting some sleep and energetic efficiency associations may be stronger in middle-aged and younger-old adults than in older age groups. Longitudinal studies are needed to elucidate relationships among objective sleep measures, CRF, and energetic efficiency in these populations.

Acknowledgments:

Funding:

This work was supported in part by the National Institute on Aging (grant number R01AG050507) and the Intramural Research Program, National Institute on Aging, NIH and Research and Development Contract HHSN-260-2004-00012C. JAS was supported by U01AG057545. AAW was supported by Grant Number K01 AG076967 from the National Institute on Aging, National Institutes of Health.

Footnotes

Conflicts of Interest: Adam Spira received payment for serving as a consultant for Merck, received honoraria from Springer Nature Switzerland AG for guest editing special issues of Current Sleep Medicine Reports, and is a paid consultant to Sequoia Neurovitality, BellSant, Inc., and Amissa, Inc. JAS is a consultant for Edwards Lifesciences and the Villages, Inc.

JAS, APS, AAW, EMS, and LF are on the editorical board of JGMS.

Data Availability:

Data supporting the results of this study are available on request from the BLSA team.

References

  • 1.Luyster FS, Strollo PJ, Zee PC, Walsh JK. Sleep: A Health Imperative. Sleep. 2012;35(6):727. doi: 10.5665/SLEEP.1846 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Williams JA, Naidoo N. Sleep and cellular stress. Curr Opin Physiol. 2020;15:104–110. doi: 10.1016/j.cophys.2019.12.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Mander BA, Winer JR, Walker MP. Sleep and Human Aging. Neuron. 2017;94(1):19. doi: 10.1016/J.NEURON.2017.02.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Li J, Vitiello MV., Gooneratne NS. Sleep in Normal Aging. Sleep Med Clin. 2018;13(1):1. doi: 10.1016/J.JSMC.2017.09.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Da Silva AA, De Mello RGB, Schaan CW, Fuchs FD, Redline S, Fuchs SC. Sleep duration and mortality in the elderly: a systematic review with meta-analysis. BMJ Open. 2016;6(2). doi: 10.1136/BMJOPEN-2015-008119 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Full KM, Huang T, Shah NA, et al. Sleep Irregularity and Subclinical Markers of Cardiovascular Disease: The Multi-Ethnic Study of Atherosclerosis. J Am Heart Assoc. 2023;12(4):27361. doi: 10.1161/JAHA.122.027361 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.St-Onge MP, Grandner MA, Brown D, et al. Sleep Duration and Quality: Impact on Lifestyle Behaviors and Cardiometabolic Health: A Scientific Statement From the American Heart Association. Circulation. 2016;134(18):e367. doi: 10.1161/CIR.0000000000000444 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Kuo PL, Schrack JA, Shardell MD, et al. A roadmap to build a phenotypic metric of ageing: insights from the Baltimore Longitudinal Study of Aging. J Intern Med. 2020;287(4):373–394. doi: 10.1111/JOIM.13024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Fleg JL, Morrell CH, Bos AG, et al. Accelerated longitudinal decline of aerobic capacity in healthy older adults. Circulation. 2005;112(5):674–682. doi: 10.1161/CIRCULATIONAHA.105.545459 [DOI] [PubMed] [Google Scholar]
  • 10.Phan DQ, Zheng C, Thai T, et al. Cardiorespiratory Fitness and Mortality in Patients Aged 60 to 90 Years. Am J Cardiol. 2022;170:132–137. doi: 10.1016/J.AMJCARD.2022.01.035 [DOI] [PubMed] [Google Scholar]
  • 11.Sui X, Sarzynski MA, chul Lee D, Kokkinos PF. Impact of Changes in Cardiorespiratory Fitness on Hypertension, Dyslipidemia and Survival: An Overview of the Epidemiological Evidence. Prog Cardiovasc Dis. 2017;60(1):56–66. doi: 10.1016/J.PCAD.2017.02.006 [DOI] [PubMed] [Google Scholar]
  • 12.Schrack JA, Simonsick EM, Chaves PHM, Ferrucci L. The Role of Energetic Cost in the Age-Related Slowing of Gait Speed. J Am Geriatr Soc. 2012;60(10):1811. doi: 10.1111/J.1532-5415.2012.04153.X [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Schrack JA, Knuth ND, Simonsick EM, Ferrucci L. “IDEAL” Aging is Associated with Lower Resting Metabolic Rate: The Baltimore Longitudinal Study of Aging. J Am Geriatr Soc. 2014;62(4):667. doi: 10.1111/JGS.12740 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Schrack JA, Zipunnikov V, Simonsick EM, Studenski S, Ferrucci L. Rising Energetic Cost of Walking Predicts Gait Speed Decline With Aging. J Gerontol A Biol Sci Med Sci. 2016;71(7):947. doi: 10.1093/GERONA/GLW002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Abellan Van Kan G, Rolland Y, Andrieu S, et al. Gait speed at usual pace as a predictor of adverse outcomes in community-dwelling older people an International Academy on Nutrition and Aging (IANA) Task Force. J Nutr Health Aging. 2009;13(10):881–889. doi: 10.1007/S12603-009-0246-Z [DOI] [PubMed] [Google Scholar]
  • 16.Buracchio T, Dodge HH, Howieson D, Wasserman D, Kaye J. The trajectory of gait speed preceding mild cognitive impairment. Arch Neurol. 2010;67(8):980–986. doi: 10.1001/ARCHNEUROL.2010.159 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Shih YL, Shih CC, Chen JY. The association between walking speed and risk of cardiovascular disease in middle-aged and elderly people in Taiwan, a community-based, cross-sectional study. PLoS One. 2020;15(7). doi: 10.1371/JOURNAL.PONE.0235277 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Manini TM. Energy Expenditure and Aging. Ageing Res Rev 2010;9(1):1. doi: 10.1016/J.ARR.2009.08.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Schrack JA, Simonsick EM, Ferrucci L. The energetic pathway to mobility loss: an emerging new framework for longitudinal studies on aging. J Am Geriatr Soc. 2010;58 Suppl 2(Suppl 2). doi: 10.1111/J.1532-5415.2010.02913.X [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Valenti G, Bonomi AG, Westerterp KR. Quality Sleep Is Associated With Overnight Metabolic Rate in Healthy Older Adults. The Journals of Gerontology: Series A. 2017;72(4):567–571. doi: 10.1093/GERONA/GLW107 [DOI] [PubMed] [Google Scholar]
  • 21.Christie AD, Seery E, Kent JA. Physical activity, sleep quality, and self-reported fatigue across the adult lifespan. Exp Gerontol. 2016;77:7–11. doi: 10.1016/J.EXGER.2016.02.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Collins KJ, Schrack JA, Vanswearingen JM, et al. Randomized Controlled Trial of Exercise to Improve Walking Energetics in Older Adults. Innov Aging. 2018;2(3). doi: 10.1093/GERONI/IGY022 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Sharma S, Kavuru M. Sleep and Metabolism: An Overview. Int J Endocrinol. 2010;2010:12. doi: 10.1155/2010/270832 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Oudegeest-Sander MH, Eijsvogels THM, Verheggen RJHM, et al. Impact of Physical Fitness and Daily Energy Expenditure on Sleep Efficiency in Young and Older Humans. Gerontology. 2012;59(1):8–16. doi: 10.1159/000342213 [DOI] [PubMed] [Google Scholar]
  • 25.Ramania NS, Tommy TA, Bagus BW. The effects of differences in sleep quality and quantity on VO2max levels. Advances in Rehabilitation. 2020;34(4):11–17. doi: 10.5114/AREH.2020.100968 [DOI] [Google Scholar]
  • 26.Antunes BM, Campos EZ, Parmezzani SS, Santos RV., Franchini E, Lira FS. Sleep quality and duration are associated with performance in maximal incremental test. Physiol Behav. 2017;177:252–256. doi: 10.1016/J.PHYSBEH.2017.05.014 [DOI] [PubMed] [Google Scholar]
  • 27.Aili K, Åström-Paulsson S, Stoetzer U, Svartengren M, Hillert L. Reliability of Actigraphy and Subjective Sleep Measurements in Adults: The Design of Sleep Assessments. J Clin Sleep Med 2017;13(1):39. doi: 10.5664/JCSM.6384 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Santos CCL Dos, Matharoo AS, Cueva EP, et al. The Influence of Sex, Age, and Race on Coronary Artery Disease: A Narrative Review. Cureus. 2023;15(10):e47799. doi: 10.7759/CUREUS.47799 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Ruggiero AR, Peach HD, Gaultney JF. Health Psychology and Behavioral Medicine An Open Access Journal Association of sleep attitudes with sleep hygiene, duration, and quality: a survey exploration of the moderating effect of age, gender, race, and perceived socioeconomic status. Health Psychol Behav Med. 2019;7(1):19–44. doi: 10.1080/21642850.2019.1567343 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Rossman I Normal Human Aging: The Baltimore Longitudinal Study of Aging. JAMA. 1986;255(7):960–960. doi: 10.1001/JAMA.1986.03370070114046 [DOI] [Google Scholar]
  • 31.Williams OA, An Y, Armstrong NM, et al. Apolipoprotein E ε4 allele effects on longitudinal cognitive trajectories are sex- and age-dependent. Alzheimers Dement. 2019;15(12):1558. doi: 10.1016/J.JALZ.2019.07.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Rabinowitz JA, Zhou Y, Chen D, et al. Associations of personality traits with actigraphic sleep in middle-aged and older adults. Scientific Reports 2024 15:1. 2025;15(1):1–10. doi: 10.1038/s41598-024-80426-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Kushida CA, Chang A, Gadkary C, Guilleminault C, Carrillo O, Dement WC. Comparison of actigraphic, polysomnographic, and subjective assessment of sleep parameters in sleep-disordered patients. Sleep Med. 2001;2(5):389–396. doi: 10.1016/S1389-9457(00)00098-8 [DOI] [PubMed] [Google Scholar]
  • 34.Owusu JT, Rabinowitz JA, Tzuang M, et al. Associations Between Objectively Measured Sleep and Cognition: Main Effects and Interactions With Race in Adults Aged ≥50 Years. J Gerontol A Biol Sci Med Sci 2023;78(3):454. doi: 10.1093/GERONA/GLAC180 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Bassett DR, Howley ET. Limiting factors for maximum oxygen uptake and determinants of endurance performance. Med Sci Sports Exerc. 2000;32(1):70–84. doi: 10.1097/00005768-200001000-00012 [DOI] [PubMed] [Google Scholar]
  • 36.Compher C, Frankenfield D, Keim N, Roth-Yousey L. Best practice methods to apply to measurement of resting metabolic rate in adults: a systematic review. J Am Diet Assoc. 2006;106(6):881–903. doi: 10.1016/J.JADA.2006.02.009 [DOI] [PubMed] [Google Scholar]
  • 37.Sewell KR, Erickson KI, Rainey-Smith SR, Peiffer JJ, Sohrabi HR, Brown BM. Relationships between physical activity, sleep and cognitive function: A narrative review. Neurosci Biobehav Rev. 2021;130:369–378. doi: 10.1016/J.NEUBIOREV.2021.09.003 [DOI] [PubMed] [Google Scholar]
  • 38.Ahmed SI, Jonathan K, Billingsley H, et al. Reduced Sleep Quality Is Associated With Worse Cardiorespiratory Fitness In Heart Failure With Preserved Ejection Fraction. J Card Fail. 2023;29(4):592. doi: 10.1016/J.CARDFAIL.2022.10.115 [DOI] [Google Scholar]
  • 39.Strand LB, Laugsand LE, Wisløff U, Nes BM, Vatten L, Janszky I. Insomnia Symptoms and Cardiorespiratory Fitness in Healthy Individuals: The Nord-Trøndelag Health Study (HUNT). Sleep. 2013;36(1):99–108. doi: 10.5665/SLEEP.2310 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.St-Onge MP, Perumean-Chaney S, Desmond R, et al. Gender Differences in the Association between Sleep Duration and Body Composition: The Cardia Study. Int J Endocrinol. 2010;2010. doi: 10.1155/2010/726071 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Bertisch SM, Reid M, Lutsey PL, et al. Gender differences in the association of insomnia symptoms and coronary artery calcification in the multi-ethnic study of atherosclerosis. Sleep. 2021;44(10). doi: 10.1093/SLEEP/ZSAB116 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Zou D, Wennman H, Ekblom Ö, et al. Insomnia and cardiorespiratory fitness in a middle-aged population: the SCAPIS pilot study. Sleep and Breathing. 2019;23(1):319–326. doi: 10.1007/S11325-018-1765-9/TABLES/3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Hall MH, Brindle RC, Buysse DJ. Sleep and Cardiovascular Disease: Emerging Opportunities for Psychology. Am Psychol. 2018;73(8):994. doi: 10.1037/AMP0000362 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Sharma S, Kavuru M. Sleep and Metabolism: An Overview. Int J Endocrinol. 2010;2010:12. doi: 10.1155/2010/270832 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Spaeth AM, Dinges DF, Goel N. Resting metabolic rate varies by race and by sleep duration. Obesity (Silver Spring). 2015;23(12):2349. doi: 10.1002/OBY.21198 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Jurado-Fasoli L, Mochon-Benguigui S, Castillo MJ, Amaro-Gahete FJ. Association between sleep quality and time with energy metabolism in sedentary adults. Scientific Reports 2020 10:1. 2020;10(1):1–10. doi: 10.1038/s41598-020-61493-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Fabbri E, An Y, Schrack JA, et al. Energy Metabolism and the Burden of Multimorbidity in Older Adults: Results From the Baltimore Longitudinal Study of Aging. J Gerontol A Biol Sci Med Sci. 2015;70(11):1297–1303. doi: 10.1093/GERONA/GLU209 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Arora A, Pell D, van Sluijs EMF, Winpenny EM. How do associations between sleep duration and metabolic health differ with age in the UK general population? PLoS One. 2020;15(11). doi: 10.1371/JOURNAL.PONE.0242852 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Zhou Y, Jin Y, Zhu Y, et al. Sleep Problems Associate With Multimorbidity: A Systematic Review and Meta-analysis. Public Health Rev. 2023;44:1605469. doi: 10.3389/PHRS.2023.1605469/FULL [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Cappuccio FP, Miller MA. Sleep and Cardio-Metabolic Disease. Curr Cardiol Rep. 2017;19(11). doi: 10.1007/S11886-017-0916-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Data supporting the results of this study are available on request from the BLSA team.

RESOURCES