Journal Logo


Effects of Weight Loss on Lean Mass, Strength, Bone, and Aerobic Capacity


Author Information
Medicine & Science in Sports & Exercise: January 2017 - Volume 49 - Issue 1 - p 206-217
doi: 10.1249/MSS.0000000000001074


Dietary energy restriction causes weight loss and provides powerful protection against many chronic diseases. However, the catabolic state induced by calorie restriction (CR) not only affects fat tissue but also causes undesirable catabolism of other tissues, including skeletal muscle, heart, liver, and kidney (3); many of these changes could have important functional and/or physiologic ramifications. Skeletal muscle mass decreases during weight loss from dietary restriction (2,17), which reduces muscle strength (26) and resting metabolic rate (13). CR causes bone loss (7,23), which may increase fracture risk (8). Cardiac mass seems to decrease during CR (3), which may reduce maximal cardiac output and aerobic capacity.

Exercise is an important adjunct to CR for weight loss. It contributes to the energy deficit and may also promote long-term weight maintenance after weight loss (5). Exercise may also attenuate some of the negative consequences of weight loss on lean mass (4,22), bone (24,32), and aerobic capacity (20). We previously reported that modest weight loss (~9%) induced by 1 yr of 20% CR decreased muscle mass and strength, absolute aerobic capacity, and bone mineral density (BMD), whereas full preservation of these outcomes occurred during a similar magnitude weight loss induced by endurance exercise training (EX) (33,36). However, in the absence of diet modification, the amount of exercise required to produce meaningful weight loss is substantial (~1 h daily) (16), especially for overweight and obese individuals who may have limited success in adhering to exercise programs. Therefore, it may be prudent to use more modest amounts of exercise in combination with CR. However, it is not clear if more modest amounts of exercise (e.g., ~30–40 min·d−1), when used in combination with CR to promote weight loss, are effective for preserving lean mass, muscle strength, bone, and aerobic capacity.

The purpose of the present study was to evaluate the effects of matched, modest (~7%) weight loss, induced by CR, EX, or both (CREX) on lean mass, muscle strength, bone, and aerobic capacity. We hypothesized that the lesser amount of exercise used in the CREX intervention would be sufficient to prevent some, but not all of the losses were induced by CR alone. By contrast, full preservation of lean mass, bone, and aerobic capacity would occur in response to the same magnitude of weight loss when induced by EX alone. This study is unique from most previous studies in that the energy deficit and weight loss were partially (CREX group) or fully (EX group) induced by exercise. In most previous studies, the exercise interventions did not contribute meaningfully to the weight loss, which was mainly induced by CR. Furthermore, the present study included a group that underwent weight loss solely by using exercise, which is rare in the existing scientific literature but is representative of many individuals in the population who maintain healthy body weight by performing large volumes of exercise without restricting energy intake. The data in this report were from a trial in which insulin sensitivity and related outcomes were primary outcomes. The primary outcomes have been published previously (35) and the data in the present report reflect secondary outcomes.


Study Participants

Nonsmoking, overweight (body mass index [BMI] = 25.0–29.9 kg·m−2), middle-age (45–65 yr) men and postmenopausal women were recruited from the Saint Louis Missouri metropolitan area. Participants were required to be weight stable (<3% change in body weight) and nonexercisers (<20 min per session and <3 d·wk−1) for at least 6 months before enrollment. A medical evaluation and a diagnostic ECG stress test were used to identify and exclude individuals with major chronic diseases and/or conditions for which exercise is contraindicated, such as coronary artery disease, diabetes, and musculoskeletal problems. For medications, stable dosages were required for ≥6 months before baseline testing, and the participants were advised to maintain stable dosages throughout the study. Participants provided informed written consent to participate in the study, which was reviewed and approved by the Institutional Review Boards at Saint Louis University and Washington University. The study was registered at (registry ID: NCT00777621).

Study Design and Randomization

After baseline testing, participants were randomly assigned to CR, EX, or CREX. The randomization sequence was generated by using and included multiple block sizes and stratification for sex. The allocation ratio was initially 1:1:1 and was later revised to 2:2:1 to account for greater drop-out rates in the CR and EX groups. The study statistician generated the randomization sequence, the research coordinators enrolled participants, and personnel who were not otherwise involved in the study used the allocation sequence to assign participants to the study groups. Outcomes were assessed at baseline and again after a 2-wk weight stability period which followed the active weight loss portion of the intervention (see Figure, Supplemental Digital Content 1, study timeline for weight loss interventions, weight stabilization, and outcomes testing,


The interventions were designed to create a 20% energy deficit, with the goal of reducing body weight by 6%–8% for 12–14 wk (see Figure, Supplemental Digital Content 1, study timeline for weight loss interventions, weight stabilization, and outcomes testing, The duration of the intervention was adjusted as needed to allow subjects to reach the weight loss goal. The initial prescribed energy deficit was calculated as 20% over baseline total energy expenditure and energy intake, as described previously (35). The CR and the EX prescriptions were adjusted each week as needed to yield the targeted rate of weight loss. After the weight loss goal was attained, the prescriptions were adjusted so that body weight remained stable (≤0.5 kg variation based on 3-d rolling average) for at least 2 wk before follow-up testing commenced (see Figure, Supplemental Digital Content 1, study timeline for weight loss interventions, weight stabilization, and outcomes testing, Participants measured and recorded fasted morning body weight at home each day for monitoring of weight loss progress during the intervention (weight as a study outcome was measured in the laboratory as described in the next section); they also visited our facility weekly to turn in home weight logs and for other study requirements, as described below.


The CR intervention was designed to reduce body weight by decreasing dietary energy intake while holding physical activity at baseline levels. Study dietitians counseled the participants on a weekly basis to reduce energy intake by reducing food portion sizes and by replacing energy-dense foods with those of lower energy density. Individualized guidance was based on 3-d food diaries, which were recorded weekly during the first 3 wk of the intervention and as needed thereafter as a tool for ongoing dietary monitoring and counseling during the intervention. The intervention was also designed to keep macronutrient balance within the recommended ranges (15); this was accomplished by qualitative (and as needed, quantitative nutrient analysis) evaluation of the food diaries to identify deviations from recommendations (e.g., low carbohydrate dietary patterns) and specific dietary recommendations designed to rectify problems. Participants with poor dietary compliance, as evidenced by lack of weight loss and little or no CR (based on dietary assessment), were given full food provision for 7 d at a time to improve compliance.


The exercise intervention was designed to reduce body weight by increasing exercise energy expenditure and holding energy intake constant at baseline levels. The types of exercise included cardiovascular exercise, such as brisk walking, jogging, and cycling, and functional physical activities, such as walking to work or adding yard work to the usual routine. Strength training was not included. Exercise energy expenditure prescriptions were calculated as described previously and included an adjustment to account for the energy expenditure that would have occurred in the absence of exercise (37). The participants were provided with wristwatch-type HR monitors (Polar, Kempele, Finland) to monitor and record exercise data (energy expenditure, HR, and exercise frequency and duration); these data were recorded by study personnel during weekly meetings with the participants. To achieve the weekly exercise energy expenditure goals, the participants were encouraged to exercise ~60 min or more each day and to use moderate to vigorous intensity (intensity recommendations were generally based on subjective ratings of moderate or vigorous overall physical exertion during exercise but more objective guidance was provided as needed); however, beyond these general recommendations, no specific prescriptions were provided for exercise frequency, intensity, and duration. In light of the time demands of the exercise intervention, the participants were encouraged to exercise at convenient locations such as outdoors, in a home gym, or at a conveniently located exercise facility. Free access to the university’s exercise facilities was also provided.


The CREX intervention included both CR and EX, with each contributing approximately half to the total energy deficit. Therefore, the degree of CR and the exercise volume were lower than those in the CR and the EX interventions, respectively. Otherwise, the intervention methods were the same as those described for CR and EX.

Body Weight and Composition

At baseline and final testing, fasted, morning, gowned body weight was measured on two separate days and averaged. Standing height without shoes was measured to the nearest millimeter using a wall-mounted stadiometer (PORTROD, Health o meter Professional Scales, Inc., McCook, IL).

Dual-energy x-ray absorptiometry (DXA; Lunar iDXA, software version 13.31, GE Healthcare, Madison, WI) was used to measure whole-body fat mass, whole-body and appendicular lean mass, and whole-body and regional (spine, hip, and radius) BMD and bone mineral content (BMC). Fat and lean mass data and whole-body bone data were derived from a whole-body scan. Regional scans were performed for BMD and BMC assessments of the lumbar spine (L1–L4), total hip, and wrist (distal third of the radius). Lean mass measures excluded both bone and fat mass. Upper and lower appendicular regions were automatically defined by DXA scanner’s software, with manual review and adjustments made by the DXA technician, as needed. In brief, the appendicular regions were defined as the tissue distal to a line bisecting the shoulder joint for the upper appendages and bisecting the hip joint for the lower appendages. Lean mass in the appendicular regions primarily represents skeletal muscle (14). However, lean mass may also include nonmuscle components (e.g., blood and some interstitial fluid), and DXA-based measures of lean mass may be susceptible to greater measurement error than more direct measures of muscle mass, such as computed tomography scans (9).

The DXA scanner was calibrated with a bone phantom every day before any scans were performed, and a soft tissue calibration bar was positioned in the scan area for all whole-body scans. The scans on study participants were performed in the morning after an overnight fast, and while the participant was wearing only a hospital gown and underwear. All radiopaque objects (jewelry, watches, etc.) were removed from the participant before the scan. During the scan, the DXA technician monitored the subject and scan for evidence of movement and to ensure that the appropriate region of interested was captured. Scans were repeated if needed to correct for motion artifacts or other artifacts.

Dietary Intakes and Physical Activity Levels

Energy and nutrient intakes were measured by performing computerized nutrient analyses (Food Processor SQL software; ESHA Research, Inc., Salem, OR) of 3-d food diaries, which included two weekdays and one weekend day. Estimates of calcium and vitamin D intake included supplements. Total energy expenditure was estimated as the average results from a modified Stanford 7-d physical activity recall interview (25) and accelerometry (RT3 triaxial accelerometers; StayHealthy, Inc., Monrovia, CA). Exercise energy expenditure data (EX and CREX groups only) were retrieved each week from the HR monitors. Because these data reflect gross energy expenditure, net exercise energy expenditure was calculated by subtracting the energy expenditure that would have occurred in the absence of exercise, which was approximated as the product of baseline total energy expenditure in kilocalories per minute and total exercise time. As outcomes assessments, dietary intakes and energy expenditure measures were performed once at baseline and once at the end of the active weight loss period (i.e., before weight stability). Exercise energy expenditure data from the HR monitors were collected throughout the intervention.

Muscle Strength

Concentric isometric and isokinetic knee flexor and knee extensor strength was measured with as Biodex System 4 dynamometer (Shirley, NY). Before the tests, the participants performed light-effort cycle ergometer exercise for 5 min. Furthermore, immediately before each of the isokinetic tests described below, a test-specific warm-up was performed, which consisted of three repetitions of alternating knee extension and flexion at 75% of maximal effort. Multiple practice sessions, which are often used in strength training studies to allow for neurologic strength gains, were not used in the present study because strength gains were not expected and the study did not involve strength training. The participants were seated with the hips flexed at 85° and the back supported and were secured to the dynamometer with padded straps. The seat and attachment positions on the dynamometer for each participant were recorded during the baseline test and the same setup was used for the follow-up test. Before starting each test, the participants were given clear and standardized instructions to give maximal effort during the test and continue until the technician said stop; verbal encouragement was not given during the test. Tests were performed on the right and left legs. Results were corrected for gravity effect torque of the lower extremity. The isometric tests were performed with the knee flexed at 45° and included three repetitions of alternating extension and flexion. The isokinetic tests were performed at an angular velocity of 60°·s−1 and 180°·s−1. Five repetitions of alternating extension and flexion were performed for the 60°·s−1 tests, and 30 repetitions were performed for the 180°·s−1 tests. The peak torque value from each test was used as the measure of strength; results from tests on the right and left limbs were summed. Strength data were also reported relative to body weight and relative to appendicular and lower extremity lean mass as an index of muscle specific force. The reliability of isokinetic strength testing using similar procedures to those described previously has an SE of measure of 6%–9%, with intraclass correlation coefficients of 0.90–0.96 (18,29).

Physiological Responses to Maximal-Intensity Exercise

The treadmill test started at 0% grade and a speed determined during a 5-min warm-up to elicit an HR of ~70% of age-predicted maximal HR. The grade was then increased by two percentage points every 1 to 2 min until the subject could no longer continue because of fatigue. Indirect calorimetry (MedGraphics CardioO2; Graphics Corporation, St. Paul, MN) was used to continuously measure metabolic outcomes during the test, with time-averaged values being generated every 30 s. Before each test, the pneumotach airflow meter was calibrated with a 3-L air syringe and the carbon dioxide, and oxygen analyzers were calibrated with medical grade calibration gases. The highest 30-s average oxygen uptake was considered maximal oxygen uptake (V˙O2max). The tests were also evaluated for meeting the criteria for “true V˙O2max” based meeting ≥2 of the following criteria: 1) maximal RER (RERmax) ≥1.10; 2) measured maximal HR ≥ predicted HRmax minus 10; and 3) increase in oxygen uptake of ≤150 mL·min−1 between the last two stages of the test. HR was measured continuously by using ECG; HRmax was determined by manually measuring five or more normal sinus R-R intervals. Blood pressure (BP) was measured via auscultation at rest, every 2 min during the exercise test and recovery, and during maximal exercise. Maximal exercise oxygen pulse was calculated as absolute V˙O2max divided by HRmax. Maximal exercise rate pressure product was calculated as the product of HRmax and maximal exercise systolic BP and was used as an index of myocardial work rate.

Statistical Analyses

Analyses were performed on data from subjects who provided both baseline and follow-up data and were adherent to the intervention; lack of adherence was defined as <1% weight loss. Comparisons of baseline characteristics among groups were performed using Fisher’s exact tests and ANOVA. Between-group comparisons of the changes in outcomes were performed using ANCOVA in which the baseline-to-follow-up change was the dependent variable and treatment group was the independent variable; covariates included in the initial model were baseline values of the outcome, magnitude of weight loss, and sex, but these were removed one at a time if they were not meaningful predictors of the outcome, based on P ≤ 0.15. Post hoc comparisons were performed based on the protected F-test principal and least significant difference tests. Significance values of the within-group changes were based on the least squares means from the ANCOVA. Strength data residuals were not normally distributed nor were mathematical transformations effective for normalization. Therefore, strength outcomes were analyzed by using nonparametric Kruskal–Wallis rank sum tests for between-group comparisons and Signed rank tests for within-group changes. Multiple regression analyses used a stepwise selection model, with P < 0.15 required for entry into the model. All statistical tests were two-tailed, and significance was accepted at P ≤ 0.05. Data from parametric tests are presented as arithmetic mean ± SE except for mean change values, which were adjusted for baseline values. Data from nonparametric tests are presented as medians and interquartile ranges except as noted. Analyses were performed using SAS for Windows (version 9.3; SAS Institute Inc., Cary, NC). Sample size calculations were based on glucoregulation outcomes, which have been published previously (35).



Of the 69 women and men who enrolled in the study, underwent baseline testing, and were randomly allocated to study groups, 52 completed the study and were compliant with the interventions. Twelve participants dropped out before providing follow-up data, and five participants were noncompliant as evidenced by negligible weight changes ranging from 0.5% weight loss to 1.3% weight gain. Additional details about the number of subjects recruited, screened, and enrolled are depicted in a Consort diagram in Supplemental Digital Content 2 (see Figure, Supplemental Digital Content 2, Consort diagram indicating sample sizes, and have been presented previously (35). Participant recruitment and screening commenced in April 2008, and final data collection was completed in March 2013.

Demographic and baseline characteristics did not differ among groups (Table 1). Body fat percentage was 45% for women and 35% for men, which is well above the recommended ranges of 20%–32% for women and 10%–22% for men (1). Energy intake and physical activity energy expenditure were within the typical ranges for overweight inactive adults. BMD T-scores indicate that most participants were above the −1.0 threshold for osteopenia. With the exception of hip and wrist BMD in the CR group, T-scores did not differ from normal values for young sex-matched young control subjects values (i.e., T-score of zero). One participant (female, CREX group) had a diagnosis of osteoporosis before enrollment. Eight were on stable doses of bone-affecting medications throughout the study, including thyroid medications (n = 4), transdermal or vaginal estrogen (n = 3), and antiresorptive medication (n = 1). Baseline V˙O2max was 8.2% ± 2.8% below age- and sex-specific predicted values (Table 1; P = 0.005 vs 100%) as expected for sedentary and overweight adults.

Baseline characteristics of study participants.

Dietary Intakes and Physical Activity

Energy intake, energy expenditure, and V˙O2max data have been reported previously as indicators of intervention compliance (35). Dietary intakes of energy, protein (g·d−1), carbohydrate, and fat decreased in the CR and CREX groups and did not change in the EX group (Table 2). When expressed relative to body mass, protein intake decreased in the CR group but did not change significantly in the CREX or EX groups. Calcium intake decreased by ~11% with all groups combined, with no significant difference among groups. Vitamin D decreased in the CREX group only (Table 2).

Dietary intakes in response to CR, EX, or a combination of CR and EX (CREX).

Exercise energy expenditure recorded with the HR monitors was 412 ± 26 kcal·d−1 in the EX group and 217 ± 23 kcal·d−1 in the CREX group (EX vs CREX P < 0.0001). Respective exercise frequencies were 8 ± 1 and 6 ± 1 sessions per week (P = 0.08), exercise duration was 7.4 ± 0.5 and 4.4 ± 0.5 h·wk−1 (P = 0.0002), and exercise intensity was 77% ± 1% and 74% ± 1% of HRmax (P = 0.17). Total energy expenditure increased in the EX (185 ± 53 kcal·d−1, P = 0.001) and CREX (126 ± 48 kcal·d−1, P = 0.01) groups and did not change in the CR group (−22 ± 51 kcal·d−1, P = 0.66) (between-group P = 0.02).

Body Weight and Lean Mass

As reported previously (35), body weight decreased by ~7% and fat mass decreased by ~15% in all three groups, with no difference among groups (P = 0.56; Table 3) as was intended by design. The average time to reach the weight loss goal was 16.8 ± 1.1 wk. Whole-body lean mass decreased in the CR group, with no change in the EX (P = 0.68) and CREX groups (P = 0.44) (Table 3). Lower extremity lean mass decreased by ~4% in the CR group and by ~2% in the CREX group and did not change in the EX group (P = 0.30). The lean mass of the upper extremities decreased by 2% with all groups combined, with no difference among groups (P = 0.29) (Table 3). The reductions in total, appendicular, and leg lean mass in the CR group remained significant, even when the change in dietary protein intake (g·d−1) was included as a covariate (data not shown).

Changes in lean mass and BMD in response to weight loss induced by CR, EX, or a combination of CR and EX (CREX).

Using a stepwise multiple linear regression analysis with the change in leg lean mass as the dependent variable and the change in energy intake, change in protein intake (g·d−1), baseline leg lean mass, and sex as the independent variables, only the change in energy intake (20% of the variance explained, P = 0.001) was related to the change in leg lean mass:

On the basis of this equation, a 500-kcal·d−1 reduction in energy intake is predicted to result in a 0.5-kg (~3%) reduction in leg lean mass. Similar results were observed when changes in total or appendicular lean mass were used as the dependent variable (results not shown). None of the independent variables were significant predictors of upper extremity lean mass changes.

Muscle Strength

Absolute isometric and 60°·s−1 isokinetic muscle strength did not change in response to weight loss in any of the three study groups (Table 4). With pooled data from all groups, marginally significant decreases in peak torque were observed for high velocity (180°·s−1) isokinetic knee extension and flexion (P = 0.08 and P = 0.02, respectively); these changes did not differ among groups (both P ≥ 0.39). When pooled strength data were reported relative to body weight, strength increased by 6% for knee flexion, by 5% for flexion and extension combined, and tended to increase for knee extension (P = 0.07), with no difference in the magnitude of improvement among groups. Strength relative to lower extremity lean mass did not change in any group (see Figure, Supplemental Digital Content 3, absolute lower extremity strength, strength relative to body weight and strength relative to lean mass,

Changes in isometric and concentric isokinetic muscle strength in response to weight loss induced by CR, EX, or a combination of CR and EX (CREX).


No changes were observed in spine, hip, or whole-body BMD in any of the study groups (Table 5). Wrist BMD decreased in the EX group, but this change did not differ from the CR and CREX groups, in which no significant changes were observed. The findings were similar for BMC and if changes were expressed as percentage change from baseline (data not shown). Exclusion of the eight participants who were on stable doses of medications that affect bone did not alter these findings.

Changes BMD in response to weight loss induced by CR, EX, or a combination of CR and EX (CREX).

Physiological Responses to Maximal-Intensity Exercise

At baseline, 89% of the V˙O2max tests met the criteria for “true V˙O2max,” and at follow-up, 85% of the test met the criteria. The frequency of meeting true V˙O2max criteria did not differ among groups at baseline (P = 0.60) or follow-up (P = 0.88). Absolute V˙O2max decreased by 6% in the CR group, increased by 15% in the EX group, and did not change in the CREX group (Table 6), with the magnitude of increase in V˙O2max being correlated (r = 0.38, P = 0.007) with physical activity energy expenditure (see Figure, Supplemental Digital Content 4, for scatterplots of physical activity expenditure, Similar results were observed for V˙O2max relative to lean mass. V˙O2max relative to body weight was preserved in the CR group and increased by 22% in the EX group and by 13% in the CREX group (Table 6). Likewise, maximal exercise oxygen pulse did not change in response to CR but increased in the EX (+19%) and CREX (+7%) groups. HRmax and RERmax did not change in response to any of the interventions. As reported previously (34), resting systolic BP decreased in the CR (−8 ± 2 mm Hg, P = 0.0002) and CREX (−5 ± 2 mm Hg, P = 0.01) groups and did not change in the EX group (−1 ± 2 mm Hg, P = 0.63); however, the between-group comparison was not statistically significant (P = 0.07). Resting systolic pressure decreased in all groups (CR: −3 ± 1 mm Hg, P = 0.05; EX: −3 ± 1 mm Hg, P = 0.03; CREX: −5 ± 2 mm Hg, P = 0.001), with no difference among groups (P = 0.63). Maximal exercise systolic and diastolic BP and maximal exercise rate pressure product all decreased based on all groups combined, with no differences among groups (Table 6).

Physiologic responses to maximal treadmill exercise in response to weight loss induced by CR, EX, or a combination of CR and EX (CREX).


Results from the present study show that modest weight loss from CR reduces total and lower extremity lean mass and absolute aerobic capacity. By contrast, the same amount of weight loss induced by exercise did not alter lower extremity lean mass and increased aerobic capacity. Although weight loss induced by CR or exercise has numerous benefits, these findings suggest that exercise may be preferable to CR, at least from the viewpoint of optimizing lean mass and V˙O2max. However, the amount of exercise needed to cause weight loss was substantial (7.4 ± 0.5 h·wk−1) and may be unrealistic for many overweight individuals with a history of being inactive. Furthermore, less exercise is required to achieve a given amount of weight loss if it is combined with CR. Therefore, we sought to determine whether more modest amounts of exercise (4.4 ± 0.5 h·wk−1), when used in combination with CR, would prevent decreases in lean mass and aerobic capacity. Indeed, with matched weight loss, the combination of CR and exercise attenuated the reductions in lean mass and aerobic capacity that occurred with CR. The lean mass losses in the CREX group were approximately half of that observed in response to CR, and reductions in V˙O2max were fully prevented. However, as compared with the combination of CR and exercise, weight loss induced by exercise alone yielded greater benefits, as evidenced by complete preservation of lower extremity and whole-body lean mass and increases in V˙O2max. Taken together, these findings indicate that exercise protects against weight loss–induced reductions in lean mass and absolute aerobic capacity when used alone or in combination with CR.

No losses of muscle strength and bone were observed in response to weight loss in the present study. This was unexpected and prevents us from determining whether exercise has protective effects on these parameters during weight loss. The SE of repeated isokinetic strength measures is ~6–13 N·m or 6%–9% (18,29), indicating good reliability; nonetheless, these tests may not have been sensitive enough to detect small reductions in strength that may have occurred as a consequence of the small 2%–4% decrease in lower extremity lean mass. On the basis of a recent meta-analysis of five trials (n = 98 participants), moderate CR with 6.2 ± kg weight loss reduces isokinetic knee extensor strength by 5.2% (P < 0.0001) but does not alter isokinetic knee flexor strength or handgrip strength (38). The lack of change in BMD and BMC in the CR group was also unexpected, especially in light of significant reductions in dietary protein and calcium. One explanation for this finding is that bone is slow to adapt (28) and the relatively short time frame for weight loss in the present study was not sufficient for measurable bone adaptations. A recent meta-analysis reported that hip BMD does not decrease during 3 months of CR-induced weight loss (based on three studies) but does decrease by ~1.0%–1.5% with interventions lasting ≥6 months (39). The same report indicated that spine and total body BMD do not decrease with weight loss interventions lasting up to 2 yr (39). However, in our previous study of 12 months of weight loss, we observed ~2% reductions in hip and spine BMD, and most of these changes occurred in the first 3 months with little or no change thereafter (33). This earlier study was remarkably similar to the present study in that the participants were middle-age (57 ± 3 yr) women and men, physically inactive at baseline, and were nonobese with a similar baseline BMI (27 ± 2 kg·m−2); furthermore, the degree of CR was similar to that used in the present study, with a goal of 16%–20% CR during the first 3–6 months of the intervention. An alternate explanation for the lack of change in BMD in the present study is that the magnitude of weight loss (~7%) may have been below a critical threshold for bone loss. Accordingly, correlation data from our previous study showed that bone losses occurred only when weight loss exceeded ~8%–10% (33).

Although it was not a study objective to identify causes of lean mass loss from CR, our data may provide some insights. One cause for CR-induced reductions in lean mass may be a modest disuse effect. As body weight decreases, the force-generating demands on muscle for performing activities of daily living also decrease, which results in skeletal muscle atrophy. This effect would be expected in weight-bearing muscles of the lower extremities, with little or no effect on upper extremity muscle. Accordingly, we observed CR-specific decreases in whole-body and lower extremity lean mass. These losses were prevented by exercise, likely because exercise put demands on muscle, which compensated for the reductions in demands associated with lower body weight. Another explanation for lean mass losses in response to CR might be the decrease in dietary protein that was observed in the CR group but not in the EX or CREX groups. Dietary protein has been shown to influence lean mass losses during weight loss (11,19). Because the anabolic effects of dietary protein are systemic (e.g., circulating factors, including amino acids and insulin-like growth factor 1), the decreases in upper extremity lean mass would also be restricted to the CR group, which was not the case. Furthermore, the reductions in lean mass in the CR group remained significant, even when the change in dietary protein intake was included as a statistical covariate. Finally, multiple regression analyses showed that the change in energy intake, but not the change in protein intake, was a significant predictor of the lean mass losses. Taken together, these lines of evidence suggest that reductions in total energy intake were responsible for CR-induced decreases in lean mass and that the changes in protein intake had little or no effect on lean mass.

Absolute V˙O2max decreased in the CR group, increased in the EX group, and remained unchanged in the CREX group. The changes were paralleled by changes in maximal exercise oxygen pulse, whereas HRmax was unchanged. Oxygen pulse is a function of left ventricular stroke volume (SV) and arteriovenous oxygen content difference (a-vO2diff). Therefore, the observed decrease in V˙O2max in the CR group and increase in the EX group were attributable to one or both of these factors. Exercise training increases V˙O2max by increasing both SV and a-vO2diff (30,31); therefore, it is likely that both contributed to the increases in V˙O2max in the EX group. Because SV depends on left ventricular mass (21,27), and because a-vO2diff depends on skeletal muscle capillary and mitochondrial density (6), the increases in V˙O2max and maximal oxygen pulse in the EX group suggest that cardiac tissue and skeletal muscle can adapt to exercise training despite the presence of a general catabolic state from weight loss. The contributions of changes in SV and a-vO2diff to the decreases in V˙O2max in the CR group are less clear, but both may have been involved. Cardiac mass decreases with CR-induced weight loss (3), which might reduce left ventricular chamber size and SV. Furthermore, the reductions in lean mass suggest that less muscle mass is present to extract oxygen from blood during maximal exercise, which would decrease a-vO2diff.

A beneficial adaptation that was similar in all three groups was the decrease in maximal exercise BP (based on pooled data from all three groups). Because there was no change in HRmax, this translated to a lower rate pressure product (12), suggesting a favorable reduction in myocardial work and oxygen consumption during intense exercise. However, in light of the decrease in V˙O2max in the CR group, and based on the Fick equation (where V˙O2 = cardiac output x a-vO2diff), the reduction in maximal exercise BP might have resulted from a decrease in SV and cardiac output during maximal exercise. By contrast, because V˙O2max increased in the EX group, maximal SV and cardiac output likely also increased (10). This, paired with a reduction in maximal exercise BP, indicates that total peripheral resistance to blood flow during maximal exercise decreased in the EX group (total peripheral resistance = mean arterial pressure / cardiac output). In this context, the exercise group likely benefitted from greater improvements in maximal exercise hemodynamic function than the CR group.

This study has limitations. The short intervention duration may not have been sufficient for full adaptations to occur; this may explain the lack of change in muscle strength and BMD in response to CR. Second, isometric and isokinetic strength measures may not reflect strength during functional physical activities. Furthermore, the unfamiliar nature of these tests to many study participants may contribute to measurement error, thereby increasing the risk of false-negative (type II) statistical errors. Third, circulating markers of bone turnover were not measured because bone markers change rapidly compared with BMD; such measures might have useful for evaluating bone responses to the relatively short-duration intervention. Fourth, we used DXA-based measures of lean mass, which are not direct measures of muscle mass and may have more measurement error compared with magnetic resonance imaging- or computed tomography-based measures of muscle mass (9). Fifth, we did not measure maximal exercise cardiac output, SV, and a-vO2diff; therefore, we cannot make definitive conclusions about the mechanisms for CR- and/or EX-induced changes in V˙O2max. Lastly, because the exercise prescriptions were based only on exercise energy expenditure, the frequency, intensity, and duration of exercise likely varied among subjects, which may have contributed to variability in the results.


Modest, CR-induced weight loss in overweight sedentary adults decreases whole-body and lower extremity lean mass, suggesting skeletal muscle atrophy, and reduces absolute aerobic capacity, suggesting catabolic activity on the cardiovascular system. However, when the same amount of weight loss was induced by using exercise alone or a combination of CR and exercise, these outcomes were preserved and/or improved. The CR-induced decreases in lean mass and aerobic capacity are likely “physiologic” changes that are appropriate for a lighter body weight and, therefore, should not be viewed as adverse effects. However, the inclusion of exercise preserves or improves lean mass and absolute aerobic capacity, despite reductions in body weight. This likely translates to an improvement in physical function, which is especially important in the sedentary population that was studied. Furthermore, the preservation of lean mass during weight loss may help preserve resting metabolism and, consequently, reduce the risk of weight regain. Taken together, these findings reinforce the notion that exercise is an important component of weight loss programs, not only for contributing to an energy deficit but also for optimizing physical function and possibly preserving metabolism.

The authors are grateful to the study participants for their cooperation and to the staff of the NIH/ICTS Clinical Research Unit for their skilled assistance. They thank the Saint Louis University Department of Campus Recreation for providing their study participants with access to the campus fitness facilities.

This work was supported by the National Institutes of Health, grant nos. K01 DK080886, DK56341 (Nutrition and Obesity Research Center), and UL1 RR024992 (Clinical Translational Science Award). The funding sources had no involvement in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the article for publication.

The authors have no conflicts of interest to disclose. All authors have approved the final article and declare that the results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. The results of the present study do not constitute endorsement by the American College of Sports Medicine.

Trial registration: identifier NCT00777621.


1. American College of Sports Medicine. ACSM’s Guidelines for Exercise Testing and Prescription. 9th ed. Baltimore: Williams & Wilkins; 2014.
2. Armamento-Villareal R, Aguirre L, Napoli N, et al. Changes in thigh muscle volume predict bone mineral density response to lifestyle therapy in frail, obese older adults. Osteoporos Int. 2014;25(2):551–8.
3. Bosy-Westphal A, Kossel E, Goele K, et al. Contribution of individual organ mass loss to weight loss-associated decline in resting energy expenditure. Am J Clin Nutr. 2009;90(4):993–1001.
4. Bouchonville M, Armamento-Villareal R, Shah K, et al. Weight loss, exercise or both and cardiometabolic risk factors in obese older adults: results of a randomized controlled trial. Int J Obes (Lond). 2014;38(3):423–31.
5. Catenacci VA, Ogden LG, Stuht J, et al. Physical activity patterns in the National Weight Control Registry. Obesity (Silver Spring). 2008;16(1):153–61.
6. Coggan AR, Spina RJ, King DS, et al. Skeletal muscle adaptations to endurance training in 60- to 70-yr-old men and women. J Appl Physiol. 1992;72(5):1780–6.
7. Compston JE, Laskey MA, Croucher PI, Coxon A, Kreitzman S. Effect of diet-induced weight loss on total body bone mass. Clin Sci (Lond). 1992;82(4):429–32.
8. Cummings SR, Melton LJ. Epidemiology and outcomes of osteoporotic fractures. Lancet. 2002;359(9319):1761–7.
9. Delmonico MJ, Kostek MC, Johns J, Hurley BF, Conway JM. Can dual energy X-ray absorptiometry provide a valid assessment of changes in thigh muscle mass with strength training in older adults? Eur J Clin Nutr. 2008;62(12):1372–8.
10. Ehsani AA, Ogawa T, Miller TR, Spina RJ, Jilka SM. Exercise training improves left ventricular systolic function in older men. Circulation. 1991;83(1):96–103.
11. Farnsworth E, Luscombe ND, Noakes M, Wittert G, Argyiou E, Clifton PM. Effect of a high-protein, energy-restricted diet on body composition, glycemic control, and lipid concentrations in overweight and obese hyperinsulinemic men and women. Am J Clin Nutr. 2003;78(1):31–9.
12. Gobel FL, Norstrom LA, Nelson RR, Jorgensen CR, Wang Y. The rate–pressure product as an index of myocardial oxygen consumption during exercise in patients with angina pectoris. Circulation. 1978;57(3):549–56.
13. Heilbronn LK, de Jonge L, Frisard MI, et al. Effect of 6-month calorie restriction on biomarkers of longevity, metabolic adaptation, and oxidative stress in overweight individuals: a randomized controlled trial. JAMA. 2006;295(13):1539–48.
14. Heymsfield SB, Gallagher D, Visser M, Nunez C, Wang ZM. Measurement of skeletal muscle: laboratory and epidemiological methods. J Gerontol A Biol Sci Med Sci. 1995;50 Spec No:23–9.
15. Institute of Medicine, Food and Nutrition Board. Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids (Macronutrients). Washington (DC): The National Academies Press; 2005.
16. Jakicic JM, Otto AD. Physical activity considerations for the treatment and prevention of obesity. Am J Clin Nutr. 2005;82(1 Suppl):226S–9S.
17. Janssen I, Ross R. Effects of sex on the change in visceral, subcutaneous adipose tissue and skeletal muscle in response to weight loss. Int J Obes Relat Metab Disord. 1999;23(10):1035–46.
18. Keskula DR, Dowling JS, Davis VL, Finley PW, Dell’omo DL. Interrater reliability of isokinetic measures of knee flexion and extension. J Athl Train. 1995;30(2):167–70.
19. Layman DK, Boileau RA, Erickson DJ, et al. A reduced ratio of dietary carbohydrate to protein improves body composition and blood lipid profiles during weight loss in adult women. J Nutr. 2003;133(2):411–7.
20. Nicklas BJ, Wang X, You T, et al. Effect of exercise intensity on abdominal fat loss during calorie restriction in overweight and obese postmenopausal women: a randomized, controlled trial. Am J Clin Nutr. 2009;89(4):1043–52.
21. Pluim BM, Zwinderman AH, van der LA, van der Wall EE. The athlete’s heart. A meta-analysis of cardiac structure and function. Circulation. 2000;101(3):336–44.
22. Rice B, Janssen I, Hudson R, Ross R. Effects of aerobic or resistance exercise and/or diet on glucose tolerance and plasma insulin levels in obese men. Diabetes Care. 1999;22(5):684–91.
23. Riedt CS, Cifuentes M, Stahl T, Chowdhury HA, Schlussel Y, Shapses SA. Overweight postmenopausal women lose bone with moderate weight reduction and 1 g/day calcium intake. J Bone Miner Res. 2005;20(3):455–63.
24. Salamone LM, Cauley JA, Black DM, et al. Effect of a lifestyle intervention on bone mineral density in premenopausal women: a randomized trial. Am J Clin Nutr. 1999;70(1):97–103.
25. Sallis JF. Seven-Day Physical Activity Recall. Med Sci Sports Exerc. 1997;29(6 Suppl):S89–103.
26. Santanasto AJ, Glynn NW, Newman MA, et al. Impact of weight loss on physical function with changes in strength, muscle mass, and muscle fat infiltration in overweight to moderately obese older adults: a randomized clinical trial. J Obes. 2011;2011. pii:516576, 10 pages.
27. Scharhag J, Schneider G, Urhausen A, Rochette V, Kramann B, Kindermann W. Athlete’s heart: right and left ventricular mass and function in male endurance athletes and untrained individuals determined by magnetic resonance imaging. J Am Coll Cardiol. 2002;40(10):1856–63.
28. Shapses SA, Sukumar D. Bone metabolism in obesity and weight loss. Annu Rev Nutr. 2012;32:287–309.
29. Sole G, Hamren J, Milosavljevic S, Nicholson H, Sullivan SJ. Test–retest reliability of isokinetic knee extension and flexion. Arch Phys Med Rehabil. 2007;88(5):626–31.
30. Spina RJ, Ogawa T, Martin WH 3rd, Coggan AR, Holloszy JO, Ehsani AA. Exercise training prevents decline in stroke volume during exercise in young healthy subjects. J Appl Physiol. 1992;72(6):2458–62.
31. Stratton JR, Levy WC, Cerqueira MD, Schwartz RS, Abrass IB. Cardiovascular responses to exercise. Effects of aging and exercise training in healthy men. Circulation. 1994;89(4):1648–55.
32. Villareal DT, Chode S, Parimi N, et al. Weight loss, exercise, or both and physical function in obese older adults. N Engl J Med. 2011;364(13):1218–29.
33. Villareal DT, Fontana L, Weiss EP, et al. Bone mineral density response to caloric restriction-induced weight loss or exercise-induced weight loss: a randomized controlled trial. Arch Intern Med. 2006;166(22):2502–10.
34. Weiss EP, Albert SG, Reeds DN, et al. Effects of matched weight loss from calorie restriction, exercise, or both on cardiovascular disease risk factors: a randomized intervention trial. Am J Clin Nutr. 2016;104(3):576–86.
35. Weiss EP, Albert SG, Reeds DN, et al. Calorie restriction and matched weight loss from exercise: independent and additive effects on glucoregulation and the incretin system in overweight women and men. Diabetes Care. 2015;38(7):1253–62.
36. Weiss EP, Racette SB, Villareal DT, et al. Lower extremity muscle size and strength and aerobic capacity decrease with caloric restriction but not with exercise-induced weight loss. J Appl Physiol. 2007;102(2):634–40.
37. Weiss EP, Racette SB, Villareal DT, et al. Improvements in glucose tolerance and insulin action induced by increasing energy expenditure or decreasing energy intake: a randomized controlled trial. Am J Clin Nutr. 2006;84(5):1033–42.
38. Zibellini J, Seimon RV, Lee CM, Gibson AA, Hsu MS, Sainsbury A. Effect of diet-induced weight loss on muscle strength in adults with overweight or obesity—a systematic review and meta-analysis of clinical trials. Obes Rev. 2016;17(8):647–63.
39. Zibellini J, Seimon RV, Lee CM, et al. Does diet-induced weight loss lead to bone loss in overweight or obese adults? A systematic review and meta-analysis of clinical trials. J Bone Miner Res. 2015;30(12):2168–78.


© 2017 American College of Sports Medicine