Abstract
Background
Muscle mass declines with age, while body adiposity increases. Sarcopenic obesity has been proposed to be particularly deleterious. However, previous methods for estimating muscle mass have been inadequate, and the relative contributions of total body fat versus muscle fat to adverse outcomes have been unclear.
Method
In a large cohort of older men (N = 1 017), we measured muscle mass (D3-creatine dilution), muscle density (high-resolution peripheral quantitative computed tomography in the diaphyseal tibia) as a proxy of muscle fat, and total body fat (dual-energy x-ray absorptiometry). We examined their associations with physical performance (walking speed, grip strength, chair stand time), the risk of mobility outcomes (mobility limitations, mobility disability), and the risk of death over ~5 years.
Results
In combined models, lower muscle mass and muscle density were independently associated with worse physical performance and the risk of adverse outcomes, while total body fat was minimally related to physical performance and not related to mobility outcomes or mortality. For example, the relative risks for mortality per 1 standardized unit increase in muscle density, muscle mass, and total body fat were 0.84 (95% confidence interval: 0.74, 0.96), 0.70 (0.57, 0.86), and 0.90 (0.64, 1.25), respectively.
Conclusions
Muscle mass and muscle density were associated with physical performance and adverse outcomes, and had independent, additive effects. There was little additional contribution of total body fat.
Keywords: Obesity, Outcomes, Physical performance, Sarcopenia
Sarcopenia, or reductions in the amount of muscle and in muscle quality (strength and endurance per unit of muscle mass), appears to be a component of the age-related decrement in physical performance (1) and is an important contributor to the decrease in physical function that occurs with aging (2). Sarcopenia has major adverse public health implications including injurious falls, fractures, disability, and mortality (3–5).
Total body fat mass increases with age. The combination of decreased muscle mass/function and increased body fat (sarcopenic obesity) has been proposed as a particularly strong risk factor for decreased physical performance, increased fall risk, functional decline, disability, and mortality (6–8), and has been postulated to be a major health problem (9). The underlying mechanisms that contribute to reduced physical performance and adverse outcomes in sarcopenic obesity are likely complex. One possibility is that obesity is associated with an accumulation of muscle fat. Higher intermuscular adipose tissue and intramyocellular lipid content, or myosteatosis, is reflected in lower muscle density by computed tomography (CT) (10). Myosteatosis increases with age and is consistently associated with lower muscle performance, mobility limitation, and mortality in older adults (11–15). Yet, total body fat may not be an adequate reflection of muscle lipid and the degree to which whole body fat is linked to myosteatosis is unclear. Moreover, whether myosteatosis and increased total body fat are similarly related to adverse outcomes has not been adequately described.
Moreover, most previous evaluations of sarcopenic obesity have been problematic as they relied on lean mass measures (9), which are only a gross approximation of muscle mass (16–18). We recently reported that when muscle mass is accurately measured using D3-creatine (D3Cr) dilution, the additional contribution of total body fat to performance and adverse outcomes is minimal (19). In view of this apparently limited effect of total body fat on physical function in older people, it is important to assess muscle lipid content more directly. For both research and clinical purposes, it is important to establish an understanding of the relative associations of muscle mass and muscle lipid, and total body fat, with muscle dysfunction and physical performance. That knowledge would both enable focused studies of causation as well as improve the ability to predict those at risk of poor outcomes.
In the current analysis, we examined the relationships between muscle density assessed by high-resolution peripheral quantitative computed tomography (HR-pQCT) in the diaphyseal tibia, muscle mass (using D3Cr dilution), and total body fat by dual-energy x-ray absorptiometry (DXA), and evaluated their independent associations with several measures of physical performance and adverse outcomes in the Osteoporotic Fractures in Men Study (MrOS), a longitudinal study of musculoskeletal health in older men.
Method
MrOS Cohort and Study Sample
The MrOS cohort of 5 994 men aged 65 years and older was initially recruited in 2000–2002. The study design and recruitment approaches have been described (20,21) (https://mrosonline.ucsf.edu). In 2014–2016, 2 786 surviving MrOS participants were contacted; 1 841 participated in a clinic visit, and 1 641 agreed to the D3Cr dilution protocol. Of these, 187 were excluded for incorrect completion of the protocol. As described in Cawthon et al. (22), the exclusions were due to incorrect timing of the dose or urine collection (either <72 hours or >144 hours between the dose and collection) or forgetting to take the dose or provide the specimen. Six samples were lost by the clinical center or laboratory and 23 men were excluded because of outlying values for D3Cr muscle mass/body mass more than 2 SD from the mean, most of which included values that exceeded 100% of body mass. Further, 408 were missing the CT variables. Of the remaining 1 017, all but one man had data for one or more of the cross-sectional outcomes (grip strength, walking speed, chair stands). The data for the mobility outcomes was gathered on an interim visit questionnaire 2.2 ± 0.3 years later; 894 men returned these questionnaires. Men with prevalent outcomes at the initial timepoint were not included in the analyses of incident outcomes, leaving n = 714 in the incident mobility limitation analyses and n = 832 in the incident mobility disability analyses. All 1 107 men were included in the analysis of mortality. These groups are summarized in Supplementary Figure 1.
Muscle Density by HR-pQCT
HR-pQCT images were acquired spanning a 10.2-mm section (168 slices with a 60.7-µm slice thickness) of the nondominant tibia (XtremeCT II, Scanco Medical AG, Brüttisellen, Switzerland) (23). The scan volume was centered at a distance 30% of tibial length, proximal to the distal end of the tibia. Muscle density was measured using Soft Tissue Analysis (Scanco Medical AG) adapted for second-generation HR-pQCT images. Briefly, this software automatically segments the muscle volume of interest, excluding the skin, subcutaneous fat, bone, extraosseal calcifications, and large fat deposits outside the muscle compartments (24). The mean grayscale value for the muscle volume of interest (ie, muscle density) was calculated in units of bone mineral density (mg/cm3). A standardized bone mineral density phantom was scanned previously to derive fixed linear grayscale-to-density calibration coefficients (25). The phantom was scanned on a daily basis to monitor stability of the density calibration. Supplementary Figure 2 shows the anatomical site of the HR-pQCT diaphyseal tibial scans used for the muscle density analyses. The cross-sectional images illustrate the muscle compartments identified, where muscle volume and muscle density are measured.
D3Cr Dilution Method to Estimate Muscle Mass
As described previously (26), the method involves the subject ingesting a 30-mg dose of stable isotope (deuterated) labeled creatine (D3Cr), and providing a fasting, morning urine sample 3–6 days later. Urine D3-creatinine, unlabeled creatinine, and creatine are measured using high-performance liquid chromatography and tandem mass spectrometry; these measures are then included in an algorithm to determine total body creatine pool size and thus skeletal muscle mass (26). The D3Cr dilution method has been extensively validated and is correlated with magnetic resonance imaging-derived muscle volume at r = .87, p < .001 in humans (27). In a previously published study (26), repeat measurements were made on urine samples from participants with wide range of ages revealing a coefficient of variation of 3.6%. Moreover, very small longitudinal changes in D3Cr muscle mass can be detected in infants (28), further demonstrating the precision of the technique.
Dual-Energy X-Ray Absorptiometry
Total body lean mass, appendicular lean mass, fat, and bone mineral content were assessed by whole-body DXA scans (Hologic 4500 scanners, Waltham, MA) (29).
Physical Performance, Incident Mobility Problems, and Mortality
During the clinic visit, walking speed at usual pace was measured over a 6-m course using the average of 2 trials (m/s), time to complete 5 repeated chair stands was assessed, and grip strength was measured using a Jamar dynamometer (Sammons Preston Rolyan, Bolingbrook, IL) (30). Grip strength was tested twice with each hand and the maximum of the 4 values was used in these analyses. During follow-up, MrOS participants or their surrogates were regularly contacted by postcard questionnaire and/or phone contact. On the questionnaire administered 2.2 ± 0.3 years after the Visit 4, participants answered questions about difficulty walking 2–3 blocks, climbing 10 stairs, or carrying or lifting 10 lbs. We defined mobility limitation as any new self-reported difficulty walking 2–3 blocks or climbing 10 steps or carrying or lifting 10 lb; mobility disability was defined as inability to do any of these tasks. Except in 55 cases in which documentation was pending, deaths were confirmed with death certificates and additionally with discharge summaries when available. The follow-up time for the mortality outcome was 4.9 ± 1.4 years (range 2 days to 6.6 years).
Statistical Analysis
Characteristics of the participants were compared across quartile of tibial muscle density using chi-squared tests for homogeneity categorical variables, analysis of variance for normally distributed continuous variables, and Kruskal–Wallis tests for continuous variables with skewed distributions.
Pearson partial correlations between measures of D3Cr muscle mass, total body fat, and muscle density were calculated, adjusting for age, clinic site, height, and weight. In all statistical models, the independent variables used were standardized [(observed value - mean)/standard deviation] to allow for comparisons of effect sizes. Linear regression was used to assess the relationship of the 3 predictors (D3Cr muscle mass, muscle density, and total body fat) and the outcomes of walking speed, grip strength, and the number of chair stands completed in 10 seconds, with results presented as beta coefficients and their 95% confidence intervals (CIs). Negative binominal regression was used to assess the relationship of measures of D3Cr muscle mass, total body fat, and muscle density with the risk of incident mobility limitations, and incident mobility disability. Results are presented as relative risks (RRs) and their 95% CIs per 1 standardized unit increase. Proportional hazards models were used to assess the relationship of measures of D3Cr muscle mass, total body fat, and muscle density with the risk of all-cause mortality, with results shown as hazard ratios (HRs) and their 95% CIs per 1 standardized unit increase.
A series of models were constructed for each outcome. All models were adjusted for age, clinic site, height, and weight. The first set of models (model 1) examined the association of each of the 3 predictors with the outcomes, with each predictor in a separate model. Model 2 had 2 of the predictors in each model, with muscle density paired with D3Cr muscle mass. Model 3 had D3Cr muscle mass, muscle density, and total body fat in the same model. In additional analyses, muscle area was added to model 3 to assess its independent effects. Finally, model 3 was further adjusted for race, self-reported health status, alcohol use, smoking status, comorbidity index, and oral corticosteroid use. Variance inflation factors (VIFs) were examined to verify that the multicollinearity of all variables in the model was acceptable. When all variables were included in the same model (model 3), the highest VIF was 8.77 for the variable weight. VIF values are shown in Supplementary Table 1.
Because muscle density is the most novel element of these analyses, participant characteristics are described in Table 1 by quartiles of muscle density. Potential interactions between pairs of independent variables were examined by including a term for their interaction plus the main effects (all as continuous variables) in models adjusted by age, clinic site, height, and weight. Interactions examined were muscle density with total body fat, muscle density with D3Cr muscle mass, and D3Cr muscle mass with total body fat. These potential interactions were also examined graphically by categorizing the independent variables into quartiles and examining the percentage of each event within each quartile combination. The continuous outcomes of grip strength, walking speed, and number of chair stands completed in 10 seconds were categorized (lowest quartile vs the other quartiles) for these plots.
Table 1.
Characteristics [mean ± SD or n (%)] of the MrOS Participants by Quartile of Tibial Muscle Density (mg/cm3)
| Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | ||
|---|---|---|---|---|---|
| <9.585 | 9.585 to <12.845 | 12.845 to <15.190 | ≥15.190 | ||
| Characteristics | (n = 254) | (n = 254) | (n = 254) | (n = 255) | p Value |
| Age, y | 85.7 ± 4.4 | 84.7 ± 4.0 | 83.6 ± 3.9 | 82.6 ± 3.1 | <.001 |
| Race, non-White | 15 (5.9) | 20 (7.9) | 40 (15.7) | 37 (14.5) | <.001 |
| Alcohol intake, drinks/wk | .610 | ||||
| <1 | 121 (47.6) | 114 (45.1) | 121 (47.6) | 115 (45.5) | |
| 1–13 | 117 (46.1) | 127 (50.2) | 116 (45.7) | 129 (51.0) | |
| 14+ | 16 (6.3) | 12 (4.7) | 17 (6.7) | 9 (3.6) | |
| Smoking status | .252 | ||||
| Never smoker | 90 (35.4) | 98 (38.6) | 105 (41.3) | 101 (39.6) | |
| Past or current smoker | 149 (58.7) | 144 (56.7) | 145 (57.1) | 141 (55.3) | |
| Missing smoking data | 15 (5.9) | 12 (4.7) | 4 (1.6) | 13 (5.1) | |
| Number of comorbiditiesa | .002 | ||||
| 0 | 147 (57.9) | 169 (66.5) | 170 (66.9) | 188 (73.7) | |
| 1 | 72 (28.3) | 57 (22.4) | 65 (25.6) | 55 (21.6) | |
| 2+ | 35 (13.8) | 28 (11.0) | 19 (7.5) | 12 (4.7) | |
| Oral corticosteroid use | 5 (2.0) | 9 (3.5) | 8 (3.1) | 5 (2.0) | .582 |
| Physical activity score (PASE) | 105.4 ± 67.0 | 110.7 ± 60.1 | 130.0 ± 65.0 | 131.9 ± 56.7 | <.001 |
| Walking speed, m/s | 0.97 ± 0.23 | 1.07 ± 0.23 | 1.15 ± 0.21 | 1.18 ± 0.21 | <.001 |
| Number of chair stands in 10 s | 3.0 ± 1.8 | 3.6 ± 1.6 | 4.2 ± 1.5 | 4.7 ± 1.5 | <.001 |
| Maximum grip strength, kg | 33.0 ± 7.6 | 34.5 ± 7.8 | 36.6 ± 7.7 | 38.4 ± 7.2 | <.001 |
| Weight, kg | 80.9 ± 12.7 | 79.1 ± 11.2 | 77.5 ± 11.5 | 75.3 ± 9.2 | <.001 |
| Height, cm | 171.8 ± 6.6 | 171.6 ± 6.3 | 171.5 ± 6.6 | 171.9 ± 6.1 | .894 |
| Body mass index, kg/m2 | 27.4 ± 4.0 | 26.8 ± 3.5 | 26.3 ± 3.4 | 25.5 ± 2.6 | <.001 |
| Excellent/good self-rated health | 214 (84.3) | 227 (89.4) | 235 (92.5) | 240 (94.5) | <.001 |
| D3Cr muscle mass/weight | 0.28 ± 0.04 | 0.30 ± 0.04 | 0.32 ± 0.04 | 0.34 ± 0.04 | <.001 |
| D3Cr muscle mass, kg | 22.5 ± 3.7 | 23.6 ± 4.0 | 24.5 ± 3.9 | 25.3 ± 3.5 | <.001 |
| Total body fat, kg | 23.5 ± 7.7 | 22.4 ± 6.6 | 21.1 ± 6.4 | 18.8 ± 5.5 | <.001 |
| Percent total body fat | 28.9 ± 5.9 | 28.3 ± 5.7 | 27.3 ± 5.3 | 25.0 ± 5.3 | <.001 |
| Appendicular lean mass (kg)/height (m)2 | 7.6 ± 0.9 | 7.5 ± 0.9 | 7.5 ± 0.8 | 7.5 ± 0.8 | .617 |
| Tibial muscle density, mg/cm3 | 5.8 ± 3.5 | 11.2 ± 0.9 | 14.0 ± 0.6 | 16.9 ± 1.3 | <.001 |
| Prevalent mobility limitation | 105 (41.3) | 58 (22.9) | 38 (15.0) | 18 (7.1) | <.001 |
| Prevalent dismobility | 46 (18.1) | 21 (8.3) | 12 (4.7) | 2 (0.8) | <.001 |
| Incident mobility limitationb | 48 (37.8) | 51 (29.1) | 48 (24.4) | 26 (12.1) | <.001 |
| Incident dismobilityb | 31 (17.3) | 29 (14.0) | 14 (6.5) | 9 (3.9) | <.001 |
| Follow-up for mobility outcomes, y | 2.2 ± 0.3 | 2.2 ± 0.3 | 2.2 ± 0.3 | 2.2 ± 0.3 | .994 |
| Mortality | 101 (39.8) | 69 (27.2) | 60 (23.6) | 43 (16.9) | <.001 |
| Follow-up for mortality, y | 4.5 ± 1.6 | 4.8 ± 1.4 | 5.0 ± 1.3 | 5.1 ± 1.2 | <.001 |
Notes: MrOS = Osteoporotic Fractures in Men Study; PASE = Physical Activity Scale for the Elderly. p Values for continuous variables are from analysis of variance for normally distributed variables or from a Kruskal–Wallis test for skewed data, and p values for categorical data are from a chi-squared test for homogeneity.
aComorbidities include congestive heart failure, chronic obstructive pulmonary disease, diabetes mellitus, and myocardial infarction.
bThose with prevalent mobility disability were excluded from the incident mobility disability variable; similar for mobility limitation.
All significance levels were 2-sided, and all analyses were conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC).
Results
Cohort Characteristics
The characteristics of the men included in these analyses are shown in Table 1, by quartiles of muscle density. The men were older (average age 84.2 ± 4 years, range 77–101) and most reported good or excellent health. Characteristics of the MrOS cohort at this clinic visit by quartiles of D3Cr muscle mass have been published elsewhere (22). In the analyzed cohort, there was a high proportion of men who were overweight or obese; there were 68% with %fat >25 (a common threshold for obesity in men), and 32% had %fat >30%. Fifty-three % had body mass index (BMI) >26 and 13% had BMI >30. Approximately 22% had some mobility limitation, but few had mobility disability. However, during follow-up, an additional 24% and 10% reported mobility limitation or disability, respectively. For the analysis of mortality, 273 had died and 744 were still living.
At baseline, men with lower muscle density were slightly older and more often White, were less often in good/excellent health, had higher BMI and body fat, and had lower D3Cr muscle mass (but not lower DXA appendicular lean mass) (Table 1). In addition, they had lower Physical Activity Scale for the Elderly scores, performed less well on tests of physical performance and more often had prevalent mobility limitation.
Relationships Between Tibial Muscle Density, D3Cr Muscle Mass, and Total Body Fat
Muscle density was not strongly related to total body fat or D3 muscle mass (Figure 1). The correlations (adjusted for age, clinic site, height, and weight) were r = .34 and r = −.12, respectively. D3Cr muscle mass and total body fat were moderately correlated (r = −.49) (p < .05 for all).
Figure 1.
The correlations of (A) tibial muscle density and D3Cr muscle mass and (B) tibial muscle density and total body fat. Pearson correlations with age, clinic site, height, and weight as the partial variables.
Associations of Muscle Density and Mass Measures With Physical Performance
When examined individually (adjusted for age, clinic site, height, and weight), higher muscle density and higher D3Cr muscle mass were both associated with better performance in all 3 measures of physical performance (walking speed, grip strength, and chair stands), and the standardized beta coefficients appeared to be of similar magnitude (Table 2, model 1). Moreover, when included in a model together, the associations of D3Cr muscle mass and muscle density with physical performance each remained significant (Table 2, model 2) (Figure 2). There were no interactions between muscle density and D3Cr muscle mass on grip strength and walk speed (interaction terms p > .05), so their effects appeared to be additive. The significant interaction term (p = .04) between muscle density and D3Cr muscle mass in their associations with chair stands suggests that the effects of low muscle density were most marked in those men with low D3Cr muscle mass and vice versa.
Table 2.
Associations of Muscle and Fat Measures With Physical Performance and Incident Outcomes
| Model 1: Independent Associationsa | Model 2: Bivariate Associationsb | Model 3: Associations Including 3 Measuresc | ||||||
|---|---|---|---|---|---|---|---|---|
| Tibial Muscle Density (mg/cm3) | D3Cr Muscle Mass (kg) | Total Body Fat (kg) | Tibial Muscle Density (mg/cm3) | D3Cr Muscle Mass (kg) | Tibial Muscle Density (mg/cm3) | D3Cr Muscle Mass (kg) | Total Body Fat (kg) | |
| Physical performance, β coefficient (95% CI) | ||||||||
| Walking speed (m/s) | 0.06 (0.05, 0.08)† | 0.08 (0.06, 0.09)† | –0.06 (–0.09, –0.03)† | 0.05 (0.03, 0.06)† | 0.06 (0.04, 0.07)† | 0.05 (0.03, 0.06)† | 0.06 (0.04, 0.08)† | 0.01 (–0.026, 0.04) |
| Maximum grip strength (kg) | 1.61 (1.17, 2.06)† | 2.28 (1.77, 2.79)† | –3.22 (–4.15, –2.29)† | 1.05 (0.59, 1.52)† | 1.85 (1.31, 2.39)† | 1.10 (0.63, 1.57)† | 1.37 (0.75, 1.99)† | –1.73 (–2.77, –0.68)† |
| Chair stands (N/10 s) | 0.61 (0.50, 0.71)† | 0.72 (0.60, 0.84)† | –0.76 (–0.98, –0.54)† | 0.45 (0.34, 0.55)† | 0.54 (0.42, 0.66)† | 0.43 (0.33, 0.54)† | 0.49 (0.35, 0.63)† | –0.23 (–0.46, 0.005) |
| Incident outcomes, relative risk (95% CI) | ||||||||
| Mobility limitation | 0.75 (0.64, 0.89)† | 0.76 (0.62, 0.93)‡ | 1.43 (1.01, 2.02)# | 0.79 (0.66, 0.94)‡ | 0.82 (0.66, 1.01) | 0.78 (0.66, 0.93)‡ | 0.86 (0.67, 1.11) | 1.20 (0.80, 1.81) |
| Mobility disability | 0.71 (0.57, 0.87)‡ | 0.54 (0.40, 0.72)† | 1.36 (0.83, 2.22) | 0.80 (0.64, 1.01) | 0.59 (0.43, 0.80)† | 0.82 (0.65, 1.04) | 0.53 (0.36, 0.78)‡ | 0.76 (0.42, 1.34) |
| All-cause mortality, hazard ratio (95% CI) |
0.76 (0.68, 0.86)† | 0.68 (0.58, 0.80)† | 1.27 (0.95, 1.70) | 0.82 (0.72, 0.93)‡ | 0.74 (0.62, 0.88)† | 0.84 (0.74, 0.96)‡ | 0.70 (0.57, 0.86)† | 0.90 (0.64, 1.25) |
Notes: Beta coefficients, relative risks, or hazard ratios and their 95% confidence intervals (CIs), per 1 unit standardized increase.
aEach association adjusted for age, clinic site, height, and weight.
bAdjusted for age, clinic site, height, weight, and including tibial muscle density and D3Cr muscle mass.
cAdjusted for age, clinic site, height, weight, and including D3Cr muscle mass, tibial muscle density, and total body fat measures.
† p ≤ .001. ‡p < .01. #p < .05.
Figure 2.
The percent of men with (A) the lowest quartile of walking speed as a function of tibial muscle density and D3Cr muscle mass, (B) the lowest quartile of grip strength as a function of tibial muscle density and D3Cr muscle mass, and (C) the lowest quartile of chair stands as a function of tibial muscle density and D3Cr muscle mass.
Associations of Muscle and Fat Measures With Physical Performance
While higher muscle density and D3Cr muscle mass were strongly associated with better physical performance measures, higher fat mass was associated with lower performance in all tests (Table 2, model 1). However, when included in a model together with muscle density and D3Cr muscle mass, the associations of total body fat with performance were reduced and remained significant only for grip strength (Table 2, model 3). When total body fat was included in bivariate models with either muscle density or D3Cr muscle mass, its associations with physical performance measures were attenuated; D3Cr muscle mass appeared to attenuate the associations of total body fat to a greater degree than did muscle density (Supplementary Table 2).
There was a significant interaction between muscle density and total body fat in their associations with grip strength (p for interaction = .04, Supplementary Figures 3–5)—a borderline interaction for the chair stands measure (p for interaction = .10) but no significant interaction for the walking speed assessment. There was the suggestion of an interaction between D3Cr muscle mass and total body fat for the chair stands measure (p for interaction = .07), but no significant interaction for the grip strength or walking speed (p for interaction > .10 for both).
Associations of Muscle Density and Mass Measures With Incident Mobility Outcomes
In models adjusted for age, clinic site, height, and weight, higher muscle density and D3Cr muscle mass were significantly associated with lower risks of incident mobility limitation and mobility disability (Table 2). When considered together (Table 2, model 2), the RRs of muscle density and D3Cr muscle mass with incident mobility outcomes were similar to model 1, indicating they had independent effects. The joint effects of muscle density and D3Cr muscle mass on mobility limitation and mobility disability are shown in Figure 3. Interaction terms were not significant. The risk of incident mobility limitation in men with the lowest D3Cr muscle mass and lowest muscle density was ~10-fold higher than those with the highest muscle density and D3Cr muscle mass. Similarly, the risk of incident mobility disability in those with the lowest D3Cr muscle mass and lowest muscle density was ~4-fold higher than those with the highest muscle density and D3Cr muscle mass.
Figure 3.
The percent of men with (A) incident mobility limitation as a function of tibial muscle density and D3Cr muscle mass and (B) incident mobility disability as a function of tibial muscle density and D3Cr muscle mass.
Associations of Muscle and Fat Measures With Incident Mobility Outcomes
When considered independently (adjusted for age, clinic site, height, and weight), higher total body fat was associated with a higher risk of incident mobility limitation and had a similar but nonsignificant association with mobility disability (Table 2, model 1). However, when included in a model together with muscle density and D3Cr muscle mass (Table 2, model 3), total body fat was not significantly associated with either outcome. When total body fat was included in bivariate models with D3Cr muscle mass, its associations with adverse mobility outcomes were attenuated, while in bivariate models with muscle density the associations were similar to those in univariate models (Supplementary Table 2).
There was an interaction between total body fat and muscle density in their associations with mobility limitation (p for interaction = .005) (Supplementary Figure 6); the men with the highest total body fat were at higher risk of mobility limitation regardless of their muscle density, while in men with the lowest total body fat, low muscle density was strongly associated with greater risk of mobility limitation. Although there was a suggestion of a similar interaction between D3Cr muscle mass and total body fat, the interactions between these measures and incident mobility disability were of borderline significance (p for interaction = .07) (Supplementary Figure 7).
Associations of Muscle and Fat Measures With Mortality
The risk of all-cause mortality was significantly lower in men with higher muscle density or D3Cr muscle mass (HR 0.76 [0.67, 0.86] and 0.67 [0.56, 0.81], respectively) (Table 2, model 1). Men with greater total body fat were at higher risk, albeit not significantly so. When considered together (Table 2, model 2), the HRs of muscle density and D3Cr muscle mass with mortality were similar and both remained significant, indicating they had independent effects (Figure 4). When muscle density, D3Cr muscle mass, and total body fat were included together (Table 2, model 3), higher muscle density and D3Cr muscle mass remained significantly associated with lower mortality, but total body fat was not a statistically significant risk factor for mortality.
Figure 4.
The percent of men who died during follow-up as a function of (A) tibial muscle density and total body fat, (B) D3Cr muscle mass and total body fat, and (C) tibial muscle density and D3Cr muscle mass.
Other Covariables of Interest
Muscle volume was a nonsignificant contributor when considered in models with muscle density and D3Cr muscle mass (data not shown). When models were further adjusted for race, self-reported health status, alcohol use, smoking status, comorbidity index, and oral corticosteroid use, the associations were essentially unchanged (data not shown).
Analyses Limited to Men With Increased Body Fat
To examine the validity of our results in men with increased body fat, we performed sensitivity analyses limited to those with %body fat >26 (a common definition of obesity in men). The associations were very similar to those in the entire cohort (Supplementary Table 3).
Discussion
In a large, well-characterized cohort of older men, we found that muscle density (primarily an index of muscle adiposity) and D3Cr muscle mass were not well correlated, indicating they reflect distinctly different characteristics of muscle. Each had strong, independent associations with better walking speed, grip strength, and chair stands. The effects of these 2 measures were at least additive, such that men with the lowest muscle density and lowest D3Cr muscle mass were more likely to have low physical performance than those with higher levels of each. These measures of physical performance have been linked to important downstream outcomes, and indeed in our cohort, those with either lower muscle density or D3Cr muscle mass tended to be at greater risk of mobility limitation and mobility disability during follow-up; those with both low D3Cr muscle mass and low density were at highest risk. Finally, we previously reported that D3Cr muscle mass is associated with increased mortality (16), and here show that lower muscle density and lower D3Cr muscle mass were each independently associated with an increased risk of all-cause mortality. When muscle density and D3Cr muscle mass were considered, total body fat was not associated with the risk of adverse mobility outcomes or mortality. These findings highlight the importance of assessing both muscle mass and muscle adiposity for the understanding of critical health outcomes.
A novel element of the current analysis is the inclusion of both CT measures of muscle density and an accurate measure of muscle mass (D3Cr). Importantly, our epidemiologic data strongly suggest that men with the same amount of functional muscle mass (D3Cr muscle mass) can have widely varying amounts of muscle lipid (muscle density) (Figure 1) and vice versa, suggesting they reflect unique and different aspects of muscle. The D3Cr dilution method is thought to be an index of functional muscle mass, as creatine (and creatine phosphate) in the sarcomere are found adjacent to the Z-disc and the A-band (31) and are not associated with noncontractile components of muscle such as lipid and fibrosis. Muscle density is thought to be a surrogate for the amount of lipid in muscle (myosteatosis), as muscle biopsy studies have demonstrated a moderate, negative correlation between lipid deposition and CT-derived muscle density (10). While muscle density is strongly related to muscle lipids, it has also been considered an overall marker of muscle quality. As assessed by CT, muscle density has been repeatedly associated with poor physical performance and adverse outcomes in older people (12–15), and the results of the current analyses support that conclusion. Although our muscle density assessment is specific to the tibial muscles, it is associated with several measures of physical function (including grip strength, walking speed, and chair stands), indicating that muscle lipid content of the tibial muscles is likely an indicator of whole-body skeletal muscle quality. Our results show that these adverse associations are independent of D3Cr muscle mass and implicate muscle density as a valuable measure of muscle health.
These analyses have implications for the consideration of sarcopenic obesity. In this large cohort of older men, many of whom were overweight or obese, the correlation of total body fat with muscle density was minimal, indicating total body fat is not a good indicator of muscle lipid content. Furthermore, we show that muscle density, primarily a reflection of muscle lipid content, was independently associated with important outcomes, while the effects of total body fat were much less prominent. Hence, our data suggest myosteatosis is more critical for the health outcomes we assessed than is total body fat. The combination of higher muscle fat and low D3Cr muscle mass appear to be particularly disadvantageous. We have previously reported that muscle mass is accurately measured by D3Cr dilution, and is strongly related to poor physical performance (eg, slower walking speed, slower chair stands performance) and a variety of deleterious outcomes, including mortality, mobility limitation, and incident disability in activities of daily living (ADLs) and instrumental ADLs (16,17,22,32). In addition, we recently reported that after D3Cr muscle mass is considered, total body fat is less strongly associated with these conditions (19), and the current analyses again support those findings. Overall, these results suggest that higher muscle lipid and low muscle mass, rather than increased total body fat, are most important in determining physical performance, mobility-related related adverse outcomes, and mortality. Thus, the concept of “sarcopenic obesity” might be replaced by a more sophisticated and specific consideration of the condition of low muscle mass and low muscle density (sarcopenia with myosteatosis).
The measurement of body composition, and how those measurements are used to evaluate age-related declines in physical performance, has been of considerable interest in both clinical and research settings. Fat mass can be conveniently and accurately measured by DXA. Muscle density, a reflection at least in part of lipid content (10), can be measured by CT; lower muscle density has been related to reduced physical performance and disability in aging, to hospitalizations and to greater fall and fracture risk (33–35). The measurement of muscle mass has been more challenging. Lean mass by DXA has been commonly used as a surrogate for muscle mass, and previous studies of the relative contributions of muscle mass, quality, and adiposity have uniformly utilized lean mass measures as a substitute for actual determinations of muscle mass. However, lean mass is not a measurement of muscle mass (18) and is only weakly associated with physical performance or important health outcomes (eg, falls, fractures, physical disability). Hence, previous studies of sarcopenia, and its relationship to adiposity, that relied on lean mass measures by DXA may be misleading.
This study has important strengths. It utilizes accurate state-of-the art measurements of muscle mass using D3Cr, muscle density using CT, and DXA total body fat. We examined a relatively large cohort of community-dwelling, older men, a group at risk of impaired physical performance and adverse health outcomes, in a longitudinal, observational study design. The cohort included a wide range of muscle measures and fat mass, allowing adequate testing of the hypotheses proposed. Study sites were experienced and assessment methods were standardized. A major limitation of our work is that that the participants were older men who were primarily Caucasian; our findings may not pertain to other groups (eg, women) and should be examined in larger, more diverse populations. Only a single measurement of body composition and physical function was performed (ie, at baseline) to predict adverse outcomes, and they may have changed over time. Also, we did not address other measures that have been proposed to reflect muscle quality (eg, mitochondrial function). We assessed muscle density and volume in the distal calf; there are few data concerning the consistency of these measures across anatomical regions (eg, calf vs thigh vs trunk). Although CT muscle density is heavily influenced by lipid content, there are likely other constituents of muscle tissue (eg, fibrosis) that may affect CT density results and that we cannot quantify without muscle tissue. While our cohort included a relatively high proportion of obesity (68% with %fat >25) and our sensitivity analyses indicate that our results were the same in them as in the entire cohort, there were relatively few with severe obesity. The relationships we describe could be different in that group. Finally, we purposefully addressed the relationship of muscle density, muscle mass, and total body fat in the context of physical performance, mobility problems, and mortality, primarily because sarcopenia and reduced muscle quality are most robustly associated with these outcomes. We did not consider other issues (eg, metabolic disturbances) that have been linked to sarcopenia and adiposity.
In summary, in a longitudinal study of a large cohort of older men, we found that D3Cr muscle mass and CT muscle density were both independently, and additively, associated with physical performance, the risk of mobility limitation and mobility disability, and the risk of mortality. Total body fat had inconsistent associations with physical performance and no independent effect on incident outcomes. These results emphasize the value of assessing both muscle mass and muscle density in the study of sarcopenia. They also suggest the concept of sarcopenic obesity should be reconsidered to include low D3Cr muscle mass and muscle density (myosteatosis), rather than total body fat.
Supplementary Material
Funding
This work is supported by National Institutes of Health funding. The following institutes provide support: the National Institute on Aging (NIA), the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS), and the National Center for Advancing Translational Sciences (NCATS), under the following grant numbers: U01 AG027810, U01 AG042124, U01 AG042139, U01 AG042140, U01 AG042143, U01 AG042145, U01 AG042168, U01 AR066160, and UL1 TR000128. Funding for the D3Cr muscle mass measure was provided by NIAMS (grant number R01 AR065268). GlaxoSmithKline provided in-kind support by providing the D3-creatine dose and analysis of urine samples. Nicolas Vilayphiou of Scanco Medical AG provided optimization support for the soft tissue analysis software.
Conflict of Interest
None declared. W.J.E. is listed as a coinventor on the granted patents for the D3Cr dilution method. However, he does not derive any income or royalties or own the intellectual property for the method.
Author Contributions
Study concept and design: E.S.O., W.J.E., and P.M.C. Acquisition of subjects and/or data: E.S.O., S.R.C., J.A.C., A.R.H., and P.M.C. Analysis and interpretation of data: All the authors. Preparation of the initial manuscript: E.S.O.
References
- 1. Chiles Shaffer N, Fabbri E, Ferrucci L, Shardell M, Simonsick EM, Studenski S. Muscle quality, strength, and lower extremity physical performance in the Baltimore Longitudinal Study of Aging. J Frailty Aging. 2017;6(4):183–187. doi:10.14283/jfa.2017.24 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Larsson L, Degens H, Li M, et al. Sarcopenia: aging-related loss of muscle mass and function. Physiol Rev. 2019;99(1):427–511. doi:10.1152/physrev.00061.2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Legrand D, Vaes B, Matheï C, Adriaensen W, Van Pottelbergh G, Degryse JM. Muscle strength and physical performance as predictors of mortality, hospitalization, and disability in the oldest old. J Am Geriatr Soc. 2014;62(6):1030–1038. doi:10.1111/jgs.12840 [DOI] [PubMed] [Google Scholar]
- 4. Abellan van Kan G. Epidemiology and consequences of sarcopenia. J Nutr Health Aging. 2009;13(8):708–712. doi:10.1007/s12603-009-0201-z [DOI] [PubMed] [Google Scholar]
- 5. Berger MJ, Doherty TJ. Sarcopenia: prevalence, mechanisms, and functional consequences. Interdiscip Top Gerontol. 2010;37:94–114. doi:10.1159/000319997 [DOI] [PubMed] [Google Scholar]
- 6. Zoico E, Di Francesco V, Mazzali G, et al. High baseline values of fat mass, independently of appendicular skeletal mass, predict 2-year onset of disability in elderly subjects at the high end of the functional spectrum. Aging Clin Exp Res. 2007;19(2):154–159. doi:10.1007/BF03324682 [DOI] [PubMed] [Google Scholar]
- 7. Zamboni M, Mazzali G, Fantin F, Rossi A, Di Francesco V. Sarcopenic obesity: a new category of obesity in the elderly. Nutr Metab Cardiovasc Dis. 2008;18(5):388–395. doi:10.1016/j.numecd.2007.10.002 [DOI] [PubMed] [Google Scholar]
- 8. Schaap LA, Koster A, Visser M. Adiposity, muscle mass, and muscle strength in relation to functional decline in older persons. Epidemiol Rev. 2013;35:51–65. doi:10.1093/epirev/mxs006 [DOI] [PubMed] [Google Scholar]
- 9. Batsis JA, Villareal DT. Sarcopenic obesity in older adults: aetiology, epidemiology and treatment strategies. Nat Rev Endocrinol. 2018;14(9):513–537. doi:10.1038/s41574-018-0062-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Goodpaster BH, Kelley DE, Thaete FL, He J, Ross R. Skeletal muscle attenuation determined by computed tomography is associated with skeletal muscle lipid content. J Appl Physiol (1985). 2000;89(1):104–110. doi:10.1152/jappl.2000.89.1.104 [DOI] [PubMed] [Google Scholar]
- 11. Larsen B, Bellettiere J, Allison M, et al. Muscle area and density and risk of all-cause mortality: the Multi-Ethnic Study of Atherosclerosis. Metabolism. 2020;111:154321. doi:10.1016/j.metabol.2020.154321 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Engelke K, Museyko O, Wang L, Laredo JD. Quantitative analysis of skeletal muscle by computed tomography imaging-State of the art. J Orthop Translat. 2018;15:91–103. doi:10.1016/j.jot.2018.10.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Correa-de-Araujo R, et al. Myosteatosis in the context of skeletal muscle function deficit: an interdisciplinary workshop at the National Institute on Aging. Front. Physiol. 2020;11:963. doi:10.3389/fphys.2020.00963 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Visser M, Goodpaster BH, Kritchevsky SB, et al. Muscle mass, muscle strength, and muscle fat infiltration as predictors of incident mobility limitations in well-functioning older persons. J Gerontol A Biol Sci Med Sci. 2005;60(3):324–333. doi:10.1093/gerona/60.3.324 [DOI] [PubMed] [Google Scholar]
- 15. Delmonico MJ, Harris TB, Visser M, et al. ; Health, Aging, and Body. . Longitudinal study of muscle strength, quality, and adipose tissue infiltration. Am J Clin Nutr. 2009;90(6):1579–1585. doi:10.3945/ajcn.2009.28047 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Cawthon PM, Blackwell T, Cummings SR, et al. Muscle mass assessed by the D3-creatine dilution method and incident self-reported disability and mortality in a prospective observational study of community-dwelling older men. J Gerontol A Biol Sci Med Sci. 2021;76(1):123–130. doi:10.1093/gerona/glaa111 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Duchowny KA, Peters KE, Cummings SR, et al. ; Osteoporotic Fractures in Men (MrOS) Study Research Group. . Association of change in muscle mass assessed by D3-creatine dilution with changes in grip strength and walking speed. J Cachexia Sarcopenia Muscle. 2020;11(1):55–61. doi:10.1002/jcsm.12494 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Evans WJ, Hellerstein M, Orwoll E, Cummings S, Cawthon PM. D3-creatine dilution and the importance of accuracy in the assessment of skeletal muscle mass. J Cachexia Sarcopenia Muscle. 2019;10(1):14–21. doi:10.1002/jcsm.12390 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Orwoll ES, Peters KE, Hellerstein M, Cummings SR, Evans WJ, Cawthon PM. The importance of muscle versus fat mass in sarcopenic obesity: a re-evaluation using D3-creatine muscle mass versus DXA lean mass measurements. J Gerontol A Biol Sci Med Sci. 2020;75(7):1362–1368. doi:10.1093/gerona/glaa064 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Orwoll E, Blank JB, Barrett-Connor E, et al. Design and baseline characteristics of the Osteoporotic Fractures in Men (MrOS) study—a large observational study of the determinants of fracture in older men. Contemp Clin Trials. 2005;26(5):569–585. doi:10.1016/j.cct.2005.05.006 [DOI] [PubMed] [Google Scholar]
- 21. Blank JB, Cawthon PM, Carrion-Petersen ML, et al. Overview of recruitment for the Osteoporotic Fractures in Men Study (MrOS). Contemp Clin Trials. 2005;26(5):557–568. doi:10.1016/j.cct.2005.05.005 [DOI] [PubMed] [Google Scholar]
- 22. Cawthon PM, Orwoll ES, Peters KE, et al. ; Osteoporotic Fractures in Men (MrOS) Study Research Group. . Strong relation between muscle mass determined by D3-creatine dilution, physical performance, and incidence of falls and mobility limitations in a prospective cohort of older men. J Gerontol A Biol Sci Med Sci. 2019;74(6):844–852. doi:10.1093/gerona/gly129 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Cauley JA, Burghardt AJ, Harrison SL, et al. Accelerated bone loss in older men: effects on bone microarchitecture and strength . J Bone Miner Res. 2018;33:1859–1869. doi:10.1002/jbmr.3468 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Erlandson MC, Wong AKO, Szabo E, et al. Muscle and myotendinous tissue properties at the distal tibia as assessed by high-resolution peripheral quantitative computed tomography. J Clin Densitom. 2017;20(2):226–232. doi:10.1016/j.jocd.2016.10.005 [DOI] [PubMed] [Google Scholar]
- 25. Burghardt AJ, Kazakia GJ, Laib A, Majumdar S. Quantitative assessment of bone tissue mineralization with polychromatic micro-computed tomography. Calcif Tissue Int. 2008;83(2):129–138. doi:10.1007/s00223-008-9158-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Shankaran M, Czerwieniec G, Fessler C, et al. Dilution of oral D3-creatine to measure creatine pool size and estimate skeletal muscle mass: development of a correction algorithm. J Cachexia Sarcopenia Muscle. 2018;9(3):540–546. doi:10.1002/jcsm.12278 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Clark RV, Walker AC, O’Connor-Semmes RL, et al. Total body skeletal muscle mass: estimation by creatine (methyl-d3) dilution in humans. J Appl Physiol (1985). 2014;116(12):1605–1613. doi:10.1152/japplphysiol.00045.2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Evans WJ, Scottoline B, Imam F, et al. D(3)-creatine dilution for the noninvasive measurement of skeletal muscle mass in premature infants. Pediatr Res. 2021;89:1508–1514. doi:10.1038/s41390-020-01122-w [DOI] [PubMed] [Google Scholar]
- 29. Lee CG, Boyko EJ, Nielson CM, et al. ; Osteoporotic Fractures in Men Study Group. . Mortality risk in older men associated with changes in weight, lean mass, and fat mass. J Am Geriatr Soc. 2011;59(2):233–240. doi:10.1111/j.1532-5415.2010.03245.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Cawthon PM, Fullman RL, Marshall L, et al. ; Osteoporotic Fractures in Men (MrOS) Research Group. . Physical performance and risk of hip fractures in older men. J Bone Miner Res. 2008;23(7):1037–1044. doi:10.1359/jbmr.080227 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Hill DK. The location of creatine phosphate in frog’s striated muscle. J Physiol. 1962;164:31–50. doi:10.1113/jphysiol.1962.sp007000 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Zanker J, Patel S, Blackwell T, et al. Walking speed and muscle mass estimated by the D(3)-creatine dilution method are important components of sarcopenia associated with incident mobility disability in older men: a classification and regression tree analysis. J Am Med Dir Assoc. 2020;21:1997–2002.e1991. doi:10.1016/j.jamda.2020.03.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Wang L, Yin L, Zhao Y, et al. Muscle density discriminates hip fracture better than computed tomography X-ray absorptiometry hip areal bone mineral density. J Cachexia Sarcopenia Muscle. 2020;11(6):1799–1812. doi:10.1002/jcsm.12616 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Cawthon PM, Fox KM, Gandra SR, et al. ; Health, Aging and Body Composition Study. . Clustering of strength, physical function, muscle, and adiposity characteristics and risk of disability in older adults. J Am Geriatr Soc. 2011;59(5):781–787. doi:10.1111/j.1532-5415.2011.03389.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Cawthon PM, Fox KM, Gandra SR, et al. ; Health, Aging and Body Composition Study. . Do muscle mass, muscle density, strength, and physical function similarly influence risk of hospitalization in older adults? J Am Geriatr Soc. 2009;57(8):1411–1419. doi:10.1111/j.1532-5415.2009.02366.x [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.