The structure, strength, and movement provided by bones and muscles become more important with age. Sarcopenia and osteoporosis are two common diseases of aging that often attack together.¹ These diseases seriously affect the functional capacity and daily living standards of elderly people (sitting, standing up, climbing stairs, etc.) because they can increase the incidence of falls and fractures.^2-4 Many studies have revealed and confirmed the association between those conditions and adverse health outcomes.⁵

Bone mineral density (BMD) is an important indicator of bone strength and can be affected by many factors (age, sex, weight, height, etc.).^6-8 Several studies have examined the relationships among BMD, body fat, and lean mass,^9,10 and it was reported in 2010 and 2013 that lean mass is more closely associated with BMD than body fat in osteoporosis patients.^11,12 Because lean mass and body fat are what make up the body, the body mass index (BMI) is a major determinant of osteoporosis.^13-17 However, unlike its positive association with osteoporosis, BMI is a primary metabolic risk factor of sarcopenia. Therefore, some researchers have hypothesized that the ideal BMI, i.e., a high BMI irrespective of the body fat percent, could prevent bone loss in patients with osteosarcopenia.^18,19

After a review of the previous literature, it became clear that it is essential to synthesize prior cohort studies and thereby procure a more thorough recognition and understanding of the relationships among sarcopenia, BMD, and BMI.

In addition to providing instrumental recommendations that medical and nursing institutions can use to reduce the occurrence of osteoporosis in patients with sarcopenia by carrying out preventive nutritional interventions as early as possible, this study could also lay a research foundation and provide baseline data for future longitudinal research.

In this study, a thorough process for extracting data from relevant studies was conducted. Guidelines from the European Working Group for Sarcopenia in Older People (EWGSOP), the Foundation for the National Institutes of Health (FNIH) Sarcopenia Project, and the International Working Group (IWG) and appendicular skeletal muscle mass divided by height² were used as sarcopenia diagnostic criteria when assessing the quality of the included studies and the indications in the Preferred Reporting Items for Systematic Reviews and Meta-analyses statement.²⁰ A meta-analysis is not a primary research method, such as empirical research or experimental research, but it does involve procedures such as problem development, data collection (concerning studies), data coding, data analysis, and interpretation,²¹ which suggests the rationality and scientific validity of this research methodology.

Search strategy

The final search of all databases was June 1, 2021. The research queried four Korean electronic databases, Kmbase, KISS, NDSL, and RISS, and three international databases, PubMed, ScienceDirect, and Cochrane Library. The following controlled vocabulary terms and items were adopted for the searches of all databases: (“sarcopenia”[Title/Abstract] AND (“bone densities”[Title/Abstract] OR “density”, “bone”[Title/Abstract]) OR (“bone mineral density”[Title/Abstract]) OR (“bone mineral densities”[Title/Abstract]) OR (“density”, “bone mineral”[Title/Abstract]) OR (“bone mineral content”[Title/Abstract]) OR (“bone mineral contents”[Title/Abstract])). Hand-searching was adopted to check the reference lists of the returned studies and further filter those relevant studies by identifying additional research. All included studies were written in English.²² Initially, two researchers (YD and JN) separately reviewed the titles and abstracts of the studies returned from those searches.

Study selection

Studies that satisfied the following criteria were included in the meta-analysis: (1) compared BMD and BMI data between participants with sarcopenia and those without sarcopenia; (2) reported the prevalence of osteoporosis using criteria such as hip BMD, femoral neck density, lumbar spine density, and BMI. Studies that (1) did not measure or report BMD and BMI in both sarcopenia subjects and those without sarcopenia; (2) examined subjects younger than 65 years old; or (3) were not accessible or not in English, were excluded.

Data extraction

Two authors (YD and ST) independently scrutinized the retrieved articles and synthesized relevant data from the included studies into a standardized Microsoft Excel spreadsheet. Differences in opinion between data reviewers were settled through discussion with or judgment by the corresponding author. Eventually, the following information was extracted and sorted: (1) characteristics of each study sample (e.g., sample size, demographic characteristics); (2) the site where each study was conducted; (3) parameters associated with the BMD and BMI of individuals with and without sarcopenia; and (4) subgroup divisions between females and males.

Statistical analysis

After the extracted data were confirmed, the meta-analysis process of statistical synthesis was conducted using Comprehensive Meta-analysis V3.0 for Windows (http://www.meta-analysis.com/). It has been suggested that only outcomes with at least two studies can reasonably be meta-analyzed, so outcomes from only one study were reported in the descriptive analyses. Random-effects models using standardized mean differences with 95% confidence intervals (CIs) were run to account for heterogeneity among observational studies.^23,24 The chi-square-based Cochran Q statistic and I² statistic were calculated to quantify the heterogeneity of each summary estimate, with values of I²≥50% indicating moderate heterogeneity.^25,26 P-values lower than 0.05 were considered significant. Publication bias was assessed using a visual inspection of a funnel plot and the Egger bias test.²⁷

A flow chart of the search results is provided in Figure 1. The searches identified 40,822 potential eligible studies, of which 3,097 duplicates were excluded. Then, 3,050 papers were excluded by reviewing the titles and abstracts, and 47 full-text articles were examined. After further investigation, eight studies were finally used for this meta-analysis. Those eight studies originated in Australia,^16,28 Canada,²⁹ the United States,¹⁹ Brazil,³⁰ South Korea,¹⁸ Belgium,³¹ and Argentina.³²

Study and participant characteristics

The participant characteristics and criteria used to define sarcopenia in the included studies are summarized in Table 1. A total of 18,783 participants were included in the meta-analysis, and their average age ranged from 60 to 73.7 years.

Meta-analysis

After synthesizing the selected studies, it was clear that the sarcopenia patients had lower BMDs (total hip BMD, femoral neck BMD, and lumbar spine BMD) than the control group subjects. Sarcopenia patients also had lower a BMI than the control group subjects (Table 2). The forest plots for those effect sizes are shown in Figure 2. The overall effect sizes in the random-effects analysis were: total hip BMD (d=0.560; 95% CI, 0.438 to 0.681; P<0.01), femoral neck BMD (d=0.522; 95% CI, 0.423 to 0.621; P<0.01), lumbar spine BMD (d=0.295; 95% CI, 0.111 to 0.478; P<0.01), and BMI (d=0.711; 95% CI, 0.456 to 0.996; P<0.01). Therefore, sarcopenia patients had significantly lower total hip BMD, femoral neck BMD, lumbar spine BMD, and BMI than the control subjects. The finding that sarcopenia patients had a lower BMI contradicts the results of a previous study.³³ The included studies had a large degree of heterogeneity, with I² ranging from 53.755% to 97.609%. A subgroup analysis by sex was also performed, and the results showed no sex differences (P>0.05) in the effect of total hip BMD, femoral neck BMD, lumbar spine BMD, or BMI (Table 3).

Publication bias

In the funnel chart, no publication deviation appeared in the case of left-right symmetry, but publication deviation was seen in the case of asymmetry.³⁴ As shown in Figure 3, the 13 funnel graphs were each asymmetrical at the bottom, which indicates that publication deviations might exist. The method suggested by Hansen et al.³⁴, which uses an Egger linear regression to detect publication bias, was also used. The results of the Egger linear regression analysis were as follows: T_{total hip BMD}=0.873, df_{total hip BMD}=10, P>0.05; T_{femoral neck BMD}=1.979, df_{femoral neck BMD}=17, P>0.05; T_{lumbar spine BMD}=0.852, df_{lumbar spine BMD}=17, P>0.05; T_BMI=0.42, df_BMI=11, P>0.05. That is, the possibility of publication bias in the included studies cannot be ruled out, but there is no evidence to oppugn the validity of these results.

The percentage of older people in the overall population is increasing worldwide. Sarcopenia and osteoporosis are predictors of premature death in elderly people, so attention to those conditions is important.³⁵ The general consensus on sarcopenia and osteoporosis is that the skeletal and muscular systems are interconnected, and the conditions thus share the same risk factors. For example, hormone levels (growth hormone, sex hormone, vitamin D, etc.) decrease with age, and decreased exercise, poor nutritional status, and an increase in comorbidities are expected to cause both muscle strength and bone density to decrease significantly.^36-40 Therefore, this study investigated the correlation between muscle content and BMD in elderly people to provide a theoretical basis for the simultaneous treatment of sarcopenia and osteoporosis.

Using criteria established by prior literature and the EWGSOP, patients with sarcopenia (18,783 subjects) were analyzed. The results show that muscle mass correlated positively with the BMD of the total hip, femoral neck, and lumbar spine, and the effect size of muscle mass was larger than that of BMI, which is consistent with the conclusions of earlier studies in several countries. Baumgartner et al.⁴¹ verified that muscle mass correlated significantly and positively with the BMD of the whole body, total hip, femoral neck, and lumbar spine. Doyle et al.⁴² studied the correlation between bone mass and muscle mass and found that less muscle mass correlated with lower bone mass, and more muscle mass correlated with higher bone mass. Another study found that weight and bone loss occurred simultaneously, with a small amount of muscle loss, and that the risk of falls and fractures was twice as high in people who lost weight as it was in those who kept their weight steady or gained weight.⁴³ Verschueren et al.⁴⁴ conducted a study of 979 middle-aged and elderly European males with sarcopenia (mean age, 59.6 years). Their results suggest that skeletal muscle mass, body fat, and muscle strength in the limbs are linearly correlated with the BMD of different body parts. Furthermore, they found that body fat content correlated positively with the BMD of the total hip, femoral neck, and lumbar spine, but not with the BMD of the whole body.⁴⁴ Go et al.¹¹ observed an increased incidence of sarcopenia in men older than 50 years who had osteoporosis or osteopenia. Combining the results of the multiple regression analyses in this study with the results of a previous study,³⁶ it is evident that sarcopenia patients suffer from poor nutritional status and have a lower than average BMI, which can affect muscle mass, but patients with osteoporosis need a certain amount of body fat to reduce the risk of falls and fractures.^11,45,46 This result is contrary to that of Zheng et al.⁴⁷ The subgroup analysis in this study found no significant sex differences in the total hip BMD, femoral neck BMD, lumbar spine BMD, or BMI of elderly people with sarcopenia, which is also different from previous research results.⁴⁸ Therefore, sex differences need not be considered when planning strategies to prevent the transition from sarcopenia to osteoporosis in elderly people.

One physiological factor that could explain the relationships among muscle, fat, and bone is homeostatic regulation among them.^49,50 Evidence indicates that sarcopenia and osteoporosis have similar clinical manifestations and are caused by the same factors, such as inflammation, hormonal and nutrient deficiencies, and physical inactivity, that increase the risk of musculoskeletal injury.⁵¹ Furthermore, studies have indicated that increased muscle mass could lead to the elongation of collagen fibers and periosteum at the interface and thus stimulate bone growth, which could increase blood flow to the bone and thereby strengthen of the bone. Furthermore, an increase in blood flow to the extremities is proportional to an increase in muscle mass. Therefore, a decrease anywhere in that process could lead to osteoporosis.⁵² However, some recent studies have suggested that fat is not a protective factor against osteoporosis and fractures in older adults. Conversely, adipose tissue in the body could cause oxidative stress and the synthesis of pro-inflammatory adipocytokines that can have adverse effects on bone metabolism.^53-56 It is important to note that the effect of a high BMI on BMD has not been elucidated and might vary depending on the type and distribution of fat (subcutaneous and visceral). Therefore, because sarcopenia and osteoporosis are major causes of disability, hospitalization, and high health care costs among elderly people, it is recommended that more detailed studies be conducted to elucidate the effects of fat.

The findings of this study should be considered in light of its limitations. First, the definition and measurement of sarcopenia varied in the studies analyzed here, which demonstrates that the concept of sarcopenia is still controversial. To some extent, the research results might have been affected by the different definitions used. This limitation is faced by all research about sarcopenia because sample selection is ambiguous. So far, published and officially certified operational definitions for sarcopenia have been published by Baumgartner;⁴¹ Newman; IWG;⁵⁷ EWGSOP;⁵⁷ the European Society for Clinical Nutrition and Metabolism Special Interest Group on cachexia-anorexia in chronic wasting disease;⁵⁸ the Society of Sarcopenia, Cachexia, and Wasting Disorders;⁵⁹ and the FNIH Sarcopenia Project.^60,61 EWGSOP has defined sarcopenia in men as appendicular lean mass adjusted for height squared <7.25 kg/m² combined with low hand-grip strength (<30 kg) or low gait speed (<0.8 m/sec).³ To date, no single outcome serves as a gold standard to clinically diagnose sarcopenia.³⁶ Second, the included studies did not subdivide sarcopenia patients into groups, such as pre-sarcopenia and sarcopenia, although the clinical stages of this disease differ in meaningful ways.⁶² Pre-sarcopenia and sarcopenia, which are characterized by a loss of muscle mass and muscle strength, respectively, are two major conditions responsible for the reduced functional capacity of individuals, and they increase the probability that a patient will be diagnosed with osteoporosis.^63,64 Due to the limited number of studies included here, this research was not further refined in the subgroup analysis.

In summary, sarcopenia and osteoporosis, which primarily affect the elderly population, pose a significant clinical and economic burden for diagnosis, prevention, and treatment. The medical need of these two diseases has not yet been satisfied. It was postulated that genetics, development, endocrine changes, and lifestyle could affect muscle strength, bone mass, and function. Specifically, sarcopenia is directly related to the BMD of the total hip, femoral neck, and lumbar spine and closely related to BMI. Appropriate obesity might be a protective factor against osteosarcopenia in people aged 60 years and older. These results encourage further research to investigate the effects and significance of appropriate exercise training programs, specific dietary patterns, and associated clinical interventions to increase muscle mass in elderly adults to prevent and treat BMD loss, thereby reducing the risk of osteosarcopenia.

The authors declare no conflict of interest.

Study concept and design: YD, ST, and CO; acquisition of data: YD, ST, and CO; analysis and interpretation of data: YD, ST, CO, and JN; drafting of the manuscript: YD and ST; critical revision of the manuscript: YD, ST, CO, and JN; statistical analysis: YD, ST, CO, and JN; and study supervision: YD, ST, and CO.