The Association between Sarcopenic Obesity and Depressive Symptoms in Older Japanese Adults

The effects of sarcopenic obesity, the co-existence of sarcopenia and obesity, on mood disorders have not been studies extensively. Our objective was to examine the association of depressive symptoms with sarcopenia and obesity status in older Japanese adults. We analyzed data from 1731 functionally-independent, community-dwelling Japanese adults aged 65 years or older (875 men, 856 women) randomly selected from the resident register of Kashiwa city, Chiba, Japan in 2012. Sarcopenia was defined based on appendicular skeletal muscle mass, grip strength and usual gait speed. Obesity was defined as the highest sex-specific quintile of the percentage body fat. Depressive symptoms were defined as a Geriatric Depression Scale 15-item score ≥ 6. Multiple logistic regression was employed to examine the association of depressive symptoms with four groups defined by the presence/absence of sarcopenia and obesity. The prevalence of depressive symptoms was 10.1% and the proportions of sarcopenia/obesity, sarcopenia/non-obesity, non-sarcopenia/obesity, non-sarcopenia/non-obesity were 3.7%, 13.6%, 16.9% and 65.8%, respectively. After adjustment for potential confounders, sarcopenia/obesity was positively associated with depressive symptoms compared with non-sarcopenia/non-obesity, whereas either sarcopenia or obesity alone was not associated with depressive symptoms. The association was particularly pronounced in those aged 65 to 74 years in age-stratified analysis. We conclude that our findings suggest a synergistic impact exerted by sarcopenic obesity on the risk of depressive symptoms, particularly in those aged 65 to 74 years.


Introduction
Obesity and depression are considered major threats to public health worldwide. Obesity has more than doubled since 1980 and is considered a major risk factor for various diseases including cardiovascular diseases, diabetes, osteoarthritis and cancers [1]. Depression is prevalent and a major risk factor for suicide, and is associated with considerable morbidity and mortality [2].
The association between these two major medical conditions has attracted much research interest. However, epidemiological studies to date have generated conflicting results. Some studies showed a positive association between depression and obesity [3], but others showed an inverse association [4] or U-shaped association [5]. Some studies even failed to show a significant association [6]. This inconsistency may have resulted from differences in study design, subject population, depression measurement, statistical analysis or covariates used. Another possible explanation could be related to the limited discrimination ability of body mass index (BMI) (i.e., ability to correctly classify obesity), which was used in most studies to ascertain obesity.
Obesity is abnormal or excessive fat accumulation that may impair health [1]. While BMI is the most useful measure of obesity at the population level, it does not differentiate muscle mass from fat in individuals. Individuals with low BMI may still have as much fat as those with high BMI, and higher BMI may mean greater muscle mass in some individuals. Use of BMI, an imperfect measure of adiposity, is actually considered as one of the explanations for the obesity paradox, the unexpected phenomenon that overweight or mild obesity is associated with decreased mortality compared with normal weight [7,8]. Indeed, recent large-scale prospective cohort study demonstrated that both low BMI and high body fat percentage are independently associated with increased mortality, claiming the importance of using direct measures of adiposity [9].
Use of BMI is particularly problematic when used to identify obesity in the elderly. People accumulate fat while losing lean body mass as they age. Hence, the discrimination ability of a certain BMI cut-off point with respect to body adiposity decreases with age [10]. In addition, body composition is highly heterogeneous in the elderly [11], which may weaken the association between BMI and body adiposity.
The loss of muscle mass and physical function with age needs to be accounted for as well in studying factors associated with depression in the elderly [6]. Lower muscle mass appears to be associated with increased depressive symptoms [12]. Several prospective cohort studies demonstrated a role of decreased physical performance or physical activity in the development of depression in the elderly. Lower physical performance measured as gait speed [13,14] or grip strength [15] was associated with the development of depression. Therefore, it is reasonable to hypothesize that sarcopenia, a syndrome characterized by progressive, generalized loss of skeletal muscle mass and strength with a risk of adverse outcomes [16], may be associated with depression.
In the present study, we employed percentage body fat to determine obesity, and examined the association of depressive symptoms with sarcopenia and obesity status in communitydwelling Japanese elderly. We hypothesized that both obesity and sarcopenia are associated with depressive symptoms, and that sarcopenic obesity, the co-existence of sarcopenia and obesity, is associated with higher risk of depressive symptoms compared with either alone, considering that sarcopenic obesity may carry cumulative risk derived from each of these two individual body composition phenotypes.

Participants
We analyzed data from the participants of the Kashiwa study, which is designed to characterize the biological, psychosocial and functional changes associated with aging in community-dwelling older adults. The sampling procedure was described in detail elsewhere [17]. Briefly, the inclusion criteria for study entry were age 65 years or older and functional independence (i.e., not requiring nursing care provided by long-term care insurance). The participants were Sleep quality was evaluated using the Pittsburgh Sleep Quality Index, of which Japanese version has been validated [23]. The total score ranges from 0 to 21, with a lower score indicating better sleep, and participants with a score higher than 5 are considered poor sleepers.
Social support provided by family/friends was assessed using the 6-item Lubben Social Network Scale. This scale asks about the number of family members/friends with whom one has social contact and social support, and is widely used as a validated screening tool for social isolation among community dwelling older adults [24]. The scale ranges from 0 to 30, with a higher score indicating a strong social support network, and participants with a score less than 12 are considered socially isolated.
Social connections beyond the social support provided by individual relationships with family and friends were assessed using the Social Cohesion Scale [25]. The Social Cohesion Scale asks five questions regarding the strength of neighborhood ties, such as whether people in the neighborhood can be trusted and whether they get along with each other. Each question was answered with a 5-point Likert scale, which was added to generate a total score ranging from 5 to 25, with a lower score representing greater social cohesion.

Statistical Analysis
Differences in participant characteristics among the four groups defined by sarcopenia and obesity status were examined using ANOVA test (or Kruskal-Wallis test) for continuous variables and chi-square test for categorical variables. The p-values were adjusted for multiple testing using the Hochberg procedure. [26] To evaluate the association with depressive symptoms, odds ratios (ORs) with 95% confidence intervals (CIs) for depressive symptoms were obtained by conducting multiple logistic regression analysis while controlling for covariates.
Our preliminary analysis indicated that there was no evidence of effect modification by sex on the covariates-adjusted associations of depressive symptoms with sarcopenia and obesity status (Wald p = 0.72-0.78). Therefore, data for men and women were pooled and analyzed together. As the next step of the preliminary analysis, we tested the applicability of BMI as a measure of obesity in the context of researching factors associated with depressive symptoms. We conducted covariates-adjusted multiple logistic regression with BMI as an independent variable, in the form of a linear term, a quadratic term or a binary variable with a cutoff of 25, with or without sarcopenia, in addition to interaction terms with sarcopenia when sarcopenia was in the model, but BMI was not significantly associated with depressive symptoms in any model (data not shown).
In the main analysis, we conducted multiple logistic regression analysis with obesity ascertained using percentage body fat. Covariates were selected a priori based on their association with depression, sarcopenia or obesity and added to the model in a sequential manner. The model was initially adjusted for age and sex (model 1). We first added behavior variables including food intake (five-level categorical), physical activity and sleep quality (poor sleeper vs. good sleeper) (model 2). We further adjusted for social factors including living condition (living alone or not), education level (three-level categorical, "below high school", "high school graduate", "college or higher"), social isolation and Social Cohesion Scale (model 3). In the final model, medical variables including use of antidepressants, use of statin and chronic comorbidity burden were added to the model (model 4).
Several recent studies have suggested that age modifies the effects of cardiovascular risk factors [7,[27][28][29]. Cardiovascular risk factors do not seem to exert adverse effects on the very old or frail individuals as they do on younger individuals. Analysis with an interaction term between age and sarcopenia and obesity status verified the presence of effect modification by age (Wald p < 0.001 in age-and sex-adjusted model, p = 0.02 in the fully-adjusted model). We then further divided participants into young-old (65-74 years) and old-old (! 75 years) as per previous studies [29] and conducted age-stratified analysis.
We defined depressive symptoms as GDS score of 6 or higher, but this definition may misclassify persons as being depressed when they are not, as a result of temporary stress or symptoms due to physical conditions. Therefore, we repeated all analyses with severe depressive symptoms as a dependent variable to test whether adoption of a higher cutoff point for GDS may alter the association of depressive symptoms with sarcopenia and obesity status.
Finally, because the effects of comorbid conditions on depressive symptoms may vary, we chose four medical conditions that are strongly associated with depressive symptoms, namely diabetes mellitus, stroke, heart disease, and cancer, and added the presence or absence of these medical conditions separately as a covariate.
All analyses were conducted using SAS version 9.3 (SAS Institute Inc., Cary, NC) and R statistical software version 2.15.2 (R Foundation, Vienna, Austria). Two-sided p < 0.05 was considered statistically significant.

Participant characteristics
The prevalence of depressive symptoms was 10.1% (9.6% in men and 10.5% in women). The proportions of sarcopenia/obesity, sarcopenia/non-obesity, non-sarcopenia/obesity, and nonsarcopenia/non-obesity were 3.7%, 13.6%, 16.9% and 65.8%, respectively. The characteristics of the study participants by sarcopenia and obesity status are shown in Table 1. Those with sarcopenia/obesity tended to be older, less physically active, less educated, and had greater chronic comorbidity burden and poor sleep compared to those without sarcopenia or obesity.

Association of depressive symptoms with sarcopenia and obesity status
The prevalence of depressive symptoms in each of the four groups is shown in Fig 1. The prevalence of depressive symptoms was 26.6% in the sarcopenia/obesity group, which was the highest among the four groups.
In multiple logistic regression adjusted for age and sex, only the sarcopenia/obesity group had increased risk of depressive symptoms compared with the non-sarcopenia/non-obesity group, while neither the non-sarcopenia/obesity group nor the sarcopenia/non-obesity group had increased risk of depressive symptoms (Table 2, model 1). The significant association between sarcopenia/obesity and depressive symptoms (and the lack of significant association of depressive symptoms with sarcopenia/non-obesity or non-sarcopenia/obesity) persisted after successively adjusting for behavior variables (  4). In covariate-adjusted multiple logistic regression analysis with sarcopenia and obesity as two separate independent variables, the interaction term between sarcopenia and obesity was statistically significant (p = 0.04). In the age-stratified analysis, the presence of sarcopenia or obesity, or their combination was not associated with increased risk of depressive symptoms in participants aged 75 years or older (Table 2, model 4a). However, in participants aged 65 to 74 years, the sarcopenia/obesity group only had increased risk of depressive symptoms (Table 2, model 4b). Addition of BMI to these models did not attenuate the association between sarcopenia/obesity and depressive symptoms (data not shown).

Sensitivity analyses
We repeated analyses with severe depressive symptoms as a dependent variable. The prevalence of severe depressive symptoms in each of the four groups is shown in Fig 2, and is much lower than that of depressive symptoms. Similarly, to the main analysis, only the sarcopenia/obesity group had increased risk of depressive symptoms compared with the non-sarcopenia/non-obesity group, while neither the non-sarcopenia/obesity group nor the sarcopenia/non-obesity group had increased risk of depressive symptoms in all the models, though the small number of participants with severe depressive symptoms resulted in a wider 95% CI (Table 3, model 1-4). The significant association between severe depressive symptoms and sarcopenia/obesity was also observed only in the age group between 65 and 74 years.
Additional analyses introducing four medical conditions (diabetes mellitus, stroke, heart disease, and cancer) as covariates did not attenuate the observed association between depressive symptoms and sarcopenia/obesity (data not shown).
Lastly, we conducted analysis excluding 67 participants with Mini-Mental State Exam score of 24 or lower (3.9%), which did not alter out conclusion (data not shown).

Discussion
In this cross-sectional analysis of data from 1732 community-dwelling Japanese older men and women, we demonstrated that participants with sarcopenic obesity, the co-existence of sarcopenia and obesity, had significantly increased risk of depressive symptoms compared to those without either sarcopenia or obesity. This relationship remained statistically significant after successive addition of covariates to the model, albeit with slightly diminished effect size. Our findings suggest a synergistic impact on the risk of depressive symptoms exerted by sarcopenia and obesity in the elderly, since participants with either obesity or sarcopenia did not have increased risk of depressive symptoms and the interaction term between sarcopenia and obesity was statistically significant.  The mechanisms underlying the association between depressive symptoms and sarcopenic obesity remain elusive and largely unknown. However, we speculate that there are several biological mechanisms that could potentially mediate the depression-sarcopenic obesity association. Obesity may cause a chronic inflammatory state [30] and insulin resistance [31], which, in turn, are associated with both sarcopenia [32,33] and depression [34,35]. Elevated leptin has been observed in those with sarcopenia [36]. and was associated with increased risk of depression, especially in the presence of obesity [37]. Therefore, leptin resistance might be implicated  as a mediating factor for the association between sarcopenic obesity and depression. There are other clinical factors that need to be considered. Obesity may lead to depression through its negative effects on self-image, disordered eating, and its associated stigma [38]. Several previous studies have shown the association of depression with physical performance [39] such as gait speed [13,14] or hand grip strength [15], consistent with our findings. With regard to the association between obesity and depression, previous findings are inconsistent (3)(4)(5)(6), with some studies demonstrating positive association, compatible with our findings (3). Most previous studies used BMI to ascertain obesity and our findings should be corroborated in future studies using percentage body fat to evaluate body adiposity.
We demonstrated that the association between sarcopenic obesity and depressive symptoms varies depending on age. The heightened risk of depressive symptoms associated with sarcopenic obesity was mostly observed in participants aged between 65 and 74 years, and it was not statistically significant in participants aged 75 years or older. Several recent studies have shown that the effects of cardiovascular risk factors may have different impacts in the elderly or frail population when compared to younger people. Obesity, or higher BMI, was inversely associated with mortality in elderly hospitalized patients [7]. Elevated blood pressure was associated with lower mortality risk in physically frail elderly adults who could not walk 20 feet [27]. Metabolic syndrome was associated with a lower probability of prevalent and incident functional disability in older adults [28]. The association between metabolic syndrome and cardiovascular events was observed only in patients younger than 75, but not in patients aged 75 or over [29]. Our study findings are consistent with these studies in that the adverse effects of obesity, one of the cardiovascular risk factors, is limited to relatively younger participants, but extend these studies by showing that the age difference in the adverse effects of obesity is also observed in a mental aspect. The reasons for the lack of impact of sarcopenic obesity on depressive symptoms among participants aged 75 years or older were unclear and need to be investigated in future studies. Those with depressive symptoms, particularly at extreme age, would be less likely to participate in epidemiological studies, and this possibility of healthy participant bias should be explored.
Our study has several limitations. First, we employed GDS to ascertain depressive symptoms. Although GDS is a valid measure of depressive symptoms and is widely used in clinical practice, it was originally developed as a screening tool rather than a diagnostic tool. It is possible that the depressive symptoms identified with GDS could be related to physical illness and not depression. However, the GDS has been successfully used for older adults in various settings and shown to have satisfactory psychometric properties [40]. In addition, the relationships between depressive symptoms and sarcopenic obesity were replicated when severe depressive symptoms, defined by a higher cutoff point of the GDS, was employed as an outcome. The higher cutoff point indicates more severe depression symptomatology and may increase the specificity (i.e., those identified as having severe depressive symptoms were most likely to be truly depressed). The successful replication with a more specific definition of depressive symptoms may support the validity of our findings. Second, our study was a crosssectional study, and the observed associations could be confounded by unmeasured or uncontrolled variables. In addition, data in the present study did not provide any explanation for the underlying mechanisms of the observed associations or the direction of the causality. Future studies with a prospective longitudinal cohort design will be needed to establish a causal relationship between depressive symptoms and sarcopenic obesity. Finally, the participants were exclusively functionally-independent, community dwelling Japanese older adults. The generalizability of our study findings will need to be tested using data from participants with a greater comorbidity burden or from other racial/ethnic groups.
In conclusion, we demonstrated that depressive symptoms was positively associated with sarcopenic obesity in functionally independent Japanese older adults. Sarcopenia and obesity appear to exert a synergistic impact on the risk of depressive symptoms, since either sarcopenia or obesity alone was not associated with depressive symptoms. Future research will need to be conducted to establish the causal pathways and identify mediators of the association, with particular attention to modifiable factors, so that co-occurrence of these two disorders, depression and sarcopenic obesity, could be prevented.