Introduction

Metabolic syndrome is a clustering of risk factors for cardiovascular diseases (CVD) [1] and type 2 diabetes [2]. Essential components of metabolic syndrome are glucose intolerance, hypertension, dyslipidemia, and central obesity, although there are variations in the criteria defining the syndrome [3]. Depressive status has been linked to metabolic abnormalities, such as abdominal obesity, high blood pressure, and insulin resistance [4–6]. Moreover, several studies have examined depressive state in relation to metabolic syndrome. In two cross-sectional studies, metabolic syndrome was significantly associated with a higher prevalence of depression in young U.S. women [7] and in Japanese men [8]. Similarly, metabolic syndrome was found to be a significant predictor of depression in prospective studies of middle-aged British adults [9] and Japanese male employees [10], although no such association was observed in large cross-sectional studies of Finnish young adults [11] and Norwegian adults [12]. Among Japanese populations, only two studies (one cross-sectional study [8] and one cohort study [10] in the same setting) have reported epidemiologic data on this subject, but both included data for men only. In the study reported here, we examined the association between metabolic syndrome and its components and depressive symptoms in Japanese men and women.

Methods

Details of the study procedure have been described elsewhere [13]. In brief, this cross-sectional study was based on data collected during a health survey conducted in July and November 2006 among employees of two municipal offices in northeastern Kyushu, Japan. All employees were invited to participate in the survey (n = 601). Of these 601 employees, 547 (response rate 91%), aged 21–67 years, agreed to participate. Data on anthropometric measurements, blood parameters (from venous blood sample), and lifestyle (by questionnaire) were obtained. After the exclusion of those subjects with missing information, 458 Japanese subjects (285 men and 173 women) remained for analysis. The protocol of the study was approved by the Ethics Committee of the National Center for Global Health and Medicine, and written informed consent was obtained from each subject.

Metabolic syndrome and metabolic risk factors were defined according to a modified version of the criteria of the National Cholesterol Education Program’s Adults Treatment Panel III (NCEP-ATP III) [14]. Since the survey was conducted before the implementation of the medical examination for metabolic syndrome, waist circumference was not measured. We thus used body mass index (BMI) instead, which is highly correlated with waist circumference (r = 0.874, p = 0.0001) [15] and has been used in epidemiologic studies lacking waist circumference data.

Men having three or more of the following criteria were defined as having metabolic syndrome [16] and, due to the low number of women showing high triglyceride (n = 2), women having two or more of the same criteria were defined as having semi-metabolic syndrome: (1) obesity: BMI ≥ 25 kg/m2; (2) high triglycerides: triglycerides ≥150 mg/dl; (3) low high-density lipoprotein (HDL) cholesterol: <40 mg/dl in men, <50 mg/dl in women; (4) high blood pressure: systolic blood pressure ≥130 mmHg or diastolic blood pressure ≥85 mmHg; (5) high fasting glucose (overnight fasting): ≥100 mg/dl. Subjects under the care of a physician due to hyperlipidemia, hypertension, or diabetes were deemed as having the respective risk factors, regardless of the biochemical values and medication status. Depressive symptoms were assessed using a Japanese version [17] of the Center for Epidemiologic Studies Depression (CES_D) scale [18]. Depressive symptoms were defined as a CES_D score ≥16 and severe depressive state was defined as a CES_D score ≥23. The modified Brief Job Stress Questionnaire, in which a higher score indicates greater stress, was used to assess job stress score [19].

Separate analyses were performed on the data for men and women. The differences between the two groups for continuous variables were assessed by an independent t test and those for categorical variables were assessed by the chi-square test. Logistic regression analysis was used to assess the association between depressive symptoms and metabolic syndrome and each of its components.

We performed three types of analysis: (1) crude model without any adjustment, (2) multivariate model adjusted for marital status and job stress score in men and occupational physical activity and job stress score in women (Model 1), and (3) multivariate model adjusted for age, workplace, marital status, job stress score, occupational physical activity, leisure-time physical activity, current smoking, current alcohol drinking, and total energy intake (Model 2). We included these variables in the model based on their known or potential relations either to depressive symptoms or to obesity. Statistical significance was declared when p was <0.05. All statistical analyses were performed with SAS ver. 9.1 (SAS Institute, Cary, NC).

Results

The prevalence of depressive symptoms (CES_D score ≥16) and severe depressive symptoms (CES_D score ≥23) were 36.5 and 14.0% in men and 37.6 and 14.5% in women, respectively. Among the study subjects, 17.2% of the men had metabolic syndrome (≥3 components), whereas 11.0% of women had semi-metabolic syndrome (≥2 components). As shown in Table 1, compared to subjects without depressive symptoms, those of both sexes with depressive symptoms were more likely to have high job stress score, be unmarried (men only), and be engaged in occupational physical activity (women only), and compared to subjects without severe depressive symptoms of both sexes, those with severe depressive symptoms were more likely to have high job stress score, be unmarried (men only), not be engaged in leisure-time physical activity (men only), and be engaged in occupational physical activity (women only). There were no material differences between the two groups for BMI, triglycerides, HDL cholesterol, systolic blood pressure, diastolic blood pressure, and fasting glucose. A relatively strong correlation (r = >0.4) was observed between CES_D and the job stress score (r = 0.51, p < 0.0001 in men; r = 0.41, p = 0.0001 in women), between fasting glucose and age (r = 0.43, p = 0.0001 in men; r = 0.43, p = 0.0001 in women), and between diastolic blood pressure and age (r = 0.46, p = 0.0001 in men; r = 0.20, p = 0.007 in women).

Table 1 Characteristics of study subjects
Full size table

As Table 2 shows, no significant association was observed between depressive symptoms and metabolic syndrome and each of its components in both men and women. There were no statistically significant interactions for depressive symptoms between metabolic syndrome and other independent variables. For severe depressive symptoms (CES_D score ≥23; n = 40), a positive association with fasting hyperglycemia was observed in men; crude odds ratio (95% confidence interval) and adjusted odds ratio (95% confidence interval) for Models 1 and 2 were 1.39 (0.68–2.86), 2.05 (0.91–4.63), and 2.66 (1.05–6.72), respectively.

Table 2 Logistic regression analysis between metabolic syndrome and depressive symptoms in 285 men and 173 women aged 21–67 years
Full size table

Discussion

The results of this study among a Japanese working population did not reveal any relation between depressive symptoms and metabolic syndrome in both men and women. There was also no association between depressive symptoms and each component of metabolic syndrome, with the exception of a positive association between fasting hyperglycemia and severe depressive symptoms (CES_D score ≥23). The prevalence (36.5% in men, 37.6% in women) of depressive symptoms (CES_D score ≥16) among our study population is much higher than that reported in studies on Western populations but is similar to that reported in a previous large-scale study of Japanese workers (38.3% in men, 39.8% in women) [20].

Metabolic syndrome has been shown to be associated with depressive symptoms in studies performed in Western countries [7, 9, 21–26] and Japan [8, 10], although the association is not entirely consistent [11, 12]. This inconsistency could be ascribed to the difference in the criteria of depressive symptoms and metabolic syndrome across studies. Many earlier studies [7, 9, 11, 24, 25] used NCEP-ATP criteria, which we also adopted in our study, whereas others used the criteria of the International Diabetes Foundation (IDF) [8, 10, 12].

The CES_D scale has been developed for epidemiological studies of depression, and the reliability of its Japanese version has been confirmed [17]. The CES_D scale was used in our study, as well as in a number of previous studies [22, 25] to assess depressive symptoms; other studies have used different criteria, including the Profile of Mood States (POMS) [8, 10] and Hospital Anxiety and Depression Scale (HADS_D) [21, 27].

Although previous Japanese studies performed in the workplace setting [8, 10] reported a positive relationship between metabolic syndrome and depression, we did not detect such an association in our study. This inconsistency with earlier results needs to be discussed from a methodological point of view. These two earlier studies employed different diagnostic criteria for metabolic syndrome (IDF) from those used in our study (NCEP-ATP III). However, the final proportion of all participants with metabolic syndrome in one of the earlier studies (cross-sectional study by Takeuchi et al. [8]) does not differ greatly from that found in our study (12.2 vs. 17.2%). However, there is a large difference between these studies in the proportion of men with depression. The proportion of subjects with mild depression and depression were 6.9 and 0.7%, respectively, in Takeuchi et al.’s study [8], whereas the proportion of subjects with depression (defined as CES_D ≥ 16) and severe depression (defined as CES_D ≥ 23) was 36.5 and 14.0%, respectively, in our study. This difference is likely due to the different depression scales used, and these considerable differences in the prevalence of depression may account, at least in part, for the inconsistent findings among Japanese studies.

We found a positive association between severe depressive symptoms (CES_D score ≥23) and fasting hyperglycemia in men after adjusting for potential confounders. Prospective studies in Italy [28] and Spain [29] have shown that diabetes at baseline is associated with an increased risk of developing depressive symptoms. Moreover, in a meta-analysis involving 42 cross-sectional studies, subjects with diabetes had a twofold increased prevalence of depressive symptoms than subjects without diabetes [30]. Our finding is in line with these results. However, epidemiologic data are not entirely consistent. In a retrospective cohort study performed in Canada [31], diabetes was not associated with risk of depression, a finding which the authors ascribed to undiagnosed diabetes cases. In a U.S. prospective study [32], treated diabetes predicted an increased risk of depressive symptoms, but impaired fasting glucose and untreated diabetes were associated with a lower, rather than higher, risk of developing depressive symptoms. The inconsistency in the results among studies may thus be due to differences not only in the depression scales and diabetes criteria but also in other factors, including clinical stage of diabetes and medical treatment. The relationship between diabetes and depressive symptoms needs to be confirmed in future.

The strengths of this study include high participation rate, adjustment of potential confounding factors, and the use of a validated scale of depressive symptoms. Our study has also some limitations. First, because of its cross-sectional nature, any observed associations do not indicate a causal relationship. In a 7-year follow-up study [22], metabolic syndrome at baseline predicted increased risk of anxiety, suggesting that metabolic syndrome might increase the risk of depressive symptoms, rather than vice versa. Second, the number of subjects may not be sufficiently large to detect a modest association. Third, because we have no data for waist circumference, BMI was used.

In conclusion, the results of our study on Japanese employees does not support the hypothesis that metabolic syndrome is associated with depressive symptoms. An increased prevalence of a severe depressive state among men with fasting hyperglycemia requires confirmation.