Introduction

Although obesity is believed to be a public health problem worldwide because of its negative effects on individuals’ health, its rate is seemingly increasing [1–3]. Not only has adverse impact of obesity on physical health been shown, but it is also associated with mental disorders such as increased depressive symptoms (DS) [4–11], a situation having substantial effects on public health [12].

Despite the strong effect of abdominal obesity on individuals’ physical health outcomes, there is little evidence underlying the relationship between central obesity and mental disorders such as depression [11]. It has been reported that body mass index (BMI) as an overall body obesity index was not [9, 13–16] or was even negatively associated with some forms of mental illnesses [9, 17]. Some other studies have shown that fat distribution (e.g., central and overall obesity) is differentially associated with DS [9, 18, 19]. On the other hand, it has been indicated that subcutaneous fat percentage, as an overall body obesity index, can predict obesity better than BMI and concluded that BMI as a measure of overall obesity does not account for varying proportions of muscle mass, bone, and fat, or the distribution of body fat [20, 21]. However, little evidence exists underlying the relationship between subcutaneous fat percentage and depression.

According to the literature, it seems that abdominal obesity is a stronger predictor of DS when compared to overall obesity [11]. However, since the relationship between central obesity and DS may vary according to country of origin, and there is scare evidence for developing countries [11], there is little evidence for children, adolescents, and youths; the majority of the studies did not adjust their analyses for some potential confounders, such as age, socioeconomic status, and physical activity (PA) [17, 18]; therefore, there is a need for further studies [11].

This study explores the relationship between DS with central and overall obesity in a sample of Iranian children, adolescents, and youths.

Methods

Participants and procedure

During 2015–2016, a round of cross-sectional data was collected from a sample of one thousand and fifty-two males ranging from 7 to 24 years old in the center of Ardabil Province, North West of Iran. To select children and adolescents (7–18 years), eight urban public schools were randomly selected. Then, classes were selected randomly in each school, and the children or adolescents in the classes were invited to participate in the study. 780 children and adolescents informed their consent verbally and underwent the measurements. However, since 45 participants did not deliver the signed written consent of their parents, they were excluded from the study. In addition, since 59 participants did not complete all the measurements, their data were excluded from the statistical analyses. To select youths (19-to-24-year-old male students), university students who were studying at the University of Mohaghegh Ardabili were recruited to participate in the study. To this end, from 26 physical education classes, 17 were randomly selected. The procedure of the study was explained to the students, and they were invited to participate in the study. 427 students informed their consent verbally and signed the written consent and underwent the measurements. However, since 51 students did not complete all the measurements, their data were excluded from the statistical analyses. The age of the children and adolescents was determined based on their date of birth in their school registering forms, but the university students wrote their age in the researcher made questionnaire. All measurements were carried out during regularly scheduled physical education lessons. The present study was approved by the Human Ethics Committee of the Ardabil department of Education and University of Mohaghegh Ardabili and the experiment was performed in accordance with the ethical standards of the committee and with the Helsinki Declaration. All the participants volunteered for this study and their informed consents were acquired according to the rules of the department in written signed form for the university students, adolescent (parents’ signed form), and children (parents’ signed form).

Anthropometric variables

Weight was measured in underwear and without shoes with an electronic scale (Type SECA 861) to the nearest 0.1 kg, and height was measured barefoot in the Frankfurt horizontal plane with a telescopic height measuring instrument (Type SECA 225) to the nearest 1 mm [22, 23]. BMI was calculated as body weight in kilograms divided by the square of height in meters.

Subcutaneous fat skinfold was measured as a general obesity index. Subcutaneous fat skinfold has been suggested as a more reliable overall obesity index than BMI [20, 21]. Lange skinfold caliper was used to assess triceps and calf skinfold thickness on the children and adolescents’ right side of the body and the average of three measures was calculated for each site [24, 25]. Then, body fat percentage was calculated according to Slaughter et al. [26]. For the university students, body fat percentage was determined from the average of three measures of the thickness of three sites on the right side of the body (chest, abdomen, and thigh) and calculated according to Jackson–Pollock method [27].

Waist circumference (WC), as a central body obesity index, was measured at a point immediately above the iliac crest on the midaxillary line at minimal respiration to the nearest 0.1 cm [28].

Depressive symptoms (DS)

Children’s Depression Inventory (CDI) [29], the most widely used measure of DS in children [4, 5], with good support for its reliability and validity [30, 31], was used to obtain the DS of the children (7–12 years old). Children were recommended to fill out the CDI questionnaire by their parents’ supervision. For the participants equal to or younger than 13 years old, the Beck Depression Inventory-II (BDI-II) [32] was employed as a valid measure of DS [33].

Possible covariates/confounders

The participants whose age was between 13 and 18 years were requested to classify themselves in one of the five stages of pubertal maturation status (PMS) defined by Tanner and Whitehouse [34]. Socioeconomic status (SES) was determined on the basis of parents’ education and occupation status by collecting data of monthly household income and the highest level of education [4, 5, 22, 24]. Physical activity (PA) for the children and adolescents was obtained using the PA Questionnaire-children (PAQ-C) [35] and PA Questionnaire-adolescents (PAQ-A) [36], respectively, with some alternations to suit to our society [4, 22, 24]. Both questionnaires have been shown valid for PA inventory [37, 38]. For the university students, the 12-month recall Baecke PA questionnaire was used [39] as a valid and reliable PA [PA at work, PA during leisure excluding sport (PAL) and sport during leisure time (PAS)] inventory [40, 41]. Since almost all the students had not jobs, the questions regarding PA at work were excluded.

Statistical analysis

The subjects were categorized into three groups as follows: children (7–12 years), adolescents (13–18 years), and youths (19–24 years). General characteristics (descriptive statistics) of the categorized groups are shown in Table 1. Before further analysis, all data were screened for problems of skewness, kurtosis, and outliers. Hierarchical linear regression analyses were conducted as follows: covariates (including age, SES, PMS, and PA) were included in the first step and obesity indices (WC and fat percentage) were included in the second step, separately. All calculations were performed using SPSS, version 21.0. The significance level was set at P < 0.05.

Table 1 General characteristics of the participants
Full size table

Results

Multiple linear regression analysis revealed significant relationship of DS with age (P < 0.01) and PA (P < 0.05) in the children (Table 2; step 1 and step 2). Positive but not significant relationship was observed between fat percentage and DS in the children (Table 2, step 2, standardized β = 0.085; P = 0.13). Introducing WC to the model had significant effect (F change = 5.4; P < 0.05) and showed significant relationship with DS in the children (Table 2, step 2, standardized β = 0.13; P < 0.05).

Table 2 Hierarchical regression analysis between DS with central and overall obesity indices after adjustment for the covariates in the children (7–12 years; n = 356)
Full size table

Significant negative relationship was observed between DS and PA in the adolescents (Table 3, step 1 and 2, P < 0.05). Positive but not significant relationship was observed between fat percentage and DS in the adolescents (Table 3, step 2, standardized β = 0.010; P = 0.10). Introducing WC had significant effect on the model (F change = 4.9; P < 0.05) and showed significant relationship with DS in the adolescents (Table 3, step 2, standardized β = 0.14; P < 0.05).

Table 3 Hierarchical regression analysis between DS with central and overall obesity indices after adjustment for the covariates in the adolescents (13–18 years; n = 320)
Full size table

Significant negative relationship was observed between PAS and DS in the youths (Table 4, step 1 and 2, P < 0.05). No significant relationship was observed between fat percentage and DS in youths (Table 4, step 2, P > 0.05). Positive but not significant relationship was observed between WC and DS in the youths (Table 4, step 2, standardized β = 0.09; P = 0.22).

Table 4 Hierarchical regression analysis between DS with central and overall obesity indices after adjustment for the covariates in the youths (19–24 years; n = 376)
Full size table

Discussion

Using a large sample of 7-to-24-year old male participants, it was observed that greater central obesity (but not overall obesity) was significantly associated with DS in the children and adolescents. Although positive relationship was observed between central obesity and DS in the youths, the relationship was not significant in this group. The possible reasons for this relationship in the youths are not clear but may be related to the students’ changed situation from school to university and/or their mood while answering the DS questionnaire.

The findings of this study are in line with the previous studies and suggest abdominal obesity as a stronger factor impacting mental health than overall obesity [11] and may be a preferred predictor of depression in human [42]. Furthermore, the findings are in line with other studies indicating that abdominal obesity, more than overall obesity, is related to poor health outcomes, such as cardiovascular diseases, diabetes, and mortality [6–8]. However, one large epidemiological study could not demonstrate an association between WC and depression [43]. Recently, a longitudinal study by including a large sample of older individuals observed that depressed individuals showed a significantly higher increase in abdominal obesity over five years compared with their non-depressed peers, and interestingly, such a relationship was not observed for an increase in general obesity and appeared to be independent of changes in general obesity. The authors concluded that DS was rather specifically related to fat gain in the abdominal region [44].

Some mechanisms have been suggested underlying the positive relationship between abdominal obesity and depression. For instance, it has been shown that chronic stress and depression are associated with a dysregulation of the hypothalamic–pituitary–adrenal axis and elevated concentrations of cortisol [45, 46], while visceral fat is highly sensitive to cortisol concentration [47] by activating lipoprotein lipase and inhibiting lipid mobilization [48]. Moreover, depression has been reported to be associated with high levels of inflammatory markers [49] that can activate the hypothalamic–pituitary–adrenal axis [50] and, in turn, resulting in abdominal obesity. On the other hand, it seems that depressed individuals have an unhealthier lifestyle [44] that may increase their abdominal obesity. For instance, even in individuals with non-hypercortisolemic atypical depression, because of overeating, a cycling of weight loss and gain occurring throughout recurrent episodes of depression can preferentially distribute weight to visceral fat areas [51].

Significant negative relationship was observed between PA and DS for all the participants, and interestingly, in the youths, significant relationship was observed between PAS (but not PAL) and DS. The result is consistent with the results of both cross-sectional and longitudinal studies [4, 5, 52]. In a recent review study, small but significant overall effect for PA on depression in children and adolescents has been demonstrated, and it was concluded that PA may play a role in the prevention and treatment of depression in young people [53].

PA has been suggested as a protective factor against depressed mood and several biological and psychological mechanisms by distracting negative thoughts and influencing self-esteem, and therefore, improving the retrieval of positive thoughts [54]. Neurobiological mechanisms have been reported as other possibilities. For instance, it has been suggested that neurotransmitters released during PA may mediate changes in mood and DS by positively impacting psychological well-being such as self-esteem [53, 55].

Conclusion

In summary, the results indicated that abdominal obesity (but not general obesity) could significantly predict DS in the children and adolescents (but not youths). Furthermore, it was observed that higher PA was significantly associated with lower DS in all the participants. Thus, in line with other studies, mental health status may be monitored and evaluated in these age groups using abdominal obesity [11, 42, 44] and may be protected by efficient PA prescription [53, 56].

However, this study is subject to several limitations. First, the causal relationship between WC and having higher DS may not be established based on the cross-sectional nature of the study. For instance, bidirectional relationship has been reported between obesity and DS, by the existence of some evidence of the association between low SES, the poor social relationship and chronic diseases with obesity which may have predisposed people with obesity to impaired mental health [14, 42]. Second, this study could not include subjects of both sexes. As a result, further research can possibly be done by the simultaneous inclusion of male and female participants.