Introduction

According to the recent statistics, the overweight frequency among adults in Croatia is approximately 38 %, and an additional 20 % of men and women are obese [1]. It is well documented that excess weight increases the risk for many diseases and metabolic abnormalities, such as type 2 diabetes, coronary heart disease, certain malignancies, hypertension and dyslipidaemia [2]. In severely obese people, an array of psychological problems can range from negative body image and lowered self-esteem to eating disorders and clinical depression [3]. Obesity has negative effects on the activities of daily living [4] and subjective health status, especially its physical aspect [5].

The results of studies on the general population samples do not clearly support the assumption that obese people, as a group, have more psychological problems than non-obese people do. Although several studies have shown that BMI is significantly associated with mood, anxiety, and personality disorders [6, 7], other studies have suggested that there is no significant association between BMI and scores on standard psychological tests [8] or that obese people have better mental health than non-obese people [9]. In Western societies where obesity is stigmatised, being obese increases the risk of anxiety and depression and the possibility to understand the nature of the association between these two conditions is important for prevention and treatment [6, 10].

Results from prospective studies have shown that obesity is associated with future incidence of depression, and cross-sectional studies suggested significant positive association between obesity and depression, particularly in females [7]. Several reviews have found that depression is more common among persons with extreme obesity; when the BMI is higher than 40 kg/m2, the likelihood of experiencing a major depression episode may increase by five times [9]. Additionally, several studies have demonstrated that obese women experience more psychological problems than men [7, 11, 12].

One of the reasons obese persons may have a higher risk for depression and other affective disturbances might be their impaired health-related quality of life (HRQoL) [13]. Empirical studies have consistently indicated that increases in weight are associated with deterioration in HRQoL, particularly physical health. In addition to the fact that excessive BMI primarily affects an individual’s physical well-being, numerous studies have shown that different domains of physical HRQoL are differently affected by increased BMI. Yancy et al. [14], for example, showed that men with BMI of ≥25 kg/m2, compared with individuals of normal weight, had significantly lower scores on the Bodily Pain Subscale. Those with BMI ≥35 kg/m2 had lower scores on the Physical Functioning Subscale, and those with BMI ≥40 kg/m2 had lower scores on the Role Limitations due to Physical Problems. Jagielski et al. [15] have found that with increasing BMI, there were more problems in physical functioning, mobility, self-care and performing usual activities, reduced ability to work and overall quality of life.

Several studies have indicated gender differences between relative weight and physical health. Laaksonen et al. [16] have demonstrated that lower levels of BMI in women may be associated with impaired physical well-being in contrast to men. They have shown that in women, physical health gradually deteriorated with increasing BMI, even among those in the normal weight range. In men, physical health deteriorated only among those who had BMI levels exceeding 27 kg/m2, and poor physical health was associated only with obesity.

Wadden and Sarwer [13] have suggested that because obesity is a chronic state, decreases in the quality of life are likely to be long-term impairments and therefore could render obese persons more vulnerable to affective disturbances.

Depressive and anxiety symptoms are not necessarily entirely harmful and may have several positive effects. Herpertz et al. [17] have demonstrated that depressive and anxiety symptoms, as indicators of psychological stress with regard to obesity, appeared to be positive predictors of weight loss after obesity surgery. These authors stressed that the presence or the absence of emotional problems might not be predictive of weight loss; more importantly, the severity of emotional problems might be predictive of weight loss. Severe depression and anxiety may have many negative effects and interfere with a person’s ability to function in everyday life. For instance, because of the symptoms of depression, such as apathy, sadness, sleeping problems and lack of energy, the affected persons may have difficulties fulfilling their responsibilities and successfully performing their work. Additionally, they may have difficulties regarding social interactions; e.g. they may avoid others and withdraw from the community, commonly feel lonely, and often report the overuse of alcohol as “self-medication” [18]. Similar to obesity, depression significantly increases health care costs [19]. Therefore, in addition to monitoring the physical health of obese individuals, it is important to monitor the patients’ emotional status.

The present study aims to exam the differences in physical health functioning among overweight and obese people given their level of nutrition. It also aims to better understand the multidimensional correlates of HRQoL in the adult overweight and obese population. Specifically, we examined the association between body mass index, depression and anxiety as well as the potential mediating effects of physical health functioning on this association. Because of the expected gender differences and influence of age to negative affect and physical health, the analyses were separately done for men and women and the effects of age were controlled.

Method

Participants

The research was conducted on a sample of overweight and obese adults who visited their primary care physician during the 6-month study period. Overall, 273 persons (n = 143 women and n = 130 men) were enrolled in the study. The inclusion criterion of the study was BMI major than 25. The body mass index (BMI) of participants ranged from 25 kg/m2 to 49 kg/m2 (M = 31.17, SD = 5.07), and age ranged between 21 and 60 years (M = 46.68, SD = 10.62). Of the subjects, 43 % were overweight and 57 % were obese; 2.5 % completed elementary school, 9.6 % vocational and 57.9 % high school, and 30 % were undergraduate or graduate; 72.1 %, were married, 4.3 % divorced or widowed, and 23.6 % single.

Measures

Depression and anxiety were assessed using the Hospital Anxiety and Depression Scale (HADS) [20], one of the most widely used instruments for detecting depression and anxiety, with well-established validity and reliability. The HADS is a 14-item self-report questionnaire: 7 items assess anxiety (HADS-A) and 7 items assess depression (HADS-D) over the last 2 weeks. The items focus on the psychological and cognitive symptoms of these two disorders and they have a 4-point Likert-type format (values 0 through 3). In each subscale, the responses result in a total score ranging from 0 to 21, with higher scores indicating higher levels of distress. A score between 8 and 10 is considered as “a possible case”, and a score of 11 or more is considered as a case of depression or anxiety. In this sample, Cronbach’s alpha coefficient for the HADS-A was 0.83, and Cronbach’s alpha coefficient for the HADS-D was .70.

Physical health functioning was measured using the Medical Outcome Study Short-Form 36 (SF-36) [21]. It is a well-validated, widely used generic health-related quality-of-life measure, and as such, it allows measurements and comparisons of QoL across various diseases. This 36-item self-report questionnaire assesses health status within eight domains of functioning: four domains reflect physical health (physical functioning, role limitations due to physical problems, bodily pain, and general health) and four domains reflect mental health. The scores in each subscale range from 0 to 100, with higher scores representing better quality of life. Additionally, the results from these eight subscales can be compressed into two composite scores that describe physical and mental health. In this study, the physical component summary (PCS) score was used to measure physical health functioning. Cronbach’s alpha coefficient for the PCS was 0.94 in this sample.

Body mass index (BMI) was calculated as weight (kg) divided by height squared (m2).

Procedure

The study data were collected during the subjects’ visit to their primary care physician. All of the subjects were asked if they want to participate in a study that aims to examine the quality of life. Those who agreed signed an informed consent to participate in the study. The subjects were requested to complete the self-assessment questionnaires, and research personnel obtained the height and weight for each participant. The Institutional Review Board consent was obtained for this study.

Statistical analysis

All of the analyses were conducted using SPSS version 15. A two-way analysis of variance was used to test the differences among three groups of patients with different BMI categories (overweight, obesity class I, class II and class III) in quality-of-life scores between men and women. For correlation analyses involving HRQoL, BMI, and psychological measures, we used Pearson correlation coefficients. Finally, to evaluate whether different aspects of physical quality of life mediate between BMI, anxiety and depression, we used the regression analysis.

Results

We examined the differences in health-related quality of life among three groups of patients of both sex: overweight, those with obesity class I, and those with obesity class II and III. Participants with obesity class II and III were grouped together because of the small number of subjects in these categories. Hereinafter referred, we will use the term severely obese for a description of this category of patients.

The main effect of the sex of the patients was significant (Table 1); women in comparison to men had worst physical functioning, more role limitations due to physical problems, more bodily pain, and lower composite score reflecting worst physical health. There was a significant main effect of BMI on composite score reflecting physical health, and physical functioning. Employing the Scheffe post hoc test, we found that severely obese patients significantly differ from overweight patients and from the group of obese in class I. There was a significant main effect of BMI on bodily pain. Employing the Scheffe post hoc test, significant differences were found between overweight and severely obese. Severely obese patients feel more bodily pain in comparison to overweight one. There was no significant interaction between sex and BMI of patients.

Table 1 Scores of health-related quality of life by obesity classes for men and women
Full size table

Regarding the level of anxiety and depression, 50.9 % of the patients did not express anxiety symptoms, 25.1 % of the patients showed signs of possible anxiety, and 24 % of the patients had anxiety symptoms (14 % women and 10 % man). In terms of depressive symptoms, 56.8 % of the patients were not depressed, 25.3 % of the patients had possible depression, and 17.9 % of the patients were depressed (11.9 % women and 6 % man).

We aimed to verify the relationship between psychological characteristics, the quality of life, and BMI. For this purpose, we conducted a correlation analysis. A correlation matrix (Tables 2 and 3) was separately computed on the results of men and of women because of the expected gender differences in the correlation of the examined variables. As this study included participants in a wide range of age (21–60 years old), we considered the necessity to include age as a control variable in the correlation analysis. The average age of the respondents in our sample was 46.68 years. Due to changes in men, and especially in women that occur around the fifties, it is necessary to control the age because of the possibility of interfering with the interpretation of results. Factors closely related with physical and mental quality of life are menopause and andropause symptoms that are not measured in the study. Recent studies suggest that women become increasingly vulnerable during the menopausal transition to declines in physical and role function and increases in depressive symptoms [22]. These biological shifts may determine age-related physical changes (e.g. increased BMI) during mid-life, and for this reason we controlled the respondents age. Beside a zero-order correlation, a first-order partial correlation was computed to explore the relationship between measured variables, controlling for the effects of age. If the age affects BMI and different domains of physical quality of life, anxiety and depression, the partial correlation between mentioned variables should be lessened or no longer significant.

Table 2 Zero-order and first-order correlation among all measured variables for men (N = 130)
Full size table
Table 3 Zero-order and first-order correlation among all measured variables for women (N = 143)
Full size table

In men, the BMI results are unrelated to different aspects of physical functioning, anxiety, and depression when controlling for age (Table 2). We found only a low, although significant, correlation between BMI and physical functioning (r = −0.22; p < .05).

In women, the first-order correlations between BMI and physical component (PC) of QoL, physical functioning, role limitation due to physical problems, and depression were found to be statistically significant, indicating that a relationship between those variables exists above and beyond the effects of age, but the relationship is lessened after controlling the age. These correlations ranged from r = 0.22 (p < .01) between BMI and depression, to r = −0.25 (p < 0.01) between BMI and PC of QoL and physical functioning (Table 3). Higher BMI levels in women were associated with lower physical functioning, more role limitations because of physical problems, and more depression. The lower perceived physical QoL in all of the domains is related to higher levels of anxiety and depression in women. The results of the female sample also revealed that when controlling the effects of age, the relationship between BMI, bodily pain, general health, and anxiety is no longer significant. It means that age is the determinant of more bodily pain, lower general health, and more anxiety. Apparently, age affects both BMI and different domains of physical quality of life, anxiety and depression and is closely related to them.

To explore the mediational effects of the PC of HRQoL and its four subscales in relation between BMI as a predictor and the negative affect as criterion, we conducted a series of regression analyses. We evaluated the mediating effects of the PC of QoL and its subscales on the relationship of BMI to the anxiety and depression subscale scores using the Baron and Kenny approach. The component of QoL is considered to be a mediator if: (1) the initial variable (BMI) is significantly associated with the outcome (anxiety or depression), (2) the initial variable (BMI) is significantly associated with the mediator (different aspect of physical QoL), and (3) the mediator is significantly associated with the outcome (anxiety or depression) after controlling for the effects of the initial variable (BMI). Baron and Kenny [23] discussed a fourth condition concerning the reduction of the initial and outcome variables relationship for evidence of partial or complete mediation. Specifically, if a test for the partial relationship between the initial variable and the outcome variable is non-significant, this finding might suggest that the mediator is a complete (or near-complete) mediator of the effect of the initial variable on the outcome variable.

Because of the specificity related to gender, which is described in the introduction, the analysis of mediation effects was separately calculated for men and women. In addition, to exclude the effects of age on the variables in the analyses, a simple linear regression was used. Instead of the original variables, all of the mediation analyses were performed with the residuals from the above-mentioned model.

In a sample of men, the first two conditions for the use of mediation analysis were not satisfied: the correlation between predictor and criteria, and the correlation between predictor and mediators. The Baron and Kenny approach was not met; that is, there was no relationship to be mediated. The PC of quality of life, as any of its components, does not mediate between BMI and anxiety or depression in men. In a sample of women, we found the same situation for anxiety. The PC of quality of life, as any of its components, does not mediate between BMI and anxiety.

In women, the PC of QoL mediates between BMI and symptoms of depression (Table 4). There was statistically significant relationship between the BMI and depression (β = 0.21, p < 0.01). After controlling for the PC, the magnitude of this association decreased and was no longer statistically significant, indicating that the PC of QoL completely mediated the relationship between BMI and depressive symptoms in women (β = 0.10). Women with higher BMI levels have a greater likelihood of decreasing their physical quality of life, which in turn has a direct effect on the development of depressive symptoms. This mediation ratio also applies to the physical functioning and role limitations due to physical problems (Table 4). After controlling for the physical functioning, the magnitude of the association between BMI and depression decreased (β = 0.10 < β = 0.21) and was no longer statistically significant, indicating that the physical functioning completely mediated the relationship between BMI and depressive symptoms in women. Women with higher BMI levels have a greater likelihood of decreasing their physical functioning, which in turn has a direct effect on the development of depressive symptoms. After controlling for the role limitation due to physical problems, the magnitude of the association between BMI and depression also decreased (β = 0.14 < β = 0.21) and was no longer statistically significant, indicating the complete mediation between BMI and depressive symptoms in women. Women with higher BMI levels have a greater likelihood of increasing their role limitations due to physical problems, which in turn has a direct effect on the development of depressive symptoms.

Table 4 Results of regression analyses—evaluation of mediational effects of different subscales of physical quality of life in the relationship between BMI and depressive symptoms in women, when controlling for the effects of age
Full size table

Discussion

In the present research, we explored the differences in physical health functioning in a community sample of overweight and obese individuals. The results supported the hypothesis of the existence of gender differences in quality of life. The major decrease was observed for women in comparison to men, in the domains reflecting PC of quality of life, which is consistent with the literature [13, 14, 24–26] but also in physical functioning, role limitations due to physical problems, and bodily pain. Obese women report worse physical functioning than men do, especially because of the experience of fatigue and sleepiness [27]. The prevalence of most common forms of pain is higher among women then men. They report greater pain with a long-term duration, [28], and display enhanced sensitivity to most forms of pain [29]. Stone et al. [30] found that obese individuals report more daily pain compared to overweight persons, and the obesity–pain association was stronger for women than for men. Various comorbidities and functional limitations associated with obesity can adversely affect physical quality of life. Obese women in our sample have more difficulties in physical functioning; the pain that they feel is very high, and they have many difficulties in performing their daily tasks, professional and family roles.

Our assumption was that there are gender differences in quality of life and we therefore did analyses separately for men and women. There are a few reasons for analysing men and women separately. First, this is a usual way to adjust for the possible confounding effect of sex differences on the relationship between obesity and quality of life. Second, there are large differences in the distribution of risks factors related to obesity between men and women for some variables (cardiovascular risks, or major depression risk), such that subjects of both sex cannot be considered as belonging to a single population for statistical analysis [31].

For women, there is a low but significant correlation between BMI and some of the measured aspects of physical quality of life, such as physical functioning, role limitations due to physical problems and PC of HRQoL, when controlling for age. It is likely that because of their weight, they have lower physical functioning; their weight might interfere with performing gender roles, such as taking care of the household and managing the family. Other authors have obtained similar results and the encumbrance on physical quality of life appeared to affect women more than men [31]. A higher BMI level was associated with lower PC, physical functioning [32], and role limitations [33]. In part, this phenomenon can be attributed to the higher prevalence of concurrent somatic diseases and psychopathological disturbances in morbidly obese women, compared to those with lower degrees of obesity. When controlling for age in men, the only relation we found was between BMI and physical functioning. A higher BMI level appeared to affect the physical functioning, functional impairment and physical discomfort.

Another objective in this study was to examine the association between body mass index, depression and anxiety as well as the potential mediating effects of physical health functioning on this association. The results indicated that some aspects of the physical aspects of HRQoL (PC of QoL, physical functioning and role limitations) mediate the relationship between BMI and level of depression, but only in women. In our research, a higher level of body mass index decreased the quality of life, a different physical aspect of health, which became a potential risk factor for the development of depressive symptoms. This pattern is consistent with previous research indicating that women, compared to men, perceived their weight as a barrier in their physical quality of life and developed a higher rate of depression [34]. In men, this condition does not apply; men do not see their weight as a factor that diminishes or worsens their quality of life. They do not associate their weight to poorer quality of life (e.g. general health, the experience of pain, or life roles). Fabricatore et al. [35] in a research on extremely obese participants (81 % women) also found that impairments in HRQoL were common. More than 40 % of participants scored in the impaired ranges of physical functioning, physical role limitations, and bodily pain. The impairments in HRQoL were related, significantly and more strongly, to symptoms of depression than were related to BMI. Jagielski et al. [15] found that the association between quality of life and BMI remains stable after controlling for comorbid health problems, and the authors conclude that obesity independently, negatively affects quality of life. Kolotkin et al. [36] found that HRQOL was more impaired for those with higher BMIs, and women in treatment groups. Recently, Buscemi et al. [37] demonstrate that just a modest amount of weight loss can improve physical and psychological quality of life.

In our study, we did not get the results that indicate physical component of HRQoL and its different aspects as mediators between BMI and anxiety. We found that older female respondents were more anxious, but this is not the case for men. Female ageing is synonymous of a decline in physical attractiveness, and it is not surprising that middle-aged women experience anxiety about ageing [38]. Gender differences in HRQoL could be also related to the higher prevalence of psychopathology among women [39, 40] or to a greater cultural drive for thinness experienced by the female sex in Western societies [41, 42]. The presence of depressed mood, which is the most common psychological problem observed in obese women [43], can increase subjective distress induced by disease-related physical symptoms and functional impairment [39]. These are different health treats associated with obesity and they have an influence on the perception of stress in obese people. On the other hand, obese people are exposed to the negative effects of stigmatisation and discrimination related to weight. Because of high frequency of stigmatisation and poor coping strategies, the probability to develop depressive mood as consequence becomes higher [15, 44]. Murphy et al. [44] concluded that depression among obese subjects in a community sample tends to be more severe in comparison to non-obese, and gaining weight is important factor that contributes to the severity of the depressive symptoms. According to our results, a portion of the variance of depressive symptoms in women could be explained by BMI. Women are likely to be more aware of the connection between obesity and health, in contrast to men in whom this aspect of life is not of primary importance.

The results in the present study show a different pattern of functioning between men and women. In women, the quality of life is significantly weakened because of increasing body weight, which in turn produces symptoms of depression. In men, this association does not exist. Preiss et al. [45] analysed results obtained in 20 studies, and in more than half found that gender was significantly associated with the relationship between obesity and depression, such that being female conferred risk of comorbid obesity and depression. Obese women experienced a greater impairment of HRQL than their male counterparts did. This confirms previous reports in clinic-based samples, and in population studies. Knowledge of gender differences allows us to better plan therapeutic interventions for men and women with high BMI levels. It is important to identify the factors that can effectively motivate and stimulate obese people to modify their lifestyles and regimens. This research emphasised the importance of assessing the differences in the psychological functioning of women and men to promote effective weight loss treatment.

In this study, we did not include information about the presence of eating disorders in a sample of obese patients, which can potentially be one of the limitations. However, research on the relationship between binge eating disorder with quality of life is not consistent and we have decided to focus on depression and anxiety as criteria. Some studies analysed association between BED and HRQOL impairment and found that this association can be mediated by higher BMI, and a greater prevalence of mental disorders [42, 43]. After adjustment for these potential confounders, binge eating was only marginally associated with some domains of HRQL, without any impact on physical scales. Kolotkin et al. [46] suggested that factors such as BMI, psychological symptoms, and demographic variables explained the association between BED and quality of life. Because earlier studies did not control for pertinent variables, the implication that the presence of BED itself accounts for the observed differences in quality of life may not be correct, Kolotkin et al. [46] concluded. Preiss et al. [45] in their article wrote that the results of research on BED might reflect an influence of binge eating that is specific to an obese (non-surgical) treatment-seeking population. On the other hand, in our research, we had a community non-clinical obese sample. There are several other limitations to our study. One limitation is related to the sample size. Therefore, it is necessary to conduct additional research with larger samples and to include information concerning co-morbidity, which is lacking in this research. Specifically, general practitioners collected the data; however, the information related to the reason for the visit to the physician was not recorded. We can only speculate that the potential reason for visiting the primary care physician was related to some of co-morbidity of obesity. In conclusion, these findings are of clinical importance, indicating that the impairment of the quality of life in obese must be of high importance in planning treatment with these individuals.