Introduction

The aim of this first study focusing on the relationship between traits postulated by Strelau’s Regulative Theory of Temperament (RTT) and obesity was to demonstrate that temperament may contribute to the development of obesity in adults. The second aim was to show that the same temperament traits can be a risk factor for affective disorders in obese patients.

According to the World Health Organization [1], “obesity may be defined as a condition of abnormal or excessive fat accumulation in adipose tissue, to the extent that health may be impaired” (p. 6). The most common and practical way of assessing nutritional status in adults is to calculate their body mass index (BMI), which is also called the Quetelet’s Index. It is measured by dividing a weight in kilograms by height in square meters. If BMI exceeds 30 kg/m2, a person is considered obese.

Risk factors in the development of obesity include, among others, genetic factors, family lifestyle, poor activity, unhealthy diet and eating habits. These factors indicate the role of human personality in the development of obesity.

In a number of works that address the associations between personality and obesity, it was found, for example, that an increase in BMI ratio is related to high neuroticism, extraversion and low conscientiousness, as well as to the facets of these traits related to difficulties with impulse control [2]. Conscientiousness is an important feature associated with self-discipline and control over impulses. The role of conscientiousness in the development of obesity was shown in a recently conducted meta-analysis [3]. Low levels of conscientiousness in children are a good predictor of obesity in adulthood [4]. In adults, low levels of self-discipline as a component of conscientiousness and a high level of impulsivity as a component of neuroticism are distinctly associated with obesity [2]. Adults with low conscientious exhibited less healthy behaviour, including limited physical activity and poor control of their diets, both of which are associated with adiposity [3]. Obese subjects also have higher scores on scales measuring major depression, thought disorder, anxiety, somatoform disorder and alcohol dependence, and differ from the control group with regard to Cloninger’s temperament traits—that is, the obese patients showed higher levels of harm avoidance and lower scores in novelty seeking, persistence and self-transcendence in comparison with the controls [5]. However, an earlier study demonstrated that, compared with lean subjects, obese subjects scored higher in novelty seeking and lower in persistence and self-directedness [6]. Hypomania and psychopathy are also considered good predictors of BMI ratio [7]. The results obtained, however, were considered to be inconclusive, indicating the severity of pathological personality traits in the obese rather than normal dimensions of personality functioning.

The complications of obesity are primarily associated with health problems such as type 2 diabetes, hypertension, cardiovascular disease and sleep apnea, but also with low quality of life. Factors affecting the quality of life of obese people include low work achievements, stigma, disability, sexual problems and depression.

Many previous works indicated the relationship of obesity to depression [8], especially in women, as opposed to men [9, 10]. The results, comprised primarily of correlational studies, are not conclusive and do not indicate whether the depression is predictive of the development of obesity or, in turn, whether obesity is predictive of the development of depression. This difficulty may also stem from the complex neuroimmunological and physiological mechanisms underlying obesity and depression [11]. Newer studies indicate, however, that such a relationship between obesity and depression is not obvious [12] or obesity would be more strongly related to somatic-affective depressive symptoms (e.g. fatigue, lack of energy, work difficulty) than to cognitive-affective depressive symptoms (e.g. pessimism, sense of failure, social withdrawal) [13, 14]. It has also been hypothesized that obesity might be related to bipolar disorder. A common impulsive-addictive diathesis is proposed as a mediating mechanism [15].

The somatic symptoms accompanying obesity and the somatic-affective criteria for depression suggest that RTT traits [16] may be identified as a moderator for both obesity and affective disorders accompanying obesity.

The RTT assumes that temperamental traits present in humans since infancy—briskness (BR), perseveration (PE), sensory sensitivity (SS), emotional reactivity (ER), endurance (EN) and activity (AC)—are predominantly related to formal (energetic and temporal) attributes of behaviour [17], implying that these traits are important components of human behaviour in all life situations.

Emotional reactivity and perseveration show large phenotypic and genetic links to neuroticism, while briskness, endurance and activity are phenotypically and genetically correlated with extraversion, as described in the five-factor personality theory [18]. Thus, it was expected that the RTT traits will play a role similar to the Big Five traits as risk factors in the development of affective disorders in obese individuals [19].

This study had two broad objectives: (1) to demonstrate that RTT temperament traits in adults may contribute to the development of obesity and (2) to support the idea that RTT traits may also contribute to the development of affective disorders in obese patients.

It is hypothesized that (1) briskness, endurance and activity, which are involved in the regulation of behaviour designed to discharge energy, are three particularly important temperament traits in the development of obesity; and (2) high levels of ER and PE may indicate a greater risk in obese women than men of developing affective disorders.

It was thought that levels of BR, EN and AC in obese patients (men and women) would be lower than in people of normal weight—in other words, that obese people would be characterised by reduced physical activity, slowness and low mobility behaviour, with little endurance and a low fatigue threshold. Sensory sensitivity does not seem to be particularly important in obesity. Emotional reactivity and perseveration may increase, reflecting the tendency of obese people to react more strongly to distress and to repeat and maintain the same behaviour and reactions, which reduces energy expenditure.

A high level of emotional reactivity and perseveration may be more typical of obese women than obese men. Women, in general, have higher levels of both these traits [20, 21]. In two previous studies, a positive correlation was shown between levels of ER and PE with cancer trauma symptoms [22] and with trauma symptoms in HIV-positive men and women [23]. In both studies cited, women had higher levels of ER and PE than men. Perseveration also proved to be a good predictor of somatic anxiety [24].

Methods

Participants

All participants, in both groups, were only white Caucasians. The participants from the clinical sample included 163 obese bariatric patients, comprised 59 men and 104 women aged 19–80 years (M = 45.38; SD = 12.60). In this group, height (in cm) ranged from 152 to 198 (M = 170.95; SD = 10.05), weight (in kg) ranged from 100 to 214 (M = 127.85; SD = 20.15), and BMI ranged from 35.4 to 64.5 (M = 43.80; SD = 5.95). All patients were treated surgically at Warsaw Medical University Hospital and the Collegium Medicum in Bydgoszcz (Poland). The inclusion criteria were BMI ≥ 35 kg/m2, age ≥18 years, and the existing or anticipated ineffectiveness of conservative treatment of obesity. In all patients, gastric bypass surgery had been performed. All patients had previously been unsuccessfully treated by interventions that included changes to eating habits, intense physical activity and pharmacological weight loss support. Surgical treatment was based on medical indicators, with the consent of the patients. Ninety percent of patients had at least one disease accompanying their obesity (hypertension, abnormal levels of blood glucose/diabetes, lipid disorders or obstructive sleep apnea). Patients with absolute contraindications for bariatric surgery, such as pregnant women, individuals with uncontrolled psychotic or depressive disorders, alcohol or drug dependencies, and patients with cancer or extreme heart disease, were excluded from surgery and did not participate in the study. Psychiatric and medical assessments of candidates for bariatric surgery were conducted at least 3 months prior to hospitalization. In the obese patients group, 55 participants (33.7 %) had attained higher education, 79 participants (48.5 %) had attained secondary education, and 29 participants (17.8 %) had attained less than secondary education.

The control group comprised 202 subjects recruited from the general population, with 89 men and 113 women aged 19–72 years (M = 47.06; SD = 12.00). In this group, height (in cm) ranged from 150 to 195 (M = 169.26; SD = 9.74), weight (in kg) ranged from 47 to 88 (M = 62.93; SD = 9.45), and BMI ranged from 19.3 to 24.7 (M = 21.84; SD = 1.49). The inclusion criteria were a BMI in the healthy weight range, age ≥ 18 years, and a lack of self-reported physical (somatic) or mental problems. In the control group, 75 participants (36.2 %) had attained higher education, 102 participants (49.3 %) had attained secondary education, and 30 participants (14.5 %) had attained primary education. Both groups were equivalent by gender [χ 2(1) = 0.71, p = 0.399] and by age [F (1, 364) = 1.70, p = 0.193]. The study was anonymous and participation was voluntary. Participants were not remunerated. All participants were informed about the purpose and approach of the study, and they provided written consent to participate.

The research project was accepted by the local Research Ethics Commission at the Faculty of Psychology, University of Warsaw.

Assessment

Temperament traits were assessed with the FCB-TI [25]. This inventory includes 120 items (20 items per scale). Respondents answered yes or no to each question. The inventory includes the following scales [17]: Briskness is defined as a tendency to react quickly, to maintain a high tempo in performing activities and to shift easily in response to changes in the surroundings from one behaviour (reaction) to another; Perseveration is defined as a tendency to continue and to repeat behaviour, and to experience emotional states after the cessation of stimuli (situations) that evoke this behaviour or these states; Sensory sensitivity is understood as an ability to react to sensory stimuli of low stimulative value, whereas emotional reactivity is a tendency to react intensively to emotion-generating stimuli, expressed in high emotional sensitivity and low emotional endurance—that is, a limited ability to control emotional reactions in response to strong emotogenic stimuli; Endurance is defined as an ability to respond adequately to situations demanding long-lasting or highly stimulative activity and/or under intensive external stimulation; and Activity refers to a tendency to undertake behaviours of high stimulative value or to provide, by means of behaviour, strong stimulation from the surroundings. The Cronbach’s alpha scores derived from the present samples were as follows (coefficients for the control group are given in the parentheses): Briskness 0.83 (0.82); Perseveration 0.78 (0.76); Sensory Sensitivity 0.72 (0.76); Emotional Reactivity 0.84 (0.81); Endurance 0.80 (0.86); and Activity 0.81 (85).

The temperament assessments were collected by the pollsters: medical students in hospitals and students of psychology in the control group. Among obese patients, temperament assessment was conducted 1 day prior to surgery.

Statistical analysis

The statistical analysis was performed with IBM SPSS Statistics 21 [26]. Since variances differed significantly between groups, the Mann–Whitney U test for two independent samples (Z values) was used.

Results

Table 1 shows differences in the studied variables between the obese and control groups.

Table 1 Means and standard deviations comparisons for temperament traits in obese patients group (n = 163) and control group (n = 202)
Full size table

Compared with the control group, the obese patients had significantly lower briskness, endurance and activity, and significantly higher perseveration. No significant difference in sensory sensitivity and emotional reactivity was found between the two groups.

The basic descriptive statistics for obese men and women are presented in Table 2 along with differences between the studied groups. The differences between men and women in the control group are shown comparatively in the same table.

Table 2 Means and standard deviations comparisons for temperament traits in obese male group (n = 59) and obese female group (n = 104)
Full size table

As shown by the data in Table 2, obese women had higher levels of PE and ER than obese men. Obese women also showed lower levels of activity compared with obese men. No significant differences were found between the two groups with regard to briskness, sensory sensitivity and endurance. In the control group, women had higher levels of PE and ER than men, but they had significantly lower briskness, endurance and activity. No significant gender difference in sensory sensitivity was found between men and women in the control group.

Discussion

This was the first study focused on RTT traits and the risk of obesity and affective disorders in obese adults. The first objective of this study was to demonstrate that RTT traits in adults may contribute to the development of obesity. The results indicate that obese individuals, irrespective of gender, tend to react slowly, to maintain a low tempo of activity and to show limited mobility of behaviour (low briskness). Obese persons may be described as not being disposed to making an intensive effort, as people who tire relatively quickly (low endurance) and as people who undertake only a few new tasks or activities (low activity) (see Table 1). Low levels of BR, EN and AC may be a consequence of clinical obesity that persists for many years. The large body weight of obese people limits their physical activity. This interpretation is supported by the fact that, unlike the control group, there were no differences in the levels of BR and EN between obese men and women (see Table 2).

The competitive hypothesis is that RTT temperament traits are already present in children and can be considered behaviour moderators leading to the development of overweight conditions and obesity in childhood and adolescence [16]. The high stability of RTT temperament traits during an individual’s lifetime suggests that temperament traits may also affect behaviour related to eating habits in adults [27]. Low levels of BR, EN and AC can reduce regular engagement in physical activity in children and adolescents, and promote sedentary behaviour, which is already a risk factor for the development of obesity in children [28]. It was also shown that PE is associated with being forced to eat pre-imposed portions in full (despite one’s sense of satiety), as perceived by adult women. A higher childhood PE reflects more internalized pressure from a woman’s mother to finish her food, with a likelihood that this eating pattern will persist from childhood into womanhood [29].

Low levels of BR, EN and AC in patients with clinical obesity are not conducive to non-invasive therapeutic action, like physical activity. This hypothesis also finds some support in the results of studies on the role of conscientiousness in the development of obesity [2, 3]. Conscientiousness is a trait correlated positively with BR and EN [25]. The results of recent studies indicate that it is possible to predict that young adults will have problems with obesity in midlife based on an assessment of conscientiousness, i.e. individuals low in conscientiousness are less likely to engage in active lifestyles and maintain healthy diets [30]. It is suggested, then, that low levels of BR, EN, and AC in earlier stages of development, in interaction with biological and environmental factors, may contribute to the development of obesity over the human lifespan.

To summarize this part of the study, it is possible to formulate a general hypothesis for further cohort studies. Temperament traits whose low levels in the earlier life stages are conducive to the development of overweight and obese individuals are further reduced in obese adults under the influence of the somatic effects of obesity. Low levels of BR, EN and AC may also be related to somatic-affective depressive symptoms, like fatigue, a lack of energy or a sense of insufficiency [13, 14].

The second objective of the study was to support the idea that RTT traits may also contribute to the development of affective disorders in obese patients. Compared to the control group, obese patients were found to be prone to longer-lasting reactions (high PE), which may result directly from their life experiences and low levels of briskness [25].

High levels of ER and PE, and low levels of BR, EN and AC were found in patients with bipolar disorder [16] (see Fig. 1).

Fig. 1
figure 1

Comparison of temperament traits in a group of obese patients, patients with bipolar disorder (BD) and controls

Full size image

Our latest study showed that patients with bipolar disorder and recurrent major depression are characterised by a high level of ER and a low level of BR, EN and AC (unpublished data).

The analysis of differences in temperament traits between men and women, however, suggests a greater risk of psychopathology in obese women than men. It has been shown previously that high levels of PE and ER, and low levels of BR, EN and AC, which are typical in obese women, are a risk factor for anxiety disorders [19, 22, 23, 31]. It is worth noting that the distribution of gender-related traits was similar in both the obese and control groups. In addition, the obese group may be additionally affected by social factors such as stigma, as indicated by some studies [32, 33], which increases the risk of psychopathology.

Compared with obese men, obese women have higher levels of ER and are willing to hold negative emotions for a longer period of time (high PE). These high levels of ER, as an equivalent of neuroticism, may contribute to the development of affective disorders in women. At the same time, activity as an RTT temperament trait related to extraversion, which is a buffer against the development of affective disorders, is decreased in women. Previous studies have shown that people with depression—in the case of both patients and the general population—have high levels of neuroticism and low levels of extraversion [34–39]. In this regard, the results observed in obese men and women seem consistent with the results related to neuroticism and extraversion. High levels of ER and PE may be factors in obese women’s vulnerability to the development of affective disorders, particularly depression. The risk seems to be even greater in obese people because insufficient physical activity can foster depression [40].

The overall finding from the study is that low levels of BR, EN and AC as temperamental traits responsible for the energetic aspects of behaviour may be responsible for the risk of obesity development and for obese people’s failure to reduce their body weight. The high ER and PE levels that accompany the lower levels of BR, EN and AC, especially in obese women, suggest their greater vulnerability to the development of affective disorders associated with obesity, in comparison with obese men.

It is possible that temperament-oriented programs could be useful in the treatment of obesity. The configuration of extremely low levels of temperament traits, such as BR, EN and AC, may help identify individuals who, despite weight loss and the resolution of obesity-related diseases, will not be able to “adapt” to the new life conditions. These findings also confirm that psychological care of patients at risk for obesity should focus on protecting against the emotional consequences of obesity, especially in obese women. Of course, our suggestions should be verified in a randomized controlled trial.

A significant limitation of the present study was the use of only one tool for diagnosing temperament; the introduction of other temperament dimensions would allow for verification of a number of hypotheses formulated here. Another important limitation was the lack of measurement of RTT traits in patients after bariatric surgery, which would assess the dynamics of analyzed traits. Neither stigma nor social support was controlled in the obese group. The group of participants should be expanded to include people who are overweight and obese, but not bariatric patients.