Food addiction and emotional eating are associated with intradaily rest–activity rhythm variability

The aim of the present investigation was to study the associations among parameters characterizing eating behavior and actimetry-derived indices of circadian rhythm of motor activity. The study involved 81 healthy participants (average age: 21.5 ± 9.6 y, women: 77.8%). Each study participant provided personal data, filled out the Yale Food Addiction Scale and the Dutch Eating Behavior Questionnaire, and wore a wrist actimeter for 7 consecutive days to record motor activity. Using time series treatments, we obtained: (a) three cosinor-derived parametric indices [Medline Estimating Statistics of Rhythm (MESOR), amplitude, and acrophase], and (b) four non-parametric indices [interdaily stability, intradaily variability (I.V.), most active 10-h period (M10), and least active 5-h period] characterizing the 24-h rhythm of motor activity. A multiple regression analysis adjusted for age, sex, and BMI was performed to assess the associations among the studied indicators. It was shown that I.V. is a predictor of symptoms of food addiction (β = 0.242, P = 0.037) and emotional eating (β = 0.390, P = 0.004), MESOR is a predictor of symptoms of food addiction (β = 0.342, P = 0.003), and M10 predicts restraint (β = 0.257, P = 0.015) and emotional eating (β = 0.464, P = 0.001). It was shown for the first time that an increase in symptom counts of food addiction is associated with an increase in the average level and fragmentation of 24-h rhythm of motor activity. Level V, cross-sectional descriptive study.


Introduction
The circadian system (CS) is a complexly organized, multilevel system that was formed in the process of evolution. It consists of specialized molecules, cells, and organs that are present in almost all organisms living on the Earth's surface, including humans [1]. On one hand, it ensures the maintenance of an endogenous circadian rhythm of biochemical, physiological, and behavioral reactions in the body; on the other, it ensures synchronization of endogenous rhythms with the daily rhythm of environment [1]. CS plays a crucial role in ensuring the adaptation of living organisms to life in a 24-h environment [2]. Despite the fact that modern humans' dependence on environmental factors has significantly decreased, the role of CS in our adaptation to life in a social environment is extremely important. This is evidenced in the results of numerous studies on the impact of CS dysfunction observed during jetlag and shift work on human well-being and even health. It was shown that shift work is a risk factor for cognitive, psycho-emotional, and eating disorders [3][4][5][6].
The recently developed Munich Chronotype Questionnaire (MCTQ) [7], which allows the state of the human circadian system to be quantified in real conditions [7], has been used to conduct extensive population studies [8]. As a result of these studies, another form of circadian misalignment 1 3 was described-social jetlag (SJL)-which arises due to a mismatch between biological and social clocks [9]. The incidence rate of SJL was found to be highest in individuals with a late chronotype, i.e., those whose biorhythms do not coincide with the rhythms of social life in modern society, for example, an early start to work or classes. As compared with other forms of circadian misalignment, SJL has a weaker negative impact on human health; however, at the same time, it is much more common, especially among young students. It has been shown that the detection rate of SJL varies from 40.1% in Japan [10] and 69% in Central Europe [11], to 86.4% in Russia [12]. Numerous studies have identified a number of negative consequences of SJL, including impaired academic performance [13], psycho-emotional state [14], and sleep quality [15] of young people. Individuals with SJL have unhealthy dietary preferences [15,16] and signs of obesity [17,18], metabolic syndrome [19], and type 2 diabetes [20]. Given that eating disorders are one of the risk factors for obesity and metabolic disorders, it is to be expected that there is an association between eating disorders and the circadian misalignment.
SJL has been shown to be associated with increased consumption of nicotine [9] and alcohol [21,22] , indirectly confirming the hypothesis that SJL is a predictor of FA. However, a direct study of association between SJL and FA conducted on a large sample did not confirm this hypothesis [28]. There was also no association of SJL with restraint, external, and emotional eating behaviors [29]. One of the possible reasons for the negative result is that the subjective method (MCTQ) used in this study to assess the state of the circadian system was not sensitive enough to identify the association between eating behavior and circadian misalignment.
Currently, in scientific research and clinical practice, actimetry is widely used as an objective method for assessing the state of CS. Actimetry has been successfully used for the objective assessment of sleep function characteristics [ [36,37]. Significant changes in the non-parametric indices of 24-h RAR were also observed in patients with BED, e.g., an increase in interdaily stability (IS) and a decrease in the least active 5-h period (L5) and most active 10-h period (M10) [37]. Currently, there are no studies devoted to the study of 24-h RAR in individuals with FA. An association between FA and emotional eating has been shown in previous studies [25,38,39]. Given this fact, it can be assumed that FA and emotional eating will show similarities in the profiles of the daily rhythm of motor activity.
The aim of this study was to analyze the relationships between indicators characterizing eating behavior and objective indicators characterizing the daily rhythm of activity and rest.

Study design
This publication presents the results of the second phase of research, which was dedicated to the study of the external and internal factors associated with FA. In the first stage, the associations among the FA detection rate, socio-demographic, anthropometric, and psychophysiological indicators of healthy residents of four Russian cities were assessed. The results of the first stage were published earlier [28,39]. The first-stage participants from Syktyvkar and Tyumen (Lat./ Long.: 61.7/50.9 and 57.2/65.5, respectively) were invited to participate in the second phase of the study to assess 24-h RAR using actimetry. The inclusion criteria were as follows: signing informed consent and filling in all the first-phase questionnaires. Exclusion criteria: night/shift work and sleep disorders.

Study participants
The study was conducted from February 2019 to March 2020 upon 81 voluntary and anonymous participants aged 18-62 years. Most of the study participants were university students. In addition, teachers, researchers, and postgraduates took part in the study. The research program was approved by the Ethics Committee of the Institute of Physiology of Komi Scientific Center of the Ural Branch of the Russian Academy of Sciences. Each study participant signed an informed consent form.

Instruments
Each study participant indicated their sex, age, height, and weight, and completed the Yale Food Addiction Scale (YFAS) [25] and the Dutch Eating Behavior Questionnaire (DEBQ) [40]. Weight and height were used to calculate body mass index (BMI) as weight in kilograms divided by height in meters squared. Sex-and age-specific BMI percentiles were calculated using BMI growth charts [41]. Table 1 shows mean values of the studied indicators, standard deviations, skewness, and kurtosis. Almost all indicators, with the exception of L5, have normal distribution, so they can be used in the analyses without any modifications. A modified indicator (2 + lnL5), which has a normal distribution was used in the analyses.

YFAS
In this study, we used a Russian translation of YFAS. In the previous study [39], satisfactory internal consistency of YFAS-Rus was shown. The validity of YFAS-Rus was confirmed by the existence of a significant association between FA and the Emotional Eating subscale of DEBQ [39]. The scale consists of 25 questions and describes 7 diagnostic criteria for drug addiction (such as tolerance, withdrawal, and loss of control) and clinically significant eating disorders. The results of the treatment are presented in the form of (a) a quantitative indicator equal to the sum of confirmed symptoms (SC, varying within the range of 0 to 7), and (b) a qualitative indicator corresponding to a clinically significant eating disorder in the presence of three or more symptoms of FA. The Cronbach's α for this sample was 0.87.

DEBQ
In this study, we used Russian translation of DEBQ. In the previous study [39], satisfactory internal consistency of DEBQ-Rus was shown. The test consists of 33 questions divided into 3 subscales evaluating 3 types of eating behavior: restraint (DEBQ restr , 10 points), external (DEBQ extern , 10 points), and emotional (DEBQ emo , 13 points). A Likert-type scale ranging from 1 to 5 is used in response to all items. To evaluate the results, the sum of points for each subscale is calculated separately and divided by the number of points. The Cronbach's α values for DEBQ, DEBQ restr , DEBQ extern , and DEBQ emo were 0.88, 0.92, 0.52, and 0.94, respectively.

Wrist actimetry
Each participant wore a Daqtometer v2.4 actimeter (Daqtix, Germany) on the wrist of their nondominant hand continuously for 1 week. The activity was measured at 1 Hz and the values were summed up once a minute and expressed in arbitrary units (a.u.). The study participants kept a diary in which they indicated the times that they removed the actimeter to engage in activities involving water, as well as the times that they went to bed in the evening and got out of bed in the morning. The Chronos-Fit software package (Zuther P. and Lemmer B., Germany) was used to calculate the following parametric indices: Midline Estimating Statistic of Rhythm (MESOR), amplitude (A) = the maximum departure of the value of a sine wave from the average value, and acrophase (φ) = the time period in a cycle during which the cycle peaks, especially the upper part of a sine wave [42,43]. The whole dataset from each person was approximated by a sine function with a fixed period (tau = 24 h). In addition, non-parametric indices were also calculated: M10, L5 [44], IS, and intradaily variability (I.V.) [45].

Statistical analyses
We used SPSS version 20 (SPSS, Inc., Chicago, IL, USA) software package for the statistical analysis. Descriptive statistics (mean, standard deviation) of variables were calculated. Multiple regression analyses were performed, in which SC, DEBQ restr , DEBQ extern , and DEBQ emo were used as dependent variables, and sex, age, BMI, Mesor, A, φ, M10, 2 + lnL5, IS, and I.V. were used as independent variables (predictors). A stepwise inclusion procedure was used to determine the final set of predictors used in the model. To assess multicollinearity, the variance inflation factor (VIF) was assessed. A predictor was excluded from the analysis if VIF was ≥ 5.

Results
The results of the study showed that 13.6% of the individuals we examined were overweight or obese ( Table 2). The incidence of FA in study participants was 12.3% (Table 2), and the mean SC of FA was 2.31 ± 1.73 (Table 1). SC of FA were higher in women, in persons with overweight/ obesity, and in young adults (Table 3, Model 1). DEBQ restr were higher in individuals with overweight/obesity, and in females (Table 3, Model 2), whereas DEBQ extern was higher in females (Table 3, Model 3). There was a positive association between MESOR and SC of FA (Table 3, Model 1), indicating that individuals with pronounced signs of FA had a higher average daily level of motor activity.
A positive association of I.V. with SC of FA and DEBQ emo was noted (Table 3, Models 1&4). These data indicate that the fragmentation of the circadian rhythm of motor activity is characteristic of persons with SC of FA and emotional eating.
A positive association of M10 with DEBQ restr and DEBQ emo were noted ( Table 3, Model 2&4). These data indicate that persons with restraint and emotional types of eating behavior are characterized by an increased level of motor activity in the daytime.

Discussion
To the best of our knowledge, this is the first report on the association between FA and actimetry-derived 24-h RAR characteristics. Studies focused on 24-h RAR characteristics in BED patients have been previously conducted [36,37]. Individuals with BED have been shown to have a decrease in amplitude and MESOR [36,37]. Women with BED during multidisciplinary therapy [37] had higher IS values than the control group, and lower L5 and M10 values, and no statistically significant differences in I.V. and φ of 24-h RAR. The authors concluded that multidisciplinary therapy leads to improved sleep-wake rhythm characteristics in patients with BED, primarily IS and L5. In this study, we demonstrated that the nature of the association between FA and 24-h RAR characteristics differs significantly from those in BED patients. We noted a positive association between SC of FA and MESOR and did not detect an association between SC of FA and the amplitude and acrophase of 24-h RAR. In addition, we demonstrated for the first time that, of all the studied non-parametric indicators of 24-h RAR, only the I.V. index was significantly associated with SC of FA. It has been repeatedly shown that BED and FA have similarities [46][47][48]. Both eating disorders are more common in women, and in overweight individuals. Both types of eating disorders are characterized by a close association with impulsivity and depression [46]. Moreover, the difference between these two types of eating disorders, mainly related to behavior, was noted. There was the least close correlation between the behavioral component of BED and SC of FA [48]. It can be assumed that actimetry-derived 24-h RAR is a sensitive method for detecting differences between these two types of eating disorders. To test the validity of this assumption, it is necessary to conduct comparative studies of 24-h RAR characteristics in BED and FA. The present study showed that an objective assessment of motor activity rhythm is a convenient and sensitive method for studying the associations between the states of the circadian system and eating behavior. Two features of motor activity rhythm characteristic of persons with FA and emotional eating were identified. First, it was shown that individuals with symptoms of FA and emotional eating had a more pronounced fragmentation of the 24-h rhythm of motor activity. Previously, it was shown that FA is more often detected in young people with a late chronotype [27]. At the same time, there was no association between SJL and FA [28]. This negative result may be due to the fact that MCTQ is not sensitive enough to study the association between the states of the circadian system and eating behavior. SJL is primarily a measure of the instability of the phase of circadian rhythm during a calendar week. Judging by the data we obtained, fragmentation of the circadian rhythms of motor activity is a characteristic trait of persons with FA and emotional eating. The relationship between FA and emotional eating has been repeatedly noted in the past [25, 38,39], indicating similarities between these two characteristics of eating behavior.
Second, we showed that persons with high SC of FA have a higher average daily level of motor activity (MESOR), and that the profile of the daily rhythm of motor activity of people with restraint and emotional eating behaviors is characterized by an increased level of motor activity during the daytime (M10). These results were unexpected, since it is well known that FA is a risk factor for obesity [25,39]. Obesity, in turn, is associated with a decrease in the level of motor activity [49,50]. However, there is quite often a weak or nonlinear association between SC of FA and the morphological signs of obesity [51][52][53]. In addition, FA and emotional eating are just some of the many risk factors for obesity. A distinctive character trait of people with FA and emotional eating is impulsivity of behavior [54,55]. A positive association between impulsivity and the level of motor activity in the daytime has been repeatedly noted in previous research [56,57].
The data obtained indicate the presence of a significant association between FA and socio-demographic indicators: higher SC of FA was more often detected in women and in overweight people, and there was also a decrease in the SC of FA with age. These data are consistent with the results of previous studies [58,59] and indicate that the study was conducted on a representative sample.
The data obtained have theoretical and practical significance. The theoretical significance lies in the fact that these data make a certain contribution to the development of the FA concept. There are still a number of specialists who express doubt concerning the need to distinguish FA as an independent eating disorder [60] on the grounds that there are no specific features distinguishing this disorder from existing ones, for example, BED. Our data show that such differences exist and can be measured using an objective method. The practical significance lies in the fact that the proposed quantitative indicator can be used to specify the diagnosis of FA and monitor the course of remission of an eating disorder.

Strength and limitations
To our knowledge, this is a first investigation of 24-h RAR characteristics in persons with SC of FA. The study has a number of limitations. A significant bias toward women among the study participants could have affected the results of the analysis. The cross-sectional design of the study did not allow us to judge the causal associations between the studied indicators. The study involved practically healthy people whose anthropometric data and eating behavior were evaluated by self-assessment, so the results obtained cannot be used in clinical practice.

Conclusions
Overall, the current study shows that actimetry is a sensitive tool for studying the relationship between 24-h rhythms and eating behavior. There was an increase in fragmentation of the rhythm of motor activity in individuals with symptoms of FA and emotional eating. There was also an increased level of motor activity in people with FA and restraint and emotional eating.

What is already known on this subject?
Circadian misalignment is associated with unhealthy food preferences and increased risk of obesity. Previous studies have not found an association between a subjective measure of circadian misalignment (social jetlag) and FA.

What your study adds?
This study shows, for the first time, that an increase in intradaily variability-an objective indicator of 24-h RAR impairment-is positively associated with SC of FA. Preliminary data were also obtained indicating that the characteristics of 24-h RAR in individuals with FA are different from those in individuals with BED.