Appetite 82 (2014) 202–207
Contents lists available at ScienceDirect
Appetite j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / a p p e t
Research report
Self-control and parental control mediate the relationship between negative emotions and emotional eating among adolescents Hong Zhu, Xingwei Luo, Taisheng Cai *, Zhihua Li, Wenli Liu Medical Psychological Institute of The Second Xiangya Hospital, Central South University, Key Laboratory of Psychiatry and Mental Health of Hunan Province, Hunan Province Technology Institute of Psychiatry, Middle Ren-Min Road 139, Changsha, Hunan 410011, China
A R T I C L E
I N F O
Article history: Received 16 December 2013 Received in revised form 25 June 2014 Accepted 27 June 2014 Available online 30 July 2014 Keywords: Self-control Parental control Negative emotion Emotional eating Multiple mediation model Adolescents
A B S T R A C T
Objectives: The study was conducted to simultaneously investigate the mediating effects of parental control and adolescents’ self-control on the relationship between adolescents’ negative emotions and emotional eating, and to determine pathways with the greatest effect among these variables. Methods: Negative emotions, emotional eating, parental control, and self-control were investigated in 594 high school students (average age = 16.70, SD = 1.09) in Changsha City, China. Results: High levels of negative emotions and parental control and low levels of self-control were strongly related to high levels of emotional eating in adolescents. In addition to the direct relationship between negative emotions and emotional eating, there was a mediating effect observed through low self-control and high parental control. The mediational effect of parental control was non-significant in adolescent boys. Furthermore, negative emotions related to emotional eating through the effect of parental control on adolescents’ self-control. The degree to which both mediators explained the relationship between negative emotions and emotional eating ranged from 52.6% to 66.8%, and self-control had a stronger mediational effect than did parental control. Conclusion: The results indicate that both self-control and parental control should be considered in designing preventative measures against emotional eating in adolescents. Adolescent self-control training could also assist in preventing emotional eating. © 2014 Elsevier Ltd. All rights reserved.
Introduction Emotional eating is defined as “eating in response to a range of negative emotions . . . to cope with negative emotions” (Faith, Allison, & Geliebter, 1997, p. 439). Emotional eating in children causes excessive intake of food that is generally rich in fat and sugar, which results in increased body weight (Elfhag & Morey, 2008; Webber, Hill, Saxton, Van Jaarsveld, & Wardle, 2008). In addition, emotional eating is a risk factor for bulimia nervosa (Engelberg, Steiger, Gauvin, & Wonderlich, 2007) and an important predictor of binge eating in preadolescents and adults (Allen, Byrne, La Puma, McLean, & Davis, 2008). Meanwhile, studies related to unhealthy eating and emotional eating have suggested that adolescence is a critical period in the development of unhealthy eating behaviors and eating disorders (Striegel-Moore & Bulik, 2007). Previous studies have found that emotional eating in adolescents is affected by family environment, including parental eating habits as well as parenting and feeding practices (de Lauzon-Guillain, Musher-Eizenman, Leporc, Holub, & Charles, 2009; Faith, Scanlon, Birch, Francis, & Sherry, 2004; Topham et al., 2011) and personal
* Corresponding author. E-mail address:
[email protected] (T. Cai). http://dx.doi.org/10.1016/j.appet.2014.06.106 0195-6663/© 2014 Elsevier Ltd. All rights reserved.
factors, including stress level, stress coping styles, and personality (Elfhag & Morey, 2008; Larsen, van Strien, Eisenga, & Engels, 2006; Wallis & Hetherington, 2004). Few studies have examined possible mediators of the relationship between negative emotions and emotional eating. To our knowledge, one study has investigated the relationship between depression and emotional eating. This study used structural equation modeling and suggested that alexithymia and impulsivity might mediate this relationship (Ouwens, van Strien, & van Leeuwe, 2009). No studies have examined the potential mediators of family environment in this relationship. In addition, previous studies have insufficiently isolated the effects of potential mediators, and thus the relative importance of each mediator remains elusive. In this study, we hypothesized that self-control and parental control might both be potential mediators in the relationship between negative emotions and emotional eating. Each of these potential mediators will be examined simultaneously to determine the most influential pathways.
Self-control and adolescent emotional eating Self-control refers to the ability to override or alter one’s inner responses, and involves consciously interrupting the stream of thought, altering emotions, and restraining undesirable impulses
H. Zhu et al./Appetite 82 (2014) 202–207
and impulsive behaviors (Tangney, Baumeister, & Boone, 2004). Blankstein and Polivy (1982) defined self-control as the ability to affect, regulate, and control one’s emotional behaviors. Elfhag and Morey (2008) found that poor self-control was an important factor for eating due to negative emotions, suggesting that the inhibition of eating and difficulty in controlling eating behavior are crucial components of emotional eating. However, self-control was also affected by negative emotion. Davis and Claridge (1998) indicated that negative emotions might change individuals’ appraisal of dietary maintenance, impair cognitive control of eating, and thereby enable the tendency to overeat. Heatherton and Baumeister (1991) proposed escape theory which asserts that some individuals narrow their attention to the immediate stimulus environment (including emotional distress) to shift attention away from negative emotions. In accordance with escape theory, Ouwens et al. (2009) found that depression was indirectly associated with emotional eating through impulse regulation. Aside from emotional eating, escape theory has been applied to other phenomena such as excessive drinking, smoking, eating, and impulsive destructive behaviors.
Parental control and adolescent emotional eating Parenting also plays an important role in children’s emotions and behavior. Compared to parental warmth, the effect of parental control on child development is relatively unclear (Chen, Liu, & Li, 2000). Parental control includes interfering, demanding compliance, overprotection, and providing less affection (Favaretto, Torresani, & Zimmermann, 2001). Although several researchers have suggested that a certain level of control is beneficial for learning and maintaining appropriate behavior, some studies have indicated that parental high control was also significantly related to unhealthy behaviors in children (e.g., Barber, 1996; Maccoby & Martin, 1983). Topham et al. (2011) suggested that it was important to study the relationship between family factors (e.g., parenting styles) and children’s emotional eating to help us better understand the development of emotional awareness, impulse regulation, and responses to negative emotions in children. However, few studies have examined the role of parenting in children’s emotional eating. Snoek, Engels, Janssens, and van Strien (2007) found that low maternal support, high psychological control, and high behavioral control were associated with increased emotional eating during childhood. They concluded that emotional eating might be a result of inadequate parenting. In addition, negative emotions are significantly related to parental control. For instance, Paulussen-Hoogeboom, Stams, Hermanns, and Peetsma (2007) found that higher negative emotions among children were associated with higher levels of parental restrictive control, characterized by power assertion, negativity, intrusiveness, hostility, over-controlling behavior, and/or over-involvement. Rubin, Stewart, and Chen (1995) also found that children’s negative emotions (e.g., social fearfulness) might elicit restrictive control from their parents. Based on these theories, we constructed a model in which self-control and parental control might both act as potential mediators of the relationship between negative emotion and emotional eating (Fig. 1). In addition, Finkenauer, Engels, and Baumeister (2005) found that parents can foster the development of certain aspects of their children’s character in order to influence their emotions and behavior. For instance, parenting can indirectly influence adolescent emotional and behavioral problems through children’s self-control. Thus, we constructed a second model that added a pathway from parental control to self-control (Fig. 2).
203
Fig. 1. Model 1.
Methods Subjects Participants were 594 tenth- and eleventh-grade students, of whom 258 were boys and 336 were girls. Their average age was 16.70 (SD = 1.09; age range 15–18). Body mass index (BMI) ranged from 16 to 29, with an average of 20.06 (SD = 2.03). Fifty-two participants were overweight/obese, and 542 were normal weight/lean. Socioeconomic status (SES) was determined by monthly family income (RMB) and divided into low (5000) levels. In terms of SES, 39.4% of subjects were low, 44.9% of subjects were medium, 8.9% were high, and 6.8% were missing. Procedures Survey data collectors were trained together on data collection procedures and contending with challenges such as cleaning data and dealing with missing data. Immediately after being informed about the goal of the study, participants completed the Positive and Negative Affect Scale, Emotional Eating Scale, Parental Bonding Instrument (PBI), and the Self-Control Scale. Data on weight, height, and family income were also collected. The total time to complete all questionnaires was about 20–30 min. Surveys were collected immediately after they were completed. This study was approved by the ethical committee of Central South University, and all subjects provided written informed consent. Measures Emotional eating The emotional eating scale (Arnow, Kenardy, & Agras, 1995) has been used to investigate eating behavior in response to negative emotions. We used the Chinese version of the Negative Emotional Eating subscale of the Emotional Eating Scale (Zhu, Cai, Chen, & Zhang, 2013), which contains three factors: eating in response to anxiety
Fig. 2. Model 2.
204
H. Zhu et al./Appetite 82 (2014) 202–207
(four items), depression (nine items), and hostility (five items). Participants were asked to respond to questions about their desire to eat when experiencing certain emotions (e.g., “Do you have a desire to eat when you are sad, irritated, worried, or lonely?”) on a 5-point Likert scale ranging from 1 (never) to 5 (very strong). The total scores for anxiety, depression, and hostility represent total negative emotional eating. Higher scores indicate a stronger desire to eat. In the current study, Cronbach’s α was 0.88 (boys: 0.87, girls: 0.92) for negative emotional eating. Self-control The Self-Control Scale (Tangney et al., 2004) has been used to measure self-control ability. Previous studies have shown that higher self-control scores were associated with fewer unhealthy eating behaviors. In the present study, we used the Chinese version of the Self-Control Scale (Hu, Chen, & Cai, 2012), which contains 19 items (e.g., “I do certain things that are bad for me if they are fun,” “I often act without thinking through all the alternatives”). Items are rated on a 5-point Likert scale, ranging from 1 (not at all) to 5 (very much), with higher scores indicating better self-control. Cronbach’s α for this measure was 0.80 (boys: 0.77, girls: 0.80) in the current study. Negative emotion The Negative Affect dimension of the Positive and Negative Affect Schedule (Watson, Clark, & Tellegen, 1988) has been used to evaluate individuals’ negative affect. The Chinese version (Huang, Yang, & Ji, 2003) was used, which included 10 items. Participants were asked to evaluate the magnitude of their emotions over the past 2 weeks (e.g., “Do you feel worried or afraid over the past 2 weeks?”) on a 5-point Likert scale ranging from 1 (not at all/very mild) to 5 (very strong). Cronbach’s α was 0.70 (boys: 0.70, girls: 0.73) in the current study, indicating good internal consistency. Parental control The Parental Bonding Instrument (PBI; Parker, Tupling, & Brown, 1979) has been used to evaluate children’s beliefs about the rearing style of their parents. The PBI questions are answered separately for mothers and fathers. We used the six-item Parental Control subscale (e.g., “My father/mother tries to control everything I do,” “My father/mother is invading my privacy,” “My father/mother overprotects me”) of the Chinese version (Chu, Zhou, Yang, & Liu, 2009); responses for each item followed a 4-point Likert scale ranging from 0 (not true at all) to 3 (absolutely true). The Chinese version measures overprotective and authoritarian parenting, meaning that parents are overly involved in their children’s life and restrict their activities. The obtained maternal and paternal scores can be used separately. In the current study, Cronbach’s α for paternal and maternal control was 0.72 (boys: 0.70, girls: 0.73) and 0.77 (boys: 0.80, girls: 0.76), respectively. Body mass index BMI was used as a measure of relative weight and was calculated by weight (kg) divided by height (m2). According to the Group of China Obesity Task Force (2004) report, the BMI cut-off points for being overweight are 23.1 and 23.4 for boys and girls, respectively, at age 15; 23.5 and 23.7 for boys and girls, respectively, at age 16; 23.8 for both genders at age 17; and 24 for both genders at age 18. Data analysis Analyses of descriptive statistics as well as t-tests, ANOVAs, and testing relationships between negative emotions, emotional eating, self-control, and parental control were performed using SPSS 17.0 (SPSS Inc., USA).
Table 1 T-tests and ANOVAs for the effects of gender, age, BMI, and SES on emotional eating. Emotional eating
Gender Boys Girls Age 15 16 17 18 BMI Normal weight/lean Overweight/obese Socioeconomic status Low Medium High
t/F
P
11.14 10.47
−2.14*
0.03
41.00 39.67 40.25 39.63
11.32 9.95 11.55 10.85
0.28
0.84
40.09 40.97
10.88 10.26
−0.51
0.61
39.80 40.93 40.46
10.44 11.08 10.93
1.90
0.13
M
SD
38.45 40.36
* P < .05.
Mplus 6.0 (Muthén & Muthén, 2010) was also used for data analysis, which consisted of two steps. The first step evaluated the two models to identify the most suitable model. We examined multiple fit indexes, namely, chi-squared statistics, the goodness-of-fit index (GFI), normed fit index (NFI), root mean square error of approximation (RMSEA), Tucker–Lewis index (TLI), comparative fit index (CFI), Akaike information criterion (AIC), and Bayesian information criterion (BIC). Values of ≥0.9 on the GFI, NFI, and CFI (Hu & Bentler, 1999) and values of ≤0.08 on the RMSEA (McDonald & Ho, 2002) indicate good fit. Smaller AIC and BIC values indicate a better-fit model (Burnham & Anderson, 2002). The second step analyzed the multiple mediating effects of self-control and parental control by using a bootstrapping approach in both male and female samples, and compared the mediation effects. Direct effects represent the influence of one variable on another, unmediated by any other variable. In contrast, indirect effects represent the effect of one variable on another, with the relationship mediated by at least one other variable (Bollen & Stine, 1990). Results Descriptive statistics There was a significant effect of gender on emotional eating, such that girls were more likely to experience emotional eating; however, there were no significant differences by age, BMI, or SES (Table 1). Descriptive statistics (means and standard deviations) for negative emotions, emotional eating, self-control and parental control, and the Pearson correlation coefficients between these variables are shown in Table 2. Negative emotions were positively correlated with emotional eating (r = 0.21) and parental control (rs = 0.20 to 0.26);
Table 2 Pearson correlations between negative emotions, emotional eating, self-control, and parental control.
1. Negative emotions 2. Emotional eating 3. Maternal control 4. Paternal control 5. Self-control Mean SD ** P < .01.
1-NE
2-EE
3-MC
4-PC
5-SC
.21** .20** .26** −.36** 20.42 5.01
.12** .17** −.36** 40.16 10.83
.62** −.24** 5.59 3.34
−.21** 4.39 2.69
67.44 9.93
H. Zhu et al./Appetite 82 (2014) 202–207
Table 3 Goodness-of-fit of the two models.
Model 1 Model 2
χ2
df
GFI
CFI
NFI
TLI
RMSEA
AIC
BIC
9.06 5.68
3 2
0.98 0.99
0.97 0.99
0.96 0.98
0.88 0.96
0.09 0.06
44.47 31.68
97.12 88.71
Notes: GFI, NFI, TLI and CFI ≥ 0.9, RMSEA ≤ 0.08. Smaller AIC and BIC values indicate a better model.
self-control was negatively correlated with all of the variables examined (rs = −0.21 to −0.36). Testing the two models In model 2, RMSEA, AIC, and BIC were smaller than in model 1, indicating that model 2 was a better fit than was model 1 (Table 3). Thus, the mediation effects within model 2 will be examined. Testing for path coefficients As shown in Table 4, negative emotions predicted differences in parental control and self-control among the male and female samples. Self-control also predicted differences in emotional eating. Parental control predicted differences in emotional eating in the female sample, but not in the male sample. In addition, parental control significantly predicted differences in self-control. Testing for direct and indirect effects As shown in Table 5, except for the mediation of parental control in boys, the mediating effects of parental control and self-control on the relationship between negative emotion and emotional eating were significant. Specifically, for the male sample, the sum of indirect effects [including the pathway through self-control alone (NE→SC→EE), parental control alone (NE→PAC→EE) and the pathway through parental control and self-control (NE→PAC→SC→EE)] was 0.133 (SE = 0.02, 95% CI: 0.094–0.173), and the direct effect was 0.12 (SE = 0.01, 95% CI: 0.101–0.140). Thus, the indirect effects accounted for 52.6% of the total effect. For the female sample, the sum of the indirect effects was 0.201 (SE = 0.02, 95% CI: 0.162–0.240), and
Table 4 Standardized path parameter estimates of the multiple mediator models. Path
Boys (n = 258) EST
Direct effect: negative emotion predicting emotional eating NE→EE Mediational path: negative emotion predicting mediators NE→PAC NE→SC Mediational path: mediators predicting emotional eating PAC→EE SC→EE Mediational path: mediator predicting another mediator PAC→SC
SE
Girls (n = 336) t
EST
SE
t
0.12
0.01
3.85*
0.10
0.01
3.37*
0.29 −0.27
0.04 0.11
3.45* −3.92*
0.27 −0.35
0.04 0.11
4.25* −6.70*
0.04 −0.36
0.01 0.04
1.14 −5.58*
0.21 −0.36
0.02 0.02
3.65* −5.48*
−0.23
0.28
−3.30*
−0.16
0.25
−2.66*
Notes: NE = negative emotions, EE = emotional eating, PAC = parental control, SC = self-control, EST = estimate value, SE = standard error. * P < .05.
205
Table 5 The indirect effects of self-control and parental control on the relationship between negative emotions and emotional eating. Indirect effect
Boys (n = 258)
Girls (n = 336)
EST
SE
P
EST
SE
P
NE→PAC→EE NE→SC→EE NE→PAC→SC→EE
0.011 0.097 0.025
0.00 0.02 0.00
>0.05