Relationships Between Learned Self-Regulatory Skills and Changes in Weight-Loss Behaviors: Mediation by Barriers Self-Efficacy

Abstract

This study aimed to assess changes in learned self-regulatory skills and barriers self-efficacy associated with theory-based behavioral weight-loss treatments with different curricular emphases, and to evaluate mechanisms of short- and long-term changes in exercise and eating behaviors via self-regulatory skills usage to inform theory and improve lagging intervention effects. Women with obesity volunteered and were randomly allocated into 1-year community-based behavioral weight-loss treatments with either a high (n = 37) or moderate (n = 41) focus on building self-regulatory skills to overcome lifestyle barriers. They were periodically assessed on measures of exercise- and eating-related self-regulatory skill usage, exercise- and eating-related barriers self-efficacy, exercise outputs, fruit/vegetable intake, and body composition. Reductions in weight and waist circumference, increases in exercise- and eating-related self-regulation and barriers self-efficacy, and increases in exercise and fruit/vegetable intake were each significant overall, and significantly greater in the group with a high self-regulatory skills-building focus. Change in barriers self-efficacy significantly mediated relationships between self-regulation change and changes in exercise outputs and fruit/vegetable intake from both baseline–Month 6 and baseline–Month 12. Change in barriers self-efficacy also significantly mediated relationships between change in self-regulation for eating from baseline–Month 3 and long-term changes in self-regulation for eating. Findings supported tenets of social cognitive, self-efficacy, and self-regulation theories: and indicated the importance of emphasizing, and facilitating a high amount of instruction and rehearsal time for self-regulatory skills development within behavioral weight-loss treatments. The ability to nurture self-efficacy through overcoming lifestyle barriers was also indicated.

Keywords

self-regulation barriers self-efficacy exercise weight loss

In the United States, 42% of adults have obesity (body mass index [BMI] ≥ 30 kg/m²; Hales et al., 2020). At any given time, six in 10 women in that subgroup report trying to control their weight through behavioral changes, which is a greater proportion than their male counterparts (Kruger et al., 2004). Even after decades of proposed refinements, behavioral (nonsurgical/non-pharmacological) treatments for obesity have, with few exceptions, had only short-term effects on weight (i.e., 6–9 months after treatment start; Cooper et al., 2010; Dombrowski et al., 2014; Loveman et al., 2011; Mann et al., 2007). After 6 to 9 months, a gradual but persistent regain in weight nearly always ensues that reverses any progress made (Dombrowski et al., 2014; Jeffery et al., 2000; Loveman et al., 2011). Typically, weight loss is less than the 5% thought to be required to improve participants’ health risks (Williamson et al., 2015). Although increased exercise and fruit/vegetable intake reliably predicts weight loss (Aljadani et al., 2013; Bertoia et al., 2015; Dunn et al., 2006), adherence beyond the short term is rare (Dombrowski et al., 2014; Jeffery et al., 2000; Loveman et al., 2011; Mann et al., 2007).

In consideration of decades of treatment failures at maintaining meaningful amounts of behavioral changes and weight loss, the National Institutes of Health recently convened a multidisciplinary working group to suggest future research directions (MacLean et al., 2015). They advised that research on “innovations in behavioral strategies” investigate the establishment of behavioral skills related to the maintenance of weight loss early in the treatment process (MacLean et al., 2015, p. 11). It was also recommended that considerable attention be placed on increasing adherence to exercise. Other researchers even more emphatically suggested that progress in behavioral weight-loss interventions requires an enhanced understanding of theory-centered psychosocial processes associated with fostering and sustaining changes in exercise and eating behaviors (Baranowski et al., 2009; Rothman et al., 2004). However, comprehension of the interaction of psychosocial correlates of weight-loss behaviors remains deficient, especially in consideration of the dynamic changes that occur during treatment processes (Cooper et al., 2010; Dombrowski et al., 2014).

Variables such as self-regulation and self-efficacy have been incorporated into models predicting weight-loss behaviors and weight (Anderson et al., 2007; Annesi, 2020; Linde et al., 2006). A systematic review of 35 studies of adults with overweight and obesity found self-regulation and self-efficacy to be consistent predictors of physical activity/exercise, diet, and weight over both the short and long term (Teixeira et al., 2015). However, much of that research was cross-sectional and of brief durations, failing to account sufficiently for interrelations of changes in variables, long-term progress, and directionality of effects. This limited the findings’ applicability for informing treatments that seek sustained improvements in weight-loss behaviors. Also, the authors of the review pointed to a need for stronger methodological quality in future related investigations, especially in the prediction of dietary change where confidence in findings was limited by few available studies (Teixeira et al., 2015). In an untested model that emerged from a review of previous research, it was suggested that self-regulation and self-efficacy impacts weight loss through an interaction between increased exercise and dietary compliance (Baker & Brownell, 2000).

The psychosocial factors of self-regulation and self-efficacy are consistent with established theories of behavior change such as self-regulation theory (Vohs & Baumeister, 2016), self-efficacy theory (Bandura, 1997), and social cognitive theory (Bandura, 1986). For example, self-regulation theory posits one’s self-regulatory skills usage will directly counter barriers to behavioral changes such as exercise-induced discomforts, cravings for unhealthy foods, and slow progress. Alternatively, self-efficacy theory emphasizes that feelings of ability to complete a task while overcoming difficulties (i.e., barriers self-efficacy; Bandura, 1997) strongly influences behavioral changes. Social cognitive theory suggests an interaction of self-regulation and self-efficacy whereby increased utilization of self-regulatory skills will directly influence behavioral changes, while also operating on those changes through improvements in barriers self-efficacy (possibly by demonstrating to one’s self they have sufficient skills to overcome challenges and improve goal-driven behaviors; Bandura, 1986). Theory has been less clear in the prediction of how variables such as self-regulation will diminish or strengthen over time. This is important in the weight-loss arena in which initial behavioral improvements do not typically predict sustained changes or consistent progress in weight. Few studies have investigated the conditions in which self-regulatory skills are maintained, increased, or depleted over time in applied (non-laboratory) settings (Vohs & Baumeister, 2016). Feelings of ability (i.e., self-efficacy) might affect one’s progress at utilizing newly learned self-regulatory skills (Bandura, 2005).

Use of theory-driven predictive models—applicable to behavioral weight-loss treatment development—might have the best chance of finally facilitating reliable improvements in both short- and long-term effects (Gillison et al., 2015). Treatment research has rarely decomposed their primarily outcome-focused findings to determine mechanisms of effects that are critically needed for advancing both theory and interventions. However, in extending the theoretical paradigm by Baker and Brownell (2000) on the effects of exercise on nutrition change and sustained weight loss, a series of field-based studies led to a proposal of a causal chain model (Spencer et al., 2005) intended to inform treatment processes over both the short and long term (Annesi, 2020). Using a translational behavioral medicine perspective, that research suggested improvements in self-regulatory skills used to overcome lifestyle barriers explained much of the variance in weight-loss behavior changes. It also posited that rehearsed self-regulatory skills would sustain or increase over time, rather than deplete (Baumeister & Vohs, 2016). In addition, the model proposed that changes in self-efficacy mediate those effects (Annesi, 2020).

Although preliminary suggestions for advancing self-regulatory skills development and self-efficacy emerged (Annesi, 2020), the research also had gaps requiring attention. For example, it was not known what proportion of treatment time dedicated to self-regulatory skills development would maximize effects in that area and in behavioral changes. Although a high emphasis might seem most beneficial, participants could also have an adverse reaction to a myopic treatment focus that does not equally instruct them in (possibly expected and sought) nuances of exercise training and nutrition. In addition in the formative research, measurement of self-regulation often failed to directly assess effects on the skills trained within the associated treatments, and the measurement of self-efficacy often did not assess barriers self-efficacy specifically. Moreover, when analyses assessed longitudinal effects, they often assumed directionality in associated relationships rather than investigate the prediction of changes over a longer period via shorter-term changes, as had been suggested (Cromwell et al., 1994). Through such, an increased confidence in the directionality of relationships between short- and long-term psychosocial changes and associated changes in weight-loss behaviors could better guide behavioral treatment processes.

Thus, the present investigation aimed to address key issues that could further inform researchers and practitioners toward an intervention model applicable for large-scale dissemination. To maximize external validity, the study was conducted in community wellness facilities where women with obesity enrolled in treatments having either a high or moderate emphasis on self-regulatory skills development. The following hypotheses were given:

Hypothesis 1 (H1): Both tested groups will demonstrate significant improvements in exercise- and eating-associated self-regulation and barriers self-efficacy, measures of exercise, fruit/vegetable intake, and body composition; however, improvements will be significantly greater in the group with the high self-regulatory skills development focus.

Hypothesis 2 (H2): Changes in barriers self-efficacy will significantly mediate the prediction of exercise and fruit/vegetable intake by self-regulatory skills usage changes.

Hypothesis 3 (H3): For both exercise- and eating-change contexts, changes in barriers self-efficacy will significantly mediate the prediction of long-term changes in self-regulatory skills usage by short-term changes.

Method

Participants

Participants volunteered for a weight-management program that included methods to increase exercise and healthy eating to be delivered in community health-promotion centers in the Eastern United States. Inclusion criteria were as follows: women of age ≥ 21 years; obese, but no known contraindication for safe participation; and a reported average of ≤3 days per week of moderate exercise and the consumption of <5 fruits and vegetables (combined) per day. There was random assignment to the two treatment groups and minimal attrition prior to study start associated with illness (n = 2), transportation problems (n = 4), and reported inclusion criteria not verified by study staff (n = 2), which did not significantly differ by group. The high self-regulation group (n = 37) and moderate self-regulation group (n = 41) had no significant difference in age (M_overall = 47.9 years, SD = 8.2 years, range = 21–60 years), BMI (M_overall = 35.0 kg/m², SD = 3.1 kg/m², range = 30.0–40.7 kg/m²), education (overall 58% bachelor’s degree or above), and racial/ethnic make-up (overall 79% White, 12% Black, and 9% of other groups combined). Nearly all participants (89%) had a middle family income of between US$50,000 and US$100,000 per year. The Kennesaw State University Institutional Review Board (IRB) approved the study protocol, and IRB-approved written consent was obtained from all participants prior to entry into the study.

Measures

Self-regulation

Using the taxonomy provided by Michie et al. (2011) as a guide for the articulation of self-regulatory skills, previously validated scales of 10 items each (Annesi & Marti, 2011; Saelens et al., 2000) were reduced to eight items each to measure participants’ use of the specific skills (e.g., self-talk/cognitive restructuring, relapse prevention) incorporated within the present treatments’ curricula. The two scales separately focused on self-regulation related to exercise and eating. Sample items (and the associated skill based on the Michie et al.’s, 2011, taxonomy) for the self-regulation for exercise measure were as follows: “When I get off-track with my exercise plans, I work to quickly get back to my routine” (relapse prevention) and “I set exercise goals” (goal setting related to behavior change). Sample items for the self-regulation for eating measure were as follows: “I make formal agreements with myself regarding my eating” (behavioral contracting) and “I try to recruit others to support my eating plans” (recruiting social supports). Response options ranged from 1 (never) to 4 (often). They were summed for a possible score range of 8 to 32, for each measure. In adults with obesity, internal consistencies were Cronbach’s alpha = .81 and .79, and test–retest reliabilities over 2 weeks were .74 and .78, respectively (Annesi & Marti, 2011). For the present sample, Cronbach’s alpha = .82 and .85, respectively.

Barriers self-efficacy

The measurement of barriers self-efficacy (also termed self-regulatory efficacy), or the confidence one has to use their internal resources to overcome difficulties (i.e., barriers) to successfully completing a behavior (Bandura, 1997), consisted of two scales with separate exercise and eating foci. Barriers self-efficacy for exercise was measured by the five items of the Exercise Self-Efficacy Scale (Marcus et al., 1992). Sample items were as follows: “How confident are you that you can persist with exercising when you feel you do not have the time?” and “How confident are you that you can persist with exercising when you are tired?” Response options ranged from 1 (not at all confident) to 11 (very confident). They were summed for a possible score range of 5 to 55. Internal consistencies were Cronbach’s alpha = .76 to .82, and test–retest reliabilities over 2 weeks were .74 to .78 (Marcus et al., 1992). For the present sample, Cronbach’s alpha = .84.

Barriers self-efficacy for eating was measured by the 20 items of the Weight Efficacy Lifestyle Scale (Clark et al., 1991). Four items each addressed challenges to healthy eating related to categories of social pressures (e.g., “I can resist eating even when others are pressuring me to eat”), positive activities (e.g., “I can resist eating when I am watching TV”), physical discomforts (e.g., “I can resist eating when I feel physically run down [very tired]”), high food availabilities (e.g., “I can resist eating when there are many different kinds of foods available”), and negative emotions (e.g., “I can resist eating when I am depressed [or feeling down]”). Response options ranged from 0 (not confident) to 9 (very confident). The mean score from each category was calculated, then summed for a possible score range of 0 to 45. Internal consistencies within categories were Cronbach’s alpha = .76 to .82 (Clark et al., 1991), and they were Cronbach’s alpha = .78 to .86 for the present sample.

Weight-loss behaviors

Exercise outputs were measured by the Leisure-Time Physical Activity Questionnaire (Godin, 2011). The metric of metabolic equivalents (METs; 1 MET = 3.5 mL of O₂/kg/min) per week were represented by number of exercise bouts of at least 15 min, with intensities ranging from 3 METs (e.g., easy walking) to 9 METs (e.g., running), which were summed. For example, three bouts of easy walking would yield a Leisure-Time Physical Activity Questionnaire MET score of 9. Concurrent validity was indicated through strong correspondences with exercise stress test (β = .57) and accelerometer (β = .45) scores (Jacobs et al., 1993; Pereira et al., 1997). Test–retest reliability over 2 weeks was .74 (Pereira et al., 1997).

The measure of fruit/vegetable intake was based on the recalled daily consumption of fruits (e.g., pear, orange [one small, or 118 mL]) and vegetables (not fried; for example, carrots, peas [118 mL]) during the previous week. Portion sizes on the measure corresponded to data from the U.S. Department of Health and Human Services and U.S. Department of Agriculture (2015) and U.S. Department of Agriculture (2017). In correspondence with suggestions for validating measures of diet (National Institutes of Health/National Cancer Institute, 2020), correlations of scores with comprehensive food frequency recall instruments and the Block Food Frequency Questionnaire were strong at βs = .70 to .85, and test–retest reliabilities over 3 weeks were .77 to .83 in women with obesity (Block et al., 1986; Mares-Perlman et al., 1993). Scores on the present measure were also inversely associated with the intake of other food groups (i.e., sweets, bread, dairy) and weight (Annesi, 2018).

Body composition

Weight was measured in kilograms using a recently calibrated digital scale. Footwear and heavy outer-clothing were first removed. Waist circumference was measured in centimeters by a tape measure at the umbilicus, at the end of a normal expiration of breath. For both, the mean of two consecutive measurements was recorded.

Procedure

Treatment instructors were existing staff members of community wellness centers. Each possessed a national certification related to health education and/or health promotion. Both the high self-regulation and moderate self-regulation groups’ curricula consisted of similar materials, but were administered in different proportions of self-regulatory skills development and educational content (i.e., 90% self-regulation skills development/10% education and 50% self-regulation skills development/50% education, respectively). The overall treatments began by establishing participants’ exercise programs. Although the governmental recommendation of at least 150 min/week of moderate-intensity aerobic activity (Piercy et al., 2018) was discussed, it was also indicated that any increase from the participants’ initial amounts of physical activity/exercise could be beneficial. The self-regulatory skills focused upon within the exercise support portion of the curricula were as follows: goal setting related to behavior change, self-monitoring of behaviors, stimulus control, productive self-talk/cognitive restructuring, relapse prevention, recruiting social supports, contingency-based self-reward, and behavioral contracting. With their bases in social cognitive (Bandura, 1986, 2005), self-efficacy (Bandura, 1997), and self-regulation (Vohs & Baumeister, 2016) theories, each skill was described by Abraham and Michie (2008) as being within the “social cognitive theory” and “control” treatment models. The focus of the high self-regulation group was primarily limited to rehearsing the covered self-regulatory methods to increase exercises chosen by participants. In the moderate self-regulation group, more detailed education on exercise processes was also provided that included topics such as warming up and cooling down, cardiorespiratory benefits of exercise, and exercise and caloric expenditures. For both groups, the exercise support component consisted of a combination of monthly one-on-one meetings, phone contacts, and reviews of written materials that overall totaled approximately 3 hr over 6 months.

For both groups, the nutrition/eating behavior-change component began 2 months after the start of the exercise support component. The same self-regulatory skills addressed in each groups’ exercise support component were addressed in the same proportions within this component. The nutritional focus of the high self-regulation group was primarily limited to increasing the intake of fruits and vegetables and reducing the consumption of sweets, while more detailed nutrition education was provided to the moderate self-regulation group. This included topics such as evaluating your diet, fast foods, and analyzing food labels. Consistent with the exercise support component, the education time in the moderate self-regulation group was replaced with review and rehearsals of covered self-regulatory methods in the high self-regulation group. Across groups, there were also recommendations for self-weighing at least once per week and restricting caloric intake to 1,200 to 1,500 kcal/day, based on present weight. For both groups, the eating behavior-change component consisted of a combination of group meetings and phone contacts every 2 weeks, and reviews of written materials that overall totaled approximately 13 hr over 10 months.

Structured fidelity checks by study staff on approximately 10% of treatment sessions suggested sound protocol compliance. Minor problems were handled without difficulty through consultations between study staff and treatment instructors. Study staff administered measures in a private area at the planned intervals over 1 year.

Data Analyses

The 10% of missing cases met the criteria for having no systematic bias (i.e., missing-at-random; White et al., 2011). Therefore, the expectation maximization algorithm (Schafer & Graham, 2002) was used for imputation to enable the advantages of an intention-to-treat study format (Gupta, 2011). For the primary analysis, 76 total participants were required to detect a moderate effect size of f² = .15 at the statistical power of .80 (α < .05; Cohen et al., 2003). Tolerance levels of .46 to .98 indicated acceptable levels of multicollinearity in the data (Pituch & Stevens, 2016).

Mixed-model repeated measures analyses of variance (ANOVAs) first assessed overall changes in the psychosocial and behavioral measures planned for entry into regression models across their measured times between baseline and Month 12, and their time × group interactions. Although it was not a focus of the study, associations of the different treatment formats with changes in measures of body composition were also included due to the weight-loss orientation of the interventions. Corresponding effect sizes were calculated using partial eta-squared ( $η_{p}^{2} = {SS}_{Effect} / {[SS}_{Effect} + {SS}_{Error}]$ ), where .01, .09, and .25 were small, moderate, and large effects, respectively.

Using data aggregated across groups, and consistent with previous theory and research (Allison, 1990; Annesi, 2020; Fitzmaurice et al., 2004), a series of mediation analyses were then fit for prediction of the planned dependent variables of changes in exercise output, fruit/vegetable intake, and self-regulatory skills usage over 6 and 12 months (see Figures 1 and 2). Based on suggested principles using bootstrapped resampling of the data (Hayes, 2015, 2018; Rucker et al., 2011), a significant relationship between the independent and dependent variables was not first required. To enhance confidence in the directionality of relationships, principles of lagged variable analyses (e.g., score changes over a longer temporal period predicted by changes over an earlier period; Cromwell et al., 1994) were employed. SPSS Statistics Version 22 was used, incorporating the Process macro-instructional software Model 4 with 50,000 bias-corrected bootstrapped resamples (Hayes, 2015). Baseline scores of the dependent variables were controlled. Significance of mediation was assessed via a 95% confidence interval (95% CI).

Figure 1.

Mediation of the prediction of weight loss behavior changes by changes in self-regulatory skills usage through change in barriers self-efficacy.

Figure 2.

Mediation of the prediction of long-term changes in self-regulatory skills usage by changes in short-term self-regulation through change in barriers self-efficacy.

Results

There was no significant between-group difference on any study variable at baseline, F(1, 76) = 0.002 to 2.45, ps = .963 to .122. Descriptive data at each measurement time are given in Table 1.

Table 1.

Descriptive Statistics of Study Variables at All Measured Times During the Study Year.

	Baseline		Month 3		Month 6		Month 12
Measure	M	SD	M	SD	M	SD	M	SD
Self-regulation for exercise
High self-regulation focus group	16.95	4.99	25.34	3.41	25.99	3.39	25.41	3.50
Moderate self-regulation focus group	18.40	5.25	21.03	5.34	21.46	5.77	21.25	5.83
Aggregated across groups	17.71	5.15	23.08	5.00	23.61	5.27	23.22	5.27
Self-regulation for eating
High self-regulation focus group	18.14	4.45	24.79	3.73	25.86	3.45	25.66	3.14
Moderate self-regulation focus group	17.31	4.86	21.00	4.72	21.44	4.80	21.01	4.43
Aggregated across groups	17.70	4.66	22.79	4.66	23.54	4.74	23.22	4.50
Barriers self-efficacy for exercise
High self-regulation focus group	23.59	8.38			34.38	11.13	32.49	11.65
Moderate self-regulation focus group	21.71	10.57			25.80	12.75	26.61	11.14
Aggregated across groups	22.60	9.58			29.87	12.69	29.40	11.69
Barriers self-efficacy for eating
High self-regulation focus group	21.05	7.27			32.14	7.89	31.65	7.48
Moderate self-regulation focus group	21.13	7.95			25.88	7.86	26.81	7.47
Aggregated across groups	21.10	7.59			28.85	8.43	29.11	7.81
Exercise output (METs/week)
High self-regulation focus group	7.73	7.12			32.41	15.43	30.79	15.28
Moderate self-regulation focus group	8.95	8.31			21.80	14.24	22.73	12.78
Aggregated across groups	8.37	7.74			26.83	15.65	26.55	14.50
Fruit/vegetable intake (portions/week)
High self-regulation focus group	2.96	1.08			5.92	2.17	5.93	2.29
Moderate self-regulation focus group	2.63	1.37			3.91	2.16	4.20	1.94
Aggregated across groups	2.79	1.24			4.87	2.38	5.02	2.27
Weight (kg)
High self-regulation focus group	94.45	10.95			88.87	11.92	89.41	12.26
Moderate self-regulation focus group	93.21	9.97			91.15	10.00	91.34	10.97
Aggregated across groups	93.80	10.40			90.07	10.94	90.43	11.56
Waist circumference (cm)
High self-regulation focus group	106.96	9.25			100.36	10.74	101.96	11.56
Moderate self-regulation focus group	103.99	7.72			102.73	6.76	102.21	7.24
Aggregated across groups	105.39	8.56			101.60	8.89	102.09	9.47

Note. High self-regulation focus group n = 37. Moderate self-regulation focus group n = 41. Aggregated across groups N = 78. METs = metabolic equivalents.

Changes in Body Composition

There were significant overall reductions in weight and waist circumference, F(1, 76) = 35.21 and 27.16, respectively, ps < .001 ( $η_{p}^{2}$ = .32 and .26, respectively), that were each significantly greater in the high self-regulation group, F(1, 76) = 7.45, p = .008, and F(1, 76) = 6.11, p = .016 ( $η_{p}^{2}$ = .09 and .07, respectively). In each case, changes from baseline–Month 6 and baseline–Month 12 were significant, where changes from Months 6 to 12 were not. For the high and moderate self-regulation groups, those changes in weight, transposed into percentages, were −5.9%, −5.3%, +0.1%, and −2.2%, −2.0%, +0.2%, respectively.

Changes in Psychosocial Measures

There were significant overall increases in self-regulation for exercise and self-regulation for eating, F(1, 76) = 87.60 and 124.44, respectively, ps < .001 ( $η_{p}^{2}$ = .54 and .62, respectively), that were each significantly greater in the high self-regulation group, F(1, 76) = 20.77 and 14.63, respectively, ps < .001 ( $η_{p}^{2}$ = .22 and .16, respectively). In each case, changes from baseline–Months 3, 6, and 12 were significant, where changes during all other temporal intervals were not.

There were significant overall increases in barriers self-efficacy for exercise and barriers self-efficacy for eating from baseline–Month 6, F(1, 76) = 28.29 and 67.10, respectively, ps < .001 ( $η_{p}^{2}$ = .27 and .47, respectively), that were each significantly greater in the high self-regulation group, F(1, 76) = 5.71, p = .019, and F(1, 76) = 10.74, p = .002 ( $η_{p}^{2}$ = .07 and .12, respectively).

Changes in Weight-Loss Behaviors

There were significant overall increases in exercise and fruit/vegetable intake, F(1, 76) = 146.44 and 91.64, respectively, ps < .001 ( $η_{p}^{2}$ = .66 and .55, respectively), that were each significantly greater in the high self-regulation group, F(1, 76) = 9.29, p = .003, and F(1, 76) = 8.89, p = .004 ( $η_{p}^{2}$ = .11 and .10, respectively). In each case, changes from baseline–Months 6 and 12 were significant, where changes from Months 6 to 12 were not.

Prediction of Weight-Loss Behavior Changes

For the prediction of changes in exercise from both baseline–Month 6 and baseline–Month 12, from 3-month change in self-regulation for exercise, 6-month change in barriers self-efficacy was a significant mediator, B = 0.54, SE_B = 0.20, 95% CI = [0.21, 1.00], and B = 0.63, SE_B = 0.21, 95% CI = [0.29, 1.13], respectively (Figure 1A and 1B). The overall model R²s were significant at .41, F(3, 74) = 16.82, and .35, F(3, 74) = 16.82, respectively, ps < .001.

For the prediction of changes in fruit/vegetable intake from both baseline–Month 6 and baseline–Month 12, from 3-month change in self-regulation for eating, 6-month change in barriers self-efficacy was a significant mediator, B = 0.07, SE_B = 0.04, 95% CI = [0.02, 0.16], and B = 0.08, SE_B = 0.04, 95% CI = [0.02, 0.18], respectively (Figure 1C and 1D). The overall model R²s were significant at .34, F(3, 74) = 12.59, and .28, F(3, 74) = 9.47, respectively, ps < .001.

Prediction of Self-Regulation Over the Long Term

For the prediction of changes in self-regulation for exercise from both baseline–Month 6 and baseline–Month 12, from 3-month change in self-regulation, 6-month change in barriers self-efficacy was not a significant mediator, B = 0.05, SE_B = 0.05, 95% CI = [−0.05, 0.15], and B = 0.03, SE_B = 0.05, 95% CI = [−0.08, 0.14], respectively (Figure 2A and 2B). The overall model R²s were, however, significant at .72, F(3, 74) = 64.55, and .60, F(3, 74) = 36.35, respectively, ps < .001.

For the prediction of changes in self-regulation for eating from both baseline–Month 6 and baseline–Month 12, from 3-month change in self-regulation, 6-month change in barriers self-efficacy was a significant mediator, B = 0.22, SE_B = 0.05, 95% CI = [0.13, 0.34], and B = 0.12, SE_B = 0.05, 95% CI = [0.03, 0.23], respectively (Figure 2C and 2D). The overall model R²s were significant at .67, F(3, 74) = 50.50, and .70, F(3, 74) = 57.59, respectively, ps < .001.

Discussion

Findings clarified associations of differing degrees of self-regulatory skills training with changes in measures of body composition, weight-loss behaviors, and their posited psychosocial correlates in women with obesity. They also elucidated relationships between changes in self-regulatory skills usage, barriers self-efficacy, and exercise and eating behaviors over 6 and 12 months. Implications for researchers and practitioners in the use of theory for the improvement of behavioral obesity interventions were derived. Although groups of both high and moderate self-regulatory skills foci demonstrated the expected significant progress in the tested measures, it is likely that the significantly greater improvements in the high self-regulatory skills group was attributable to the additional curricular time spent in that area (i.e., 90% of treatment time dedicated to self-regulatory skills development vs. 50%). Although inconsistent with the vast majority of treatments that primarily attend to educating participants on the benefits of specific changes in the diet (e.g., low-carbohydrate, vegan) and detailed exercise plans (e.g., high intensity interval training, cross training) to induce behavior change (Mann et al., 2007; Swift et al., 2014), the present results not only supported the benefits of addressing the development of specific self-regulatory skills to counter common barriers to change, but suggest advantages of a nearly singular focus there. As proposed elsewhere (Annesi, 2018, 2020), the present findings suggest that education can be streamlined and targeted (e.g., increase fruits/vegetables, increase any physical activity that elevates the heart rate) so that enabling the self-regulation of behaviors are the clear focus of treatment. In many programs, self-regulation (often recognized via amorphous terms such as “will-power,” “motivation,” or “volition”), while acknowledged, is given only cursory, often atheoretical, attention. Thus, participants are provided minimal opportunities for the development and rehearsal of valuable self-regulatory skills, which necessitate sufficient cultivation (Abraham & Michie, 2008; Michie et al., 2011).

However, further research will be required to evaluate the most productive proportion of diet/exercise information and behavioral skills training, and whether that is affected by variables such as educational level, age, outcome expectations, and/or psychosocial variables that are not yet known. Research on such individualization and targeting of strategies was also suggested in the National Institutes of Health’s working group report on maintenance of weight loss (MacLean et al., 2015). Because of the great amount of information offered to the public by the popular media and commercial weight-loss programs on diet, exercise, and weight loss, it will be important to directly evaluate the acceptability of a high self-regulatory skills focus (i.e., being substantially different from what might be expected) on participants’ program adherence over the long term, which tends to be problematic (Alhassan et al., 2008). Because obesity is a chronic disease, such tolerability is needed because long-term follow-up is generally considered by behavioral researchers to be necessary (Hall & Kahan, 2018).

Findings also indicated that relationships between changes in self-regulatory skills usage and exercise output, and between self-regulation and fruit/vegetable intake, were significantly mediated by barriers self-efficacy. This supported H2 and was consistent with tenets of self-regulation theory (Vohs & Baumeister, 2016), self-efficacy theory (Bandura, 1997), and social cognitive theory (Bandura, 1986). Those findings were also in agreement with much of the previous research that did not account for directionality in those relationships as adequately as the current study (Teixeira et al., 2015). Thus, the present findings add to the extant research on the suggested decomposition of the effects of behavioral treatments on weight-loss behavior changes through psychosocial pathways (Gillison et al., 2015). Although still requiring direct testing, it is likely that as participants overcame challenges that had been problematic to them (e.g., discouragement from slower-than-desired progress, a “slip” in an exercise or eating plan associated with illness or work travel) through their increased self-regulatory skill usage, barriers self-efficacy increased.

While the relationship between improvements in self-regulation and exercise and fruit/vegetable intake over 6 and 12 months were found to be through increased self-efficacy, it is notable that the relationship between changes in self-regulation and both weight-loss behaviors were not significant over 12-month changes (see Path c′, Figure 1B and 1D). This could indicate that the effect that increased self-regulation has on weight-loss behaviors over the longer term (i.e., beyond the initial 6–9 months where weight loss is expected) is particularly dependent on its association with barriers self-efficacy over that temporal period (i.e., the belief that one can overcome lifestyle barriers over the longer term).

Although it was not a focus within the present treatment curricula, it could be valuable to ascertain relationships among changes in self-regulation, self-efficacy, and weight-loss behavior changes if specific methods are included to directly increase barriers self-efficacy. Within self-efficacy theory, Bandura (1997) suggests multiple ways in which self-efficacy can be nurtured, including the demonstrated abilities to succeed, as suggested above (i.e., “mastery experiences”; considered the most important method). Thus, in future treatment iterations, increasing (perceived) abilities to persevere through challenges could be highlighted to participants through carefully logging and highlighting even minimal progress in their proximal goals.

Longitudinal changes in weight-loss behaviors and their psychosocial correlates are high priorities due to expected lapses that reliably induce weight regain (Dombrowski et al., 2014; Jeffery et al., 2000; Loveman et al., 2011). Thus, the carry-over of short-term improvements in self-regulatory skills usage to the long-term increases addressed here is important. Self-regulation theory is unclear under what conditions depletion versus strengthening of such skills occurs (Vohs & Baumeister, 2016). The present findings indicated that early improvements in self-regulatory skills usage significantly predicted improvements over 6 and 12 months in both exercise and eating contexts. In partial support of the corresponding hypotheses, this relationship was significantly mediated by barriers self-efficacy change for eating. A careful inspection of the findings suggests that such mediation was not present in the exercise context due to the nonsignificant associations between increased barriers self-efficacy and increased exercise outputs (see Path b, Figure 2A and 2B). Extensions of this research will be required to determine if, for example, self-regulatory skills required to overcome the effort needed to initiate regular exercise are more directly transferred to long-term adherence than when decisions to eat in a healthy manner are also dependent on one’s feelings of ability to overcome associated barriers for such over the long term (which might be more situational and diverse than with exercise). Although self-regulatory skills are often treated as generic, they might have specific, yet unknown, properties when behaviors with different reinforcement/punishment contingencies are present. Thus, as the present line of behavioral weight-loss research advances, it is likely that theoretical adaptations will be required to address exercise and eating behaviors, as they have been suggested for behavioral changes associated with weight loss versus weight-loss maintenance (Annesi, 2020). Further attention to the optimal conditions for self-regulatory skills to generalize from exercise to eating changes (or vice versa) is also warranted.

Although the present findings suggested practical and theoretical treatment advances that will guide both researchers and direct practitioners, as well as address gaps in related research that should be focused upon in the future, there were also limitations that should be expressed. For example, the specific characteristics of the volunteer sample evoke a need for replications across other sample types (e.g., men, emerging adults, by race/ethnicity, among individuals with severe obesity and/or diabetes). Also, there was a reliance on self-report measures, which presented possible challenges related to participants’ recall, and responses possibly biased by social desirability and expectation effects (Rosenthal & Rosnow, 2009). This is particularly true because participants expended substantial effort within the treatment protocol and had regular interactions with instructors. The incorporation of contrast groups could partially control for expectation and social support effects in extensions of this research. Barriers self-efficacy is markedly difficult to accurately measure because the array of lifestyle barriers presented within included items (in which the respondent’s perception of ability to overcome is evaluated) might substantially differ, by individual (Buchan et al., 2015). The ideographic nature of the construct and its measurement challenges can affect scoring accuracies, especially across respondents (Teixeira et al., 2015). Although the study time frame went beyond the initial 6 to 9 months where weight loss is expected and the challenge of weight regain begins, extensions of this research should assess outcomes and their psychosocial correlates over even longer periods to better assess associated impacts on health risks during the long term. In addition, because increased exercise in low-active individuals is reliably associated with improved mood (Arent et al., 2020), and mood has been linked to improved self-regulation (Gendolla & Brinkman, 2005), future studies should consider mood as an additional mediator or moderator in extensions of this research. Finally, although addressing the study variables within a field context challenged internal validity, it also was an advantage based on the enhanced ability to readily apply findings to behavioral obesity treatments (Green et al., 2013).

In conclusion, findings supported tenets of social cognitive, self-efficacy, and self-regulation theories (Bandura, 1986, 1997; Baumeister & Vohs, 2016); and they indicated benefits for high amounts of instruction time dedicated to self-regulatory skills development within behavioral weight-loss treatments. The identified effects of barriers self-efficacy on relationships between short-term changes in self-regulation, and long-term changes in exercise outputs, fruit/vegetable intake, and self-regulatory skills usage, further inform behavioral weight-loss theory and treatment and provide guidance to practitioners. As intervention effects are further decomposed through their theory-based psychosocial correlates, revised architectures of behavioral obesity interventions will improve the lagging outcomes that have persisted for decades.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Author Biography

James J. Annesi, PhD, is a professor at the University of Alabama at Birmingham School of Health Professions, and a fellow of the American Academy of Health Behavior, The Obesity Society, and the American Psychological Association. Dr. Annesi’s research program focuses upon health behavior-change theory and methods applied to exercise adherence, controlled eating, weight management, and the effects of physical activity on mental health, body image, emotional eating, and other quality-of-life factors. In 2019, he received the American Psychological Association’s Excellence in Clinical Health Psychology Award from its Society for Health Psychology.

References

Abraham

Michie

(2008). A taxonomy of behavior change techniques used in interventions. Health Psychology, 27(3), 379–387. https://doi.org/10.1037/0278-6133.27.3.379

Alhassan

Kim

Bersamin

King

A. C.

Gardner

C. D.

(2008). Dietary adherence and weight loss success among overweight women: Results from the A TO Z weight loss study. International Journal of Obesity, 32(6), 985–991. https://doi.org/10.1038/ijo.2008.8

Aljadani

H. M.

Patterson

Sibbritt

Hutchesson

M. J.

Jensen

M. E.

Collins

C. E.

(2013). Diet quality, measured by fruit and vegetable intake, predicts weight change in young women. Journal of Obesity, 2013, Article 525161. https://doi.org/10.1155/2013/525161

Allison

P. D.

(1990). Change scores as dependent variables in regression analysis. Sociological Methodology, 20, 93–144. https://doi.org/10.2307/271083

Anderson

E. S.

Winett

R. A.

Wojcik

J. R.

(2007). Self-regulation, self-efficacy, outcome expectations, and social support: Social cognitive theory and nutrition behavior. Annals of Behavioral Medicine, 34(3), 304–312. https://doi.org/10.1007/BF02874555

Annesi

J. J.

(2018). Moderation of psychological factors in the relationship of increased fruit and vegetable intake with reductions in other food groups and weight in women with obesity. Minerva Psichiatrica, 59(1), 1–9. https://doi.org/10.23736/S0391-1772.17.01951-3

Annesi

J. J.

(2020). Sequential changes advancing from exercise-induced psychological improvements to controlled eating and sustained weight loss: A treatment-focused causal chain model. Permanente Journal, 24(3), Article 19235. https://doi.org/10.7812/TPP/19.235

Annesi

J. J.

Marti

C. N.

(2011). Path analysis of exercise treatment-induced changes in psychological factors leading to weight loss. Psychology and Health, 26(8), 1081–1098. https://doi.org/10.1080/08870446.2010.534167

Arent

S. M.

Walker

A. J.

Arent

M. A.

(2020). The effects of exercise on anxiety and depression. In Tennenbaum

Eklund

R. C.

(Eds.), Handbook of sport psychology (4th ed., pp. 872–890). Wiley.

10.

Baker

C. W.

Brownell

K. D.

(2000). Physical activity and maintenance of weight loss: Physiological and psychological mechanisms. In Bouchard

(Ed.), Physical activity and obesity (pp. 311–328). Human Kinetics.

11.

Bandura

(1986). Social foundations of thought and action: A social cognitive theory. Prentice Hall.

12.

Bandura

(1997). Self-efficacy: The exercise of control. Freeman.

13.

Bandura

(2005). The primacy of self-regulation in health promotion. Applied Psychology: An International Review, 54(2), 245–254. https://doi.org/10.1111/j.1464-0597.2005.00208.x

14.

Baranowski

Cerin

Baranowski

(2009). Steps in the design, development and formative evaluation of obesity prevention-related behavior change trials. International Journal of Behavioral Nutrition and Physical Activity, 6, Article 6. https://doi.org/10.1186/1479-5868-6-6

15.

Baumeister

R. F.

Vohs

K. D.

(2016). Strength model of self-regulation as limited resource: Assessment, controversies, update. Advances in Experimental Social Psychology, 54, 67–127. https://doi.org/10.1016/bs.aesp.2016.04.001

16.

Bertoia

M. L.

Mukamal

K. J.

Cahill

L. E.

Hou

Ludwig

D. S.

Mozaffarian

Willett

W. C.

F. B.

Rimm

E. B.

(2015). Changes in intake of fruits and vegetables and weight change in United States men and women followed for up to 24 years: Analysis from three prospective cohort studies. PLOS Medicine, 12(9), Article e1001878. https://doi.org/10.1371/journal.pmed.1001878

17.

Block

Hartman

A. M.

Dresser

C. M.

Carroll

M. D.

Gannon

Gardner

(1986). A data-based approach to diet questionnaire design and testing. American Journal of Epidemiology, 124(3), 453–469. https://doi.org/10.1093/oxfordjournals.aje.a114416

18.

Buchan

Janda

Box

Rogers

Hayes

(2015). Exercise barriers self-efficacy: Development and validation of a subscale for individuals with cancer-related lymphedema. Health and Quality of Life Outcomes, 13(1), Article 37. https://doi.org/10.1186/s12955-015-0223-7

19.

Clark

M. M.

Abrams

D. B.

Niaura

R. S.

Eaton

C. A.

Rossi

J. S.

(1991). Self-efficacy in weight management. Journal of Consulting and Clinical Psychology, 59(5), 739–744. https://doi.org/10.1037/0022-006X.59.5.739

20.

Cohen

West

S. G.

Aiken

L. S.

(2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd ed.). Lawrence Erlbaum.

21.

Cooper

Doll

H. A.

Hawker

D. M.

Byrne

Bonner

Eeley

O’Connor

M. E.

Fairburn

C. G.

(2010). Testing a new cognitive behavioural treatment for obesity: A randomized controlled trial with three-year follow-up. Behavior Research and Therapy, 48(8), 706–713. https://doi.org/10.1016/j.brat.2010.03.008

22.

Cromwell

J. B.

Hannan

M. J.

Labys

W. C.

Terraza

(1994). Multivariate tests for time series models. SAGE.

23.

Dombrowski

S. U.

Knittle

Avenell

Araújo-Soares

Sniehotta

F. F.

(2014). Long term maintenance of weight loss with non-surgical interventions in obese adults: Systematic review of meta-analyses of randomised controlled trials. British Medical Journal, 348, Article g2646. https://doi.org/10.1136/bmj.g2646

24.

Dunn

Hannan

Jeffery

Sherwood

N. E.

Pronk

N. P.

Boyle

(2006). The comparative and cumulative effects of a dietary restriction and exercise on weight loss. International Journal of Obesity, 30(1), 112–121. https://doi.org/10.1038/sj.ijo.0803046

25.

Fitzmaurice

G. M.

Laird

N. M.

Ware

J. H.

(2004). Applied longitudinal analysis. Wiley.

26.

Gendolla

G. H. E.

Brinkman

(2005). The role of mood states in self-regulation: Effects on action preferences and resource mobilization. European Psychologist, 10(3), 187–198. https://doi.org/10.1027/1016-9040.10.3.187

27.

Gillison

Stathi

Reddy

Perry

Taylor

Bennett

Dunbar

Greaves

C. J.

(2015). Processes of behavior change and weight loss in a theory-based weight loss intervention program: A test of the process model for lifestyle behavior change. International Journal of Behavioral Nutrition and Physical Activity, 12(1), Article 2. https://doi.org/10.1186/s12966-014-0160-6

28.

Godin

(2011). The Godin-Shephard Leisure-Time Physical Activity Questionnaire. Health and Fitness Journal of Canada, 4(1), 18–22. https://doi.org/10.14288/hfjc.v4i1.82

29.

Green

L. W.

Sim

Breiner

(Eds.) (2013). Evaluating obesity prevention efforts: A plan for measuring progress. Institute of Medicine of the National Academies.

30.

Gupta

S. K.

(2011). Intention-to-treat concept: A review. Perspectives in Clinical Research, 2(3), 109–112. https://doi.org/10.4103/2229-3485.83221

31.

Hales

C. M.

Carroll

M. D.

Fryar

C. D.

Ogden

C. L.

(2020). Prevalence of obesity and severe obesity among adults: United States, 2017–2018 (NCHS data brief 360). National Center for Health Statistics.

32.

Hall

K. D.

Kahan

(2018). Maintenance of lost weight and long-term management of obesity. Medical Clinics of North America, 102(1), 183–197. https://doi.org/10.1016/j.mcna.2017.08.012

33.

Hayes

A. F.

(2015). An index and test of linear moderated mediation. Multivariate Behavioral Research, 50(1), 1–22. https://doi.org/10.1080/00273171.2014.962683

34.

Hayes

A. F.

(2018). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach (2nd ed.). Guilford.

35.

Jacobs

D. R.

Ainsworth

B. E.

Hartman

T. J.

Leon

A. S.

(1993). A simultaneous evaluation of 10 commonly used physical activity questionnaires. Medicine and Science in Sports and Exercise, 25(1), 81–91. https://doi.org/10.1249/00005768-199301000-00012

36.

Jeffery

R. W.

Drewnowski

Epstein

L. H.

Stunkard

A. J.

Wilson

G. T.

Wing

R. R.

Hill

D. R.

(2000). Long-term maintenance of weight loss: Current status. Health Psychology, 19(Suppl. 1), 5–16. https://doi.org/10.1037/0278-6133.19.Suppl1.5

37.

Kruger

Galuska

D. A.

Serdula

M. K.

Jones

D. A

. (2004). Attempting to lose weight: Specific practices among U.S. adults. American Journal of Preventive Medicine, 26(5), 402–406. https://doi.org/10.1016/j.amepre.2004.02.001

38.

Linde

J. A.

Rothman

A. J.

Baldwin

A. S.

Jeffery

R. W.

(2006). The impact of self-efficacy on behavior change and weight change among overweight participants in a weight loss trial. Health Psychology, 25(3), 282–291. https://doi.org/10.1037/0278-6133.25.3.282

39.

Loveman

Frampton

G. K.

Shepherd

Picot

Cooper

Bryant

Welsh

Clegg

(2011). The clinical effectiveness and cost-effectiveness of long-term weight management schemes for adults: A systematic review. Health Technology Assessment, 15(2), 1–82. https://doi.org/10.3310/hta15020

40.

MacLean

P. S.

Wing

R. R.

Davidson

Epstein

Goodpaster

Hall

K. D.

Levin

B. E.

Perri

M. G.

Rolls

B. J.

Rosenbau

Rothman

A. J.

Ryan

(2015). NIH working group report: Innovative research to improve maintenance of weight loss. Obesity, 23(1), 7–15. https://doi.org/10.1002/oby.20967

41.

Mann

Tomiyama

Westling

Lew

A. M.

Samuels

Chatman

(2007). Medicare’s search for effective obesity treatments: Diets are not the answer. American Psychologist, 62(3), 220–233. https://doi.org/10.1037/0003-066X.62.3.220

42.

Marcus

B. H.

Selby

V. C.

Niaura

R. S.

Rossi

J. S.

(1992). Self-efficacy and the stages of exercise behavior change. Research Quarterly for Exercise and Sport, 63(1), 60–66. https://doi.org/10.1080/02701367.1992.10607557

43.

Mares-Perlman

J. A.

Klein

B. E. K.

Klein

Ritter

L. L.

Fisher

M. R.

Freudenheim

J. L.

(1993). A diet history questionnaire ranks nutrient intakes in middle-aged and older men and women similarly to multiple food records. Journal of Nutrition, 123(3), 489–501. https://doi.org/10.1093/jn/123.3.489

44.

Michie

Ashford

Sniehotta

F. F.

Dombrowski

S. U.

Bishop

French

D. P.

(2011). A refined taxonomy of behaviour change techniques to help people change their physical activity and healthy eating behaviours: The CALO-RE taxonomy. Psychology and Health, 26(11), 1479–1498. https://doi.org/10.1080/08870446.2010.540664

45.

National Institutes of Health/National Cancer Institute. (2020). Dietary assessment primer: Validation. https://dietassessmentprimer.cancer.gov/profiles/recall/validation.html

46.

Pereira

M. A.

FitzGerald

S. J.

Gregg

E. W.

Joswiak

M. L.

Ryan

W. J.

Suminski

R. R.

Utter

A. C.

Zmuda

J. M.

(1997). A collection of physical activity questionnaires for health-related research. Medicine and Science in Sports and Exercise, 29(Suppl. 6), S1–S205.

47.

Piercy

K. L.

Troiano

R. P.

Ballard

R. M.

Carlson

S. A.

Fulton

J. E.

Galuska

D. A.

George

S. M.

Olson

R. D.

(2018). The physical activity guidelines for Americans. Journal of the American Medical Association, 320(19), 2020–2028. https://doi.org/10.1001/jama.2018.14854

48.

Pituch

Stevens

(2016). Applied multivariate statistics for the social sciences. Routledge. https://doi.org/10.4324/9781315814919

49.

Rosenthal

Rosnow

R. L.

(2009). Artifacts in behavioral research. Oxford.

50.

Rothman

A. J.

Baldwin

A. S.

Hertel

A. W.

(2004). Self-regulation and behavior change: Disentangling behavioral initiation and behavioral maintenance. In Baumeister

R. F.

Vohs

K. D.

(Eds.), Handbook of self-regulation: Research, theory, and applications (pp. 130–148). Guilford.

51.

Rucker

D. D.

Preacher

K. J.

Tormala

Z. L.

Petty

R. E.

(2011). Mediation analysis in social psychology: Current practices and new recommendations. Social and Personality Psychology Compass, 5(6), 359–371. https://doi.org/10.1111/j.1751-9004.2011.00355.x

52.

Saelens

B. E.

Gehrman

C. A.

Sallis

J. F.

Calfas

K. J.

Sarkin

J. A.

Caparosa

(2000). Use of self-management strategies in a 2-year cognitive-behavioral intervention to promote physical activity. Behavior Therapy, 31(2), 365–379. https://doi.org/10.1016/S0005-7894(00)80020-9

53.

Schafer

J. L.

Graham

J. W.

(2002). Missing data: Our view of the state of the art. Psychological Methods, 7(2), 147–177. https://doi.org/10.1037/1082-989X.7.2.147

54.

Spencer

S. J.

Zanna

M. P.

Fong

G. T.

(2005). Establishing a causal chain: Why experiments are often more effective than mediational analyses in examining psychological processes. Journal of Personality and Social Psychology, 89(6), 845–851. https://doi.org/10.1037/0022-3514.89.6.845

55.

Swift

D. L.

Johannsen

N. M.

Lavie

C. J.

Earnest

C. P.

Church

T. S.

(2014). The role of exercise and physical activity in weight loss and maintenance. Progress in Cardiovascular Diseases, 56(4), 441–447. https://doi.org/10.1016/j.pcad.2013.09.012

56.

Teixeira

P. J.

Carraça

E. V.

Marques

M. M.

Rutter

Oppert

J. M.

De Bourdeaudhuij

Lakerveld

Brug

(2015). Successful behavior change in obesity interventions in adults: A systematic review of self-regulation mediators. BMC Medicine, 13, Article 84. https://doi.org/10.1186/s12916-015-0323-6

57.

U.S. Department of Agriculture. (2017). MyPlate and historical Food Pyramid resources. National Agricultural Library.

58.

U.S. Department of Health and Human Services and U.S. Department of Agriculture. (2015). 2015–2020 dietary guidelines for Americans (8th ed.). https://health.gov/dietaryguidelines/2015/guidelines/

59.

Vohs

K. D.

Baumeister

R. F.

(Eds.). (2016). Handbook of self-regulation: Research, theory, and application (3rd ed.). Guilford.

60.

White

I. R.

Horton

N. J.

Carpenter

Pocock

S. J.

(2011). Strategy for intention to treat data in randomized trials with missing outcome data. British Medical Journal, 342, Article d40. https://doi.org/10.1136/bmj.d40

61.

Williamson

D. A.

Bray

G. A.

Ryan

D. H.

(2015). Is 5% weight loss a satisfactory criterion to define clinically significant weight loss? Obesity, 23(12), 2319–2320. https://doi.org/10.1002/oby.21358