Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Apr 1.
Published in final edited form as: Psychol Health. 2021 Aug 29;38(4):429–444. doi: 10.1080/08870446.2021.1967953

Social comparisons between group members during behavioural weight loss treatment: comparison direction, scale, and associations with weight loss maintenance

Danielle Arigo 1, Savannah R Roberts 2, Meghan L Butryn 3
PMCID: PMC9382642  NIHMSID: NIHMS1771595  PMID: 34459320

Abstract

Objective:

To examine distinct types of social comparisons (i.e., self-evaluations relative to others) in behavioural weight loss groups and their relations with weight loss maintenance.

Design:

Participants (N = 127, MBMI = 35.66 kg/m2) reported on their comparisons at mid-treatment (6 months), including identification of their primary individual comparison target (group member) and perceptions of their own treatment adherence versus that of their group and identified target.

Main Outcome Measures:

Weight was assessed at baseline, mid-treatment, end-of-treatment (12 months), and 18- and 24-month follow-ups.

Results:

Comparisons with individual targets perceived as more successful with weight loss were most frequent (i.e., upward comparisons), though comparisons differed based on group versus individual targets and specific treatment behaviors (e.g., self-monitoring). Comparisons did not align with participants’ own treatment progress, suggesting that comparisons reflect more than just their objective weight loss relative to others. Relations between participants’ initial weight loss and maintenance was moderated by the type of individual target identified at mid-treatment (p = 0.02, sr = 0.27).

Conclusions:

Social comparisons in group-based weight loss treatment are multifaceted and predict long-term weight loss maintenance. Additional work is needed to determine how best to harness comparison processes to promote success in behavioural weight loss treatment.

Keywords: weight loss, social comparison, behavioural treatment, obesity, social influence

Behavioural weight loss treatment can produce clinically significant weight loss in the short term (i.e., 5%−10% weight loss over six months), but many participants in these interventions regain lost weight within the following year (Christian, Tsai, & Bessesen, 2010; MacLean et al., 2014; Varkevisser et al., 2019). Thus, sustaining weight after initial weight loss – particularly after weight loss treatment ends (i.e., weight loss maintenance) – is a key challenge for obesity treatment research and practice (Lean & Hankey, 2018). There is a critical need to improve upon the current understanding of barriers to, and facilitators of, longer-term success in behavioural weight loss programs. Given that these interventions often are delivered in groups of 8–14 individuals who meet face-to-face over 6–18 months (Curry et al., 2018), social processes such as support and accountability have been documented as important facilitators of change for participants in behavioural weight loss programs (Gorin et al., 2014; Kiernan et al., 2012; see Mohr et al., 2011).

Similarly, social comparison, or evaluating oneself relative to others (Festinger, 1954), is known to contribute to outcomes in behavioural weight loss interventions (Leahey, Kumar, Weinberg, & Wing, 2012; Oleander et al., 2013). Members of a participant’s treatment group offer a variety of relevant potential comparison targets, and each of these targets provides different information about the participant’s weight loss status. This information can motivate the participant to engage in recommended dietary and physical activity changes, to catch up to or stay ahead of group members’ weight loss progress (Merchant et al., 2017). As early success in behavioural weight loss programs and the habits associated with this success are strong predictors of weight loss maintenance post-treatment (Elfhag & Rössner, 2005), it is possible that social comparison activity early in treatment (i.e., in the first 6 months) contributes to longer-term success.

Yet, social comparison is a complex process that is associated with both positive and negative health outcomes, both within and beyond the context of weight loss (Arigo, Suls, & Smyth, 2014; Buunk et al., 1990; Schokker et al., 2010). For example, some individuals report stronger social comparison orientation than others, or the tendency to make comparisons; these individuals tend to experience elevated neuroticism, depressive symptoms, and negative affect (Gibbons & Buunk, 1999). Among behavioural weight loss treatment participants, pre-treatment level of social comparison orientation is positively associated with long-term maintenance of physical activity but negatively associated with long-term maintenance of weight loss (i.e., at 1 year post-treatment; Arigo & Butryn, 2019). These mixed findings indicate that an individual’s self-report of their general tendency to make comparisons is only part of a wider landscape linking social comparison processes to health outcomes and behaviors. Examining other facets of the comparison process could deepen our understanding of the role of social comparisons in weight loss maintenance, and inform improvements to behavioral weight loss treatment.

The association between comparisons and health outcomes (beyond the context of weight loss) is known to depend on a variety of contextual factors, particularly the direction of comparison, which refers to the comparer’s perception of their status relative to the comparison target. Comparisons to targets perceived to be “doing better” than the comparer on a relevant dimension are upward comparisons; downward comparisons are to targets perceived to be “doing worse,” and lateral comparisons are to targets perceived to be “about the same” (Wood, Taylor, & Lichtman, 1989). These perceptions of relative standing often are more influential than one’s standing when objectively assessed on the same characteristic (Klein, 1997; Wood & Wilson, 2003), and also involve perceptions of scale (Buunk, 1995). For example, the comparison target may be perceived as doing a little bit to a lot better or worse than the self (see Muller & Fayant, 2010). Previous work in the broader social comparison space also has specified that comparisons to individuals, versus comparisons to a group average, are related but distinct processes (Zell & Alike, 2010). To date, however, these aspects of comparison have received little attention with respect to the context of behavioural weight loss treatment.

When applied to the setting of behavioural weight loss treatment groups, the broader social comparison literature suggests that upward comparisons toward group members can provide inspiration and information about how to improve; downward and lateral comparisons toward group members can lead to satisfaction with one’s status, and/or motivation to avoid becoming like worse-off others or being outperformed by group members (cf. Carmack-Taylor et al., 2007; Garcia, Tor, & Schiff, 2013; Rancourt, Leahey, LaRose, & Crowther, 2014). These comparisons also can have the opposite effects, however. Upward comparisons can highlight the comparer’s inferiority and can activate frustration or embarrassment (which can lead to giving up on weight loss attempts; Merchant et al., 2017); downward and lateral comparisons can indicate that the comparer is already doing well with weight loss and offer a reason to relax their efforts (Wills, 1981). This range of positive and negative responses is likely even when comparers self-select their targets (vs. being presented with a single target; Arigo, Smyth, & Suls, 2015; Van der Zee et al., 1998).

Thus, the role of comparisons in behavioural weight loss treatment outcomes such as post-treatment weight loss maintenance may depend on direction and scale of each of the selected comparison target(s), as well as how comparison information is interpreted. Alternatively, it is possible that participants’ comparisons to their group members during treatment simply reflect their own progress in the program (e.g., participants who lose the most weight make downward comparisons because downward targets are more readily accessible than upward targets), rather than an independent psychosocial process that differentially predicts future weight loss maintenance. Despite the current appreciation for social comparison as a process that can facilitate behaviour change in group-based behavioural weight loss treatment (cf. Oleander et al., 2013), these more nuanced aspects of the comparison process as they function between group members have not yet been addressed.

The goal of the present study was to provide initial insight into relations between behavioural weight loss participants’ comparisons to their treatment group members and their weight loss maintenance post-treatment, which may be useful for maximizing the potential benefit of comparison processes in this context. Participants reported on comparisons to their perceived group average and to an individual group member at the midpoint of behavioural weight loss treatment (i.e., 6 months into treatment). Percent weight lost from baseline was assessed objectively during treatment and then through 12 months post-treatment (i.e., post-treatment maintenance). The specific aims of this study were to evaluate: (1) discrepancies between participants’ ratings of their own, their group’s, and an individual comparison target’s adherence to treatment recommendations, (2) the direction and scale of comparisons to an individual group member during treatment, (3) predictors of individual comparison target selection during treatment (direction), and (4) relations between comparison target selection during treatment (direction) and weight loss maintenance for 12 months post-treatment.

Methods

Participants

The present study was approved by the Institutional Review Board at the supporting institution, and all participants provided written informed consent upon enrollment. Radio and print advertisements were used to recruit adults from the general community to a behavioural weight loss clinical trial (NCT01858714). Individuals were eligible if they were between the 18 and 70 years old, had BMIs between 27 and 45 kg/m2, reported that they did not have physical activity limitations, and completed screening procedures (e.g., returned 7-day food records). Individuals were excluded if they were currently pregnant, had any changes in medication that could affect weight, or had plans to enroll in a different weight loss program. In this sub-study, participants enrolled in three out of five waves of treatment were administered a self-report measure of social comparisons with group members (N = 127, MAge = 52.89 years, MBMI = 35.66 kg/m2). The majority were women (81.9%), married (56.8%), and identified as Caucasian (68.3%). Most participants had at least a high school diploma (94.4%) and reported household incomes less than $100,000 per year (71.8%). These demographic indicators were comparable to those of the full clinical trial sample (blinded for review).

Primary Measures

Social comparison questionnaire.

As a validated measure of social comparisons between group members was not available at the time of data collection, the authors created a study-specific measure to assess: (1) comparisons to the treatment group as a whole, and (2) comparisons to an individual treatment group member (see Appendix A). This measure asked participants to identify the group member that they compared to most often during treatment, and to rate their own, their group’s, and their individual comparison target’s adherence to program recommendations (e.g., sticking to a daily calorie goal) on a scale of 1 (none of the time) to 6 (all of the time).

With respect to the individual comparison target, participants were asked to indicate whether their identified target had lost “much more,” “somewhat more,” “about the same,” “somewhat less,” or “much less” weight than they had. (Thus, individual comparisons were restricted to the dimension of weight loss.) Response options corresponded to strong (distant) upward targets, weak (close) upward targets, lateral targets, weak (close) downward targets, and strong (distant) downward targets, respectively. Results below include descriptive information about selection for each of these five categories. Consistent with previous work (Leahey, Crowther, & Ciesla, 2011; Rancourt, Leahey, LaRose, & Crowther, 2015 Lennarz, Lichtwarck-Aschoff, Finkenauer, & Granic, 2017), strong and weak targets in each direction were collapsed to form three categories (i.e., upward, lateral, and downward) for inferential tests.

Weight loss and maintenance.

Behavioural weight loss treatment lasted for 1 year, in 2 phases. The first 6 months involved weekly group meetings with content focused on weight loss. The frequency of group meetings decreased during the second 6 months and content shifted to include discussion of continued weight loss and maintenance of weight lost to date. Assessments of weight loss took place at baseline and every 6 months for the following 2 years (i.e., 1 year of active treatment, 1 year of follow-up). Participants’ height and weight were measured by treatment staff at each assessment point and were used to calculate BMI. For each post-baseline assessment, including mid-treatment (6 months) and three post-treatment points (maintenance; 12 months [end of treatment], 18 months, and 24 months), percent weight loss was calculated as percent change from the participant’s baseline weight. Weight loss maintenance was operationalized as percent weight loss from baseline during the post-treatment assessment phase (12, 18, and 24 months post-treatment); “regain” describes reductions in percent weight loss after the end of treatment.

Potential predictors of comparison target identification.

To better understand individual characteristics and behaviors that may be associated with the identification of an individual group member as a comparison target, the following variables were evaluated for differences between participants based on their identified target (using the combined upward, lateral, and downward categories): self-reported age and gender at the start of treatment, number of pre-treatment weight loss attempts and expected weight loss (assessed with the Weight And Lifestyle Inventory [WALI]; Wadden & Foster, 2006), pre-treatment depressive symptoms (assessed with the Beck Depression Inventory [BDI-II]; Beck et al., 1996), treatment condition, treatment group assignment, and session attendance through mid-treatment (recorded by the group clinician at each treatment session).

Procedure

After completing screening and enrollment procedures, participants were randomized to one of three behavioural weight loss interventions; all interventions occurred on the same schedule over the course of one year, with the majority of sessions taking place over the first 6 months of treatment (26 sessions total; reference blinded for review). Participants were encouraged to eat between 1200 and 1800 calories per day (based on starting weight and rate of weight loss) and to engage in moderate-to-vigorous exercise for 30 or more minutes on 3–5 days per week (increasing as treatment progressed), with the goal of 5%−10% weight loss by mid-treatment. The intervention conditions differed in the combination and number of psychological techniques used (e.g., home environment changes, acceptance-based techniques). Psychology master’s- or doctoral-level clinicians facilitated group treatment sessions with 10–14 participants per group and the same groups of clinicians and participants met for each session. Random assignment to treatment condition and scheduling accommodations resulted in heterogeneity in gender, age, racial/ethnic identification, and starting weight in all groups. As part of their regular assessment batteries, participants in the last 3 out of 5 waves of treatment were asked to complete the social comparison measure created for this study. This measure was administered at mid-treatment (6 months).

Data Analysis

Participants’ ratings for their own, their group’s, and their individual comparison target’s (identified group member’s) treatment adherence were evaluated with descriptive statistics (means, standard deviations). Paired t-tests then were used to test for differences between group and self-ratings and between individual target and self-ratings (Aim 1). Individual comparison target selections (direction/scale) also were examined descriptively (Aim 2). ANOVA models were used to test for differences in pre-treatment participant characteristics (e.g., age), treatment characteristics (e.g., condition), and percent weight loss at mid-treatment between participants who selected different targets (Aim 3). As only a subset of participants (n = 82) identified an individual group member as their primary comparison target, ANOVA models used combined direction categories (i.e., upward, lateral, and downward).

Finally, two-level multilevel models with assessments nested within participants were used to test for relations between individual target selection (combined upward, lateral, and downward categories) and percent weight loss across assessments at 12, 18, and 24 months (Aim 4). Models employed restricted maximum likelihood estimation methods. To address the question of whether comparison uniquely predicts future weight loss (vs. reflects past weight loss only), percent weight loss at 6 months was examined first as a covariate; a second model tested the interaction between percent weight loss at 6 months and individual target direction for predicting weight loss maintenance. Statistical significance for inferential tests was set to p < 0.05; effect sizes are presented with respect to Cohen’s d (for t-tests) and semipartial correlation coefficients (sr, for multilevel models).

Results

At mid-treatment (6 months), participants’ ratings of adherence to treatment recommendations for their group (overall), their identified comparison target, and themselves were high, with means above 4 (adhering a good bit of the time) on a 6-point scale (see Table 1). Participants perceived themselves as doing better than the group with respect to recording their food intake and behaviors such as attending group sessions and self-weighing at home (ps < 0.05, ds = 0.17–0.35), though their group and self-ratings for adhering to physical activity and calorie goals did not differ (ps > 0.10). Conversely, participants perceived their identified target as doing better than themselves with adhering to physical activity and calorie goals (ps < 0.05, ds = 0.19–0.29), though their target and self-ratings for recording intake and attendance/self-weighing behaviors did not significantly differ (ps > 0.10). In sum, both the full group and the identified group member served as lateral targets for certain behaviors, though participants made downward comparisons with respect to self-monitoring behaviors and upward comparisons for physical activity and calorie goal adherence.

Table 1.

Group, target, and self-ratings for adherence to behavioral weight loss treatment recommendations (assessed at mid-treatment).

Group Ratings (N = 127)
Meeting Treatment Recommendation Self M (SD) Group M (SD) Difference (t, d)
Physical activity 4.33 (1.30) 4.33 (0.82) 0.00, 0.00
Calorie goals 4.35 (1.14) 4.48 (0.75) 1.16, 0.10
Recording intake 4.83 (1.27) 4.58 (0.82) −1.92*, 0.17
Other (attendance, self-weighing) 5.02 (1.05) 4.61 (0.81) −3.91**, 0.35
Individual Identified Target Ratings (N = 100)
Meeting Treatment Recommendation Self M (SD) Target M (SD) Difference (t, d)
Physical activity 4.33 (1.30) 4.75 (1.02) 2.91**, 0.29
Calorie goals 4.35 (1.14) 4.77 (1.03) 1.91*, 0.19
Recording intake 4.83 (1.27) 5.00 (0.99) 1.02, 0.10
Other (attendance, self-weighing) 5.02 (1.05) 4.92 (0.90) −1.01, 0.10
Open in a new tab

Note: Ratings on a scale of 1 (none of the time) to 6 (all of the time); negative values reflect that self-ratings were higher than those of the group/target (downward comparisons); positive values reflect that self-ratings were lower than those of the group/target (upward comparisons)

*

p < 0.05

**

p < 0.01.

Predicting and Describing Target Selection

As noted, 82 participants indicated the direction and scale of the primary group member they used as an individual comparison target, whereas 45 did not provide this information. Those who did versus did not indicate a target selection did not differ by age, gender, number of pre-treatment weight loss attempts or expected weight loss, pre-treatment depressive symptoms, treatment condition, treatment group assignment, session attendance, or percent weight loss at mid-treatment (ps > 0.15). Within upward and downward directions, strong and weak selections were combined to form three comparison target categories (upward, lateral, and downward; see Table 2). Upward targets were most popular (52.4% of identified targets), followed by lateral and downward targets (24.4% and 23.2%, respectively). Target selection across these categories did not differ by age, gender, number of pre-treatment weight loss attempts or expected weight loss, pre-treatment depressive symptoms, treatment condition, treatment group assignment, or session attendance (ps > 0.40). Thus, target selection did not appear to be driven by pre-treatment individual differences, treatment characteristics, or treatment dose.

Table 2.

Descriptive statistics for individual target identification categories.

Percent Weight Loss from Baseline (Mid-Treatment; 6 months)
All Selections Frequency (%) M (SD) Min Max
Strong Upward 19 (23%) 6.27 (5.42) −5.71 15.06
Weak Upward 24 (29%) 8.67 (3.82) 2.16 14.61
Lateral 20 (24%) 9.46 (4.75) −1.34 19.37
Weak Downward 15 (18%) 8.05 (5.30) 0.42 16.44
Strong Downward 4 (6%) 15.18 (7.30) 5.79 21.89
Combined Selections Frequency (%) M (SD) Min Max
Upward 43 (53%) 9.63 (6.34) 0.42 21.89
Lateral 20 (24%) 9.46 (4.75) −1.34 19.37
Downward 19 (23%) 7.59 (4.71) −5.71 15.06
Open in a new tab

Further, perceptions of identified targets did not appear to reflect merely participants’ success at mid-treatment. Percent weight loss did not differ between the three combined direction categories (F[2,80] = 1.45, p = 0.24); average percent weight loss was lowest among those who selected upward targets (7.6%), but did not significantly differ from those who selected downward or lateral targets (9.6% and 9.5%, respectively; contrast F[1,81] = 2.89, p = 0.10). Considerable ranges of weight loss were observed within each group of participants who identified the same direction (and direction/scale) of comparison target (see Table 2), suggesting that target identification did not uniformly correspond to participants’ own weight loss success.

Relations between Target Selection and Weight Loss Over Time

Controlling for initial weight loss (percent weight loss at 6 months), percent weight loss across the maintenance period of 12, 18, and 24 months from baseline did not differ by identified comparison target category at mid-treatment (6 months; F = 0.12, p = 0.74). However, there was a significant interaction effect between initial weight loss and identified comparison target on percent weight loss across the maintenance period (F = 5.99, p = 0.02, sr = 0.27; see Table 3), indicating that identified comparison target moderated the relation between initial weight loss and weight loss maintenance after treatment ended. As shown in Figure 1, percent weight loss after treatment cessation (i.e., maintenance) for participants with lower weight loss at mid-treatment was similar across target selection categories (<5% by 24 months), and participants who selected upward targets appeared to have the most weight regain over time. Among participants with higher weight loss at mid-treatment, those who selected upward targets showed dramatic weight regain (to ~6% by 24 months), whereas those who selected lateral and downward targets fared better (i.e., >8% weight loss by 24 months). Specifically, participants who selected lateral targets experienced the least weight regain.

Table 3.

Model estimates for predicting weight loss maintenance during 12 months post-treatment (12, 18, and 24 months from baseline; n = 82).

Controlling for Mid-Treatment Weight Loss B (SE)
Intercept 12.77 (1.08)***
Percent Weight Loss at Mid-Treatment 1.23 (1.09)***
Time (Assessment Point) −1.49 (0.44)**
Identified Target Direction −0.57 (0.74)
Identified Target Direction*Time (Assessment Point) −1.00 (0.30)
Interaction between Mid-Treatment Weight Loss and Identified Target Direction B (SE)
Intercept 12.79 (1.10)***
Percent Weight Loss at Mid-Treatment 1.15 (0.18)***
Time (Assessment Point) −1.45 (0.43)**
Identified Target Direction −0.55 (0.74)
Percent Weight Loss at Mid-Treatment*Identified Target Direction*Time (Assessment Point) −0.12 (0.05)*
Open in a new tab

Note:

*

p < 0.05

**

p < 0.01

***

p < 0.001; identified target direction = three categories of individual target selection (upward, downward, and lateral).

Figure 1.

Figure 1.

Open in a new tab

Weight loss maintenance (i.e., percent weight loss from baseline) across one year post-treatment (12 months, 18 months, and 24 months from baseline), by initial weight loss success and identified comparison target direction at mid-treatment (6 months from baseline). Initial weight loss success (i.e., percent weight loss at 6 months) displayed at +/−½ SD for illustrative purposes.

Discussion

Although somewhat preliminary, the present findings offer considerable insight into comparison processes among participants in behavioural weight loss treatment and their potential relations with weight loss maintenance post-treatment. First, results show that weight loss comparisons in group treatment can occur with respect to both the group as a whole and to individual group members, and that these comparisons look different on distinct dimensions. Specifically, participants elected to compare with individual group members who were more adherent to calorie and physical activity treatment recommendations than they were, and who lost more weight than they did, reflecting upward comparisons. In contrast, participants perceived themselves as doing better than their group with respect to food intake recording and behaviors such as session attendance, reflecting downward comparisons with the group as a whole. Although comparisons to individuals versus groups (or averages) are recognized as related but discrete processes (Zell & Alicke, 2010), to our knowledge, this study is one of the few that has captured both, in the context of a health behavior intervention.

It is possible that individual and group comparisons serve different functions. Though upward individual comparisons may be selected for their (longer-term) motivational benefits, also making self-enhancing downward group comparisons could buffer against the (shorter-term) negative emotions associated with upward comparisons, to balance between the positive effects of each (cf. Van der Zee, Oldersma, Buunk, & Bos, 1998). It also is possible that participants see individual comparison targets and the whole treatment group as differentially relevant to specific dimensions of weight loss (cf. Harrison, Marshall, Bianchi-Berthouze, & Bird, 2015), though this level of granularity with respect to comparison dimension rarely is captured in research on social comparisons of health behaviors. Further, the majority of evidence relating social comparisons to health behaviors and outcomes comes from assessing social comparison orientation (i.e., the general tendency to make comparisons; Gibbons & Buunk, 1999) or a single instance of comparison, as in experimental studies (e.g., Derlega et al., 2008). Very few studies assess multiple comparisons (or types of comparisons) made by the same person, as in the present study. Additional work in this area could highlight the uniqueness of different comparison processes and their independent or combined influences on weight loss and related outcomes (see Arigo et al., 2020).

Second, the present findings indicate that comparison target selection does not simply reflect relative treatment progress or opportunity to make certain types of comparisons. There was noteworthy heterogeneity in participants’ own weight loss within selection categories, and the distribution of target selections did not match the distribution of weight loss at the same time point: weight loss was normally distributed but selection was skewed toward upward targets. Thus, participants who experienced high initial weight loss did not default to selecting downward targets because there were fewer targets “doing better” than they were with respect to weight loss; in fact, many of these participants still chose upward targets. This observation is consistent with existing work showing discrepancies between objective and social comparison perceptions (Klein, 1997), particularly in the domain of weight (Sánchez, Dijkstra, & Visser, 2015), and indicates that social comparison operates independently from objective treatment progress.

Yet, target selection alone did not predict long-term weight loss maintenance. Based on evidence from the broader social comparison literature, it is likely that participants differed in one or more aspects of the comparison process, such that target direction did not provide the full context for understanding their future weight loss maintenance. These include their reasons or outcome expectations for comparisons to individual targets (e.g., mood enhancement, learning useful information; Bandura & Jourden, 1991; Wood, 1989), their interpretations of the information gleaned (i.e., perceived similarity; Buunk & Ybema, 1997), and whether their interpretations are motivating. In line with this perspective, and perhaps most important for understanding how social comparisons are associated with behavioural weight loss outcomes, the relation between percent weight loss at mid-treatment and weight loss maintenance during the 12 months post-treatment was moderated by the direction of comparison to an individual target at mid-treatment.

Participants with higher initial weight loss did better with respect to maintenance overall (and those with lower initial weight loss did worse), which aligns with existing work on long-term success with behavioural weight loss treatment (Elfhag & Rössner, 2005). However, patterns of weight loss maintenance differed based on identified comparison target within these broad clusters. Those who began with high initial weight loss and selected lateral targets showed the best weight loss maintenance post-treatment. It is possible that seeing one’s own treatment progress as on par with the comparison target provided motivation to continue engaging with successful but challenging treatment recommendations, as it highlighted participants’ ability to “keep pace” with a relevant peer (Helgeson & Mickelson, 1995). Those with higher initial weight loss who selected downward targets also showed impressive maintenance, possibly stemming from satisfaction with one’s performance and motivation to avoid an undesirable future state (e.g., “sticking to a calorie goal is tough but I’m doing better than they are, I can keep going and be successful;” Wills, 1981). Yet neither type of comparison appeared to be as useful for participants with lower initial weight loss.

Despite the potential for inspiration and motivation from upward targets (e.g., “improvement is possible;” Buunk & Ybema, 1997), participants who selected these targets fared worse than their counterparts. In particular, those with higher initial weight loss who selected upward targets showed steep declines post-treatment (indicating weight regain), which may reflect dejection in response to upward comparison (e.g., “I’ll never achieve what they have, why even try?” Van der Zee, Buunk, Sanderman, Botke, & Van den Bergh, 2000). These participants may have few upward targets from which to select, forcing those who prefer upward targets to compare with unusually successful group members (i.e., those who lose more than 10% of their initial weights in the first 6 months of treatment; Nackers, Ross, & Perri, 2010). As declines were much more modest among participants with lower initial weight loss who selected upward targets, it is possible that negative reactions to upward comparisons in weight loss treatment are stronger at higher levels of achievement (e.g., “I’ve worked so hard and I still can’t achieve what they have;” Muller & Fayant, 2010).

Overall, the results of this study provide compelling, if preliminary, evidence that comparison to group members in behavioural weight loss treatment is not universally beneficial or harmful. Evidence-based behavioural weight loss treatment protocols and their modern adaptations reference social comparison as a negative cognitive process that is a candidate for cognitive restructuring (Brownell, 2004; Diabetes Prevention Program Research Group, 2002), Although this may be an optimal approach for some individuals (e.g., those who use or interpret comparisons in ways that demotivate healthy weight control behaviors), considering a given treatment participant’s initial success could help to refine the use of comparison in group weight loss treatment and inform targeted maintenance treatments. For example, partner activities during group sessions could pair members with aligned comparison motives, such that each individual is paired with a target that provides benefit (e.g., motivation, self-efficacy), and discussions about restructuring comparative thoughts could emphasize shifts to positive interpretations (vs. shifts away from comparison altogether). Reassigning group membership based on weight loss at mid-treatment to maximize the benefit of naturally occurring comparisons also represents a compelling line of investigation for future research.

Finally, as noted, previous work has shown relations between baseline global social comparison orientation and weight loss maintenance post-treatment (Arigo & Butryn 2019, from the same trial as the present findings). This construct is much broader and retrospective than reports of comparisons made to group members during treatment; future work to determine the unique and combined roles of both aspects of social comparison activity would take important next steps in this area. Specifically, work that examines interactions between pre-existing comparison orientation and reported comparison targets during treatment might further help to explain individual differences in weight loss maintenance. Behavioral weight loss trials with large, diverse samples would be optimal for this purpose, to allow for drawing strong inferences about the independent and interactive roles of global social comparison orientation and selected targets during treatment for predicting post-treatment outcomes.

Strengths of this study are its novel approach to understanding the role of social comparison in group-based behavioural weight loss treatment (i.e., reported comparisons with group members), assessment of both group and single-member comparisons across multiple specific dimensions of weight control behavior, and an objective and long-term primary outcome (i.e., percent weight loss maintained over 12 months post-treatment). Limitations include a modest sample size for inferential tests and the use of a study-specific measure of social comparison. The multilevel analytic approach employed in this study was selected to maximize power from the sample size available and address the nested structure of the data (Singer, Willett, & Willett, 2003). However, replication and extension are needed to clarify the generalizability of the present findings.

As noted, the authors are not aware of a validated measure of comparison to weight loss treatment group members, and available measures of social comparison were deemed inappropriate for the current research questions. The items included in this measure were based on those of existing tools and on the authors’ understanding of the broader social comparison theory and evidence (blinded for review), and the measure used in this study requires validation in future research. Also noteworthy is that the comparison measure used in this study asked about the most frequent target of comparison, rather than the one that was most valued. Although these may be the same target for some individuals, they may be distinct for others; the possibility of observing different patterns with wording focused on valued comparisons should be investigated in future research.

In addition, a subset of participants did not provide a response to the item assessing their primary individual comparison target’s weight loss. It is not clear whether this represents true lack of comparison activity, poor memory for comparisons, reluctance to report comparison activity, or another process (Gibbons & Buunk, 1999; Helgeson & Taylor, 1993; Hemphill & Lehman, 1993). Response bias was not detected with respect to pre-treatment individual differences or treatment characteristics, though other factors not assessed here may differentiate those who do and do not report target selections. Finally, it is possible that comparisons with the treatment group and individual members may change over time (Arigo et al., 2020), and that these changes are critical to understanding weight loss maintenance patterns. Future studies in this domain could be most useful for elucidating the utility of comparisons in weight loss treatment if they have the following features: (1) assessment of responses to various types of comparisons (e.g., positive/negative interpretations or affect, how motivating the comparison is), (2) repeated or intensive assessment methods (e.g., ecological momentary assessment) to capture comparison at more than one time point, and (3) examination of temporal relations between comparisons, engagement in weight control behaviors, and longer-term weight change.

Conclusions

Together, findings from this study provide important, if preliminary, insight into the process of social comparison in group-based behavioural weight loss treatment. Results indicate that comparisons to one’s treatment group overall and to individual group members are distinct processes that occur with respect to different dimensions of weight loss (i.e., physical activity, food intake, self-monitoring), and that relations between initial weight loss and post-treatment maintenance of weight loss may depend on participants’ comparisons to individual group members during treatment. As lateral comparisons were associated with the best weight loss maintenance, understanding how these comparisons may be beneficial (and for whom) will be a valuable next step. Continuing to investigate multiple aspects of comparison and considering how best to incorporate a nuanced understanding of the comparison process into treatment are likely to improve maintenance of weight loss after participation in group-based behavioural programs.

Supplementary Material

Appendix

References

  1. Arigo D, & Butryn ML (2019). Prospective relations between social comparison orientation and weight loss outcomes. Behavioural Medicine, 45(3), 249–254. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Arigo D, Mogle JA, Brown MM, Pasko K, Travers L, Sweeder L, & Smyth JM (2020). Methods to assess social comparison processes within persons in daily life: A scoping review. Frontiers in Psychology, 10, 2909. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Arigo D, Smyth JM, & Suls JM (2015). Perceptions of similarity and response to selected comparison targets in type 2 diabetes. Psychology & Health, 30(10), 1206–1220. [DOI] [PubMed] [Google Scholar]
  4. Arigo D, Suls JM, & Smyth JM (2014). Social comparisons and chronic illness: research synthesis and clinical implications. Health Psychology Review, 8(2), 154–214. [DOI] [PubMed] [Google Scholar]
  5. Bandura A, & Jourden FJ (1991). Self-regulatory mechanisms governing the impact of social comparison on complex decision making. Journal of Personality and Social Psychology, 60(6), 941–951. [Google Scholar]
  6. Beck AT, Steer RA, & Brown GK (1996). Manual for the beck depression inventory-II. San Antonio, TX: Psychological Corporation, 1, 82. [Google Scholar]
  7. Brownell KD (2004). The LEARN program for weight management: Lifestyle, exercise, attitudes, relationships, nutrition. American Health Publishing Company. [Google Scholar]
  8. Buunk BP (1995). Comparison direction and comparison dimension among disabled individuals: Toward a refined conceptualization of social comparison under stress. Personality and Social Psychology Bulletin, 21(4), 316–330. [Google Scholar]
  9. Buunk BP, Collins RL, Taylor SE, VanYperen NW, & Dakof GA (1990). The affective consequences of social comparison: either direction has its ups and downs. Journal of Personality and Social Psychology, 59(6), 1238–1249. [DOI] [PubMed] [Google Scholar]
  10. Buunk BP, & Ybema JF (1997). Social comparisons and occupational stress: The identification-contrast model. In Buunk BP, & Ybema JF (Eds.), Health, Coping, and Well-being: Perspectives from Social Comparison Theory (p. 359–388). Mahwah, NJ: Lawrence Erlbaum Associates, Inc. [Google Scholar]
  11. Carmack-Taylor CL, Kulik J, Badr H, Smith M, Basen-Engquist K, Penedo F, & Gritz ER (2007). A social comparison theory analysis of group composition and efficacy of cancer support group programs. Social Science & Medicine, 65(2), 262–273. [DOI] [PubMed] [Google Scholar]
  12. Christian JG, Tsai AG, & Bessesen DH (2010). Interpreting weight losses from lifestyle modification trials: using categorical data. International Journal of Obesity, 34(1), 207–209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Curry SJ, Krist AH, Owens DK, Barry MJ, Caughey AB, Davidson KW, … & Kubik, M. (2018). Behavioural weight loss interventions to prevent obesity-related morbidity and mortality in adults: US preventive services task force recommendation statement. JAMA, 320(11), 1163–1171. [DOI] [PubMed] [Google Scholar]
  14. Derlega VJ, Greene K, Henson JM, & Winstead BA (2008). Social comparison activity in coping with HIV. International Journal of STD & AIDS, 19(3), 164–167. [DOI] [PubMed] [Google Scholar]
  15. Diabetes Prevention Program (DPP) Research Group. (2002). The Diabetes Prevention Program (DPP): description of lifestyle intervention. Diabetes Care, 25(12), 2165–2171. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Elfhag K, & Rössner S. (2005). Who succeeds in maintaining weight loss? A conceptual review of factors associated with weight loss maintenance and weight regain. Obesity Reviews, 6(1), 67–85. [DOI] [PubMed] [Google Scholar]
  17. Festinger L. (1954). A theory of social comparison processes. Human Relations, 7, 117–140. [Google Scholar]
  18. Garcia SM, Tor A, & Schiff TM (2013). The psychology of competition: A social comparison perspective. Perspectives on Psychological Science, 8(6), 634–650. [DOI] [PubMed] [Google Scholar]
  19. Gibbons FX, & Buunk BP (1999). Individual differences in social comparison: development of a scale of social comparison orientation. Journal of Personality and Social Psychology, 76(1), 129–142. [DOI] [PubMed] [Google Scholar]
  20. Gorin AA, Powers TA, Koestner R, Wing RR, & Raynor HA (2014). Autonomy support, self-regulation, and weight loss. Health Psychology, 33(4), 332–339. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Harrison D, Marshall P, Bianchi-Berthouze N, & Bird J. (2015). Activity tracking: Barriers, workarounds and customisation. Proceedings of the 2015 ACM International Joint Conference on Pervasive and Ubiquitous Computing, 2015, 617–621. [Google Scholar]
  22. Helgeson VS, & Mickelson KD (1995). Motives for social comparison. Personality and Social Psychology Bulletin, 21(11), 1200–1209. [Google Scholar]
  23. Helgeson VS, & Taylor SE (1993). Social Comparisons and Adjustment Among Cardiac Patients 1. Journal of Applied Social Psychology, 23(15), 1171–1195. [Google Scholar]
  24. Hemphill KJ, & Lehman DR (1991). Social comparisons and their affective consequences: The importance of comparison dimension and individual difference variables. Journal of Social and Clinical Psychology, 10(4), 372–394. [Google Scholar]
  25. Kiernan M, Moore SD, Schoffman DE, Lee K, King AC, Taylor CB, … & Perri MG. (2012). Social support for healthy behaviors: scale psychometrics and prediction of weight loss among women in a behavioural program. Obesity, 20(4), 756–764. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Klein WM (1997). Objective standards are not enough: affective, self-evaluative, and behavioural responses to social comparison information. Journal of Personality and Social Psychology, 72(4), 763–774. [DOI] [PubMed] [Google Scholar]
  27. Leahey TM, Crowther JH, & Ciesla JA (2011). An ecological momentary assessment of the effects of weight and shape social comparisons on women with eating pathology, high body dissatisfaction, and low body dissatisfaction. Behavior Therapy, 42(2), 197–210. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Leahey TM, Kumar R, Weinberg BM, & Wing RR (2012). Teammates and social influence affect weight loss outcomes in a team-based weight loss competition. Obesity, 20(7), 1413–1418. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lean M, & Hankey C. (2018). Keeping it off: The challenge of weight-loss maintenance. The Lancet Diabetes & Endocrinology, 6(9), 681–683. [DOI] [PubMed] [Google Scholar]
  30. Lennarz HK, Lichtwarck-Aschoff A, Finkenauer C, & Granic I. (2017). Jealousy in adolescents’ daily lives: How does it relate to interpersonal context and well-being?. Journal of Adolescence, 54, 18–31. [DOI] [PubMed] [Google Scholar]
  31. MacLean PS, Wing RR, Davidson T, Epstein L, Goodpaster B, Hall KD, … & Rothman AJ. (2015). NIH working group report: Innovative research to improve maintenance of weight loss. Obesity, 23(1), 7–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Merchant G, Weibel N, Pina L, Griswold WG, Fowler JH, Ayala GX, … & Patrick K. (2017). Face-to-face and online networks: College students’ experiences in a weight-loss trial. Journal of Health Communication, 22(1), 75–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Mohr D, Cuijpers P, & Lehman K. (2011). Supportive accountability: a model for providing human support to enhance adherence to eHealth interventions. Journal of medical Internet Research, 13(1), e30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Muller D, & Fayant MP (2010). On being exposed to superior others: Consequences of self-threatening upward social comparisons. Social and Personality Psychology Compass, 4(8), 621–634. [Google Scholar]
  35. Nackers LM, Ross KM, & Perri MG (2010). The association between rate of initial weight loss and long-term success in obesity treatment: does slow and steady win the race?. International Journal of Behavioral Medicine, 17(3), 161–167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Oleander EK, Fletcher H, Williams S, Atkinson L, Turner A, & French DP (2013). What are the most effective techniques in changing obese individuals’ physical activity self-efficacy and behaviour: a systematic review and meta-analysis. International Journal of Behavioural Nutrition and Physical Activity, 10(1), 29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Sánchez CM, Dijkstra SC, & Visser M. (2015). Self-perception of body weight status in older Dutch adults. The Journal of Nutrition, Health & Aging, 19(6), 612–618. [DOI] [PubMed] [Google Scholar]
  38. Rancourt D, Leahey TM, LaRose JG, & Crowther JH (2015). Effects of weight-focused social comparisons on diet and activity outcomes in overweight and obese young women. Obesity, 23(1), 85–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Schokker MC, Keers JC, Bouma J, Links TP, Sanderman R, Wolffenbuttel BH, & Hagedoorn M. (2010). The impact of social comparison information on motivation in patients with diabetes as a function of regulatory focus and self-efficacy. Health Psychology, 29(4), 438. [DOI] [PubMed] [Google Scholar]
  40. Singer JD, Willett JB, & Willett JB (2003). Applied longitudinal data analysis: Modeling change and event occurrence. Oxford University Press. [Google Scholar]
  41. Van der Zee K, Oldersma F, Buunk BP, & Bos D. (1998). Social comparison preferences among cancer patients as related to neuroticism and social comparison orientation. Journal of Personality and Social Psychology, 75(3), 801–810. [DOI] [PubMed] [Google Scholar]
  42. Van der Zee K, Buunk B, Sanderman R, Botke G, & Van den Bergh F. (2000). Social comparison and coping with cancer treatment. Personality and Individual differences, 28(1), 17–34. [Google Scholar]
  43. Varkevisser RDM, van Stralen MM, Kroeze W, Ket JCF, & Steenhuis IHM (2019). Determinants of weight loss maintenance: A systematic review. Obesity Reviews, 20(2), 171–211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Wadden TA, & Foster GD (2006). Weight and lifestyle inventory (WALI). Obesity, 14(S3), 99S–118S. [DOI] [PubMed] [Google Scholar]
  45. Wills TA (1981). Downward comparison principles in social psychology. Psychological Bulletin, 90(2), 245–271. [Google Scholar]
  46. Wood JV (1989). Theory and research concerning social comparisons of personal attributes. Psychological bulletin, 106(2), 231–248. [Google Scholar]
  47. Wood JV, Taylor SE, & Lichtman RR (1985). Social comparison in adjustment to breast cancer. Journal of Personality and Social Psychology, 49(5), 1169–1183. [DOI] [PubMed] [Google Scholar]
  48. Wood JV, & Wilson AE (2003). How important is social comparison? In Leary MR & Tangney JP (Eds.), Handbook of self and identity (p. 344–366). The Guilford Press. [Google Scholar]
  49. Zell E, & Alicke MD (2010). The local dominance effect in self-evaluation: Evidence and explanations. Personality and Social Psychology Review, 14(4), 368–384. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix
NIHMS1771595-supplement-Appendix.docx (23.7KB, docx)

RESOURCES