Elsevier

Appetite

Volume 143, 1 December 2019, 104403
Appetite

Does Health At Every Size® fit all? A group-based trajectory modeling of a non-diet intervention

https://doi.org/10.1016/j.appet.2019.104403Get rights and content

Abstract

Objective

Health At Every Size® (HAES®) interventions have been increasingly recognized as a sustainable strategy in obesity management. Nevertheless, heterogeneity among obese individuals leads to challenges as it translates in mixed responses to treatment. In this context, our objective was to identify trajectories of responses to a non-diet intervention for adult overweight/obese women to highlight profiles of responders.

Method

Based on data from a multicentric quasi-experimental study, a latent class growth modeling (LCGM) was performed. Two hundred and ten women with high body mass index (BMI ≥ 25, MBMI = 36.53) who followed a non-diet intervention offered in Health and Social Services Centres completed questionnaires at T = 0, 4 and 16 months. Outcomes used in the LCGM were intuitive eating and body esteem, two central components in HAES® interventions. Types of responders were then profiled on sociodemographic, weight, lifestyle, psychological and eating variables.

Results

The LCGM revealed a 4-trajectory model (p < .001), comprising non-responders (14.67%), moderate improvement with low maintenance responders (49.89%), moderate improvement with high maintenance responders (29.28%) and high functioning partial responders (6.56%). Analysis of variances showed significant differences between all types of responders with medium to large effect sizes on depressive symptoms, self-esteem and disinhibited eating (p < .001; η2 = .23, 0.30 and 0.16 respectively). Fewer differences were found on sociodemographic, lifestyle, health and weight variables. Overall, non-responders (14.67%) had a distinctive profile compared to the other groups by consistently expressing poorer psychological functioning, less adapted eating behaviors and reaching more frequently the clinical cutoff for severe depression (p = .001).

Conclusions

Findings strongly support the relevance of considering psychological characteristics to move towards personalized healthcare in obesity management.

Introduction

Obesity is widely known as a public health issue that accounts for quality of life impairments and many serious health conditions, such as cardiovascular diseases (GBD 2015 Obesity Collaborators, 2017; World Health Organization, 2018). Meanwhile, it is also associated with psychological distress, body dissatisfaction, stigmatization and depression, another leading cause of disease burden worldwide (Luppino et al., 2010; Petry, Barry, Pietrzak, & Wagner, 2008; Puhl & Heuer, 2010; Schwartz & Brownell, 2004). Among the several strategies developed in obesity management, the Health At Every Size® (HAES®) movement stands out as a non-diet approach, with an emphasis on self-acceptance and healthy day-to-day practices. While they acknowledge health risks associated with higher BMI, HAES® practitioners advocate the idea that health relies on a wide and more complex range of factors, not solely on body mass index (BMI), and can be improved without necessarily losing weight Bacon & Aphramor, 2011, Gaesser, Angadi, & Sawyer, 2011, Tomiyama et al., 2013. HAES® interventions aim to reduce weight stigma, which threatens both psychological and physical health (Puhl & Heuer, 2010), while also promoting healthy eating habits, such as intuitive eating and physical activity (Burgard, 2009, pp. 42–53; Robison, 2007a; 2007b). In doing so, the HAES® movement addresses simultaneously health issues associated with high BMI as well as the excessive preoccupations towards food, weight and body image among women, who are significantly targeted by societal ideals of thinness (Burgard, 2009, pp. 42–53; Saguy, 2012, Robison, 2007b). According to a recent systematic review, non-diet interventions seem to display rather positive outcomes on many levels, including psychological well-being, eating behaviors and physical health (Clifford et al., 2015; Ulian et al., 2018). They also seem to stabilize weight or even produce weight loss, though results are still inconsistent (Clifford et al., 2015). As such, it has been suggested to implement this approach as a public health policy since it shows promising results while doing no harm (Bombak, 2014; Thomas, 2006).

It is well established that obesity interventions, regardless of the approach, produce great variability in responses to treatment (Brownell & Wadden, 1991; Field, Camargo, & Ogino, 2013; Gordon-larsen, 2018; Yang, Ginsburg, & Simmons, 2013). This leads evidently to significant clinical challenges for practitioners. Several studies have pointed out the heterogeneity of obese/overweight individuals as a factor that should be acknowledged to tailor interventions to the patients’ needs and, as a result, have identified different subtypes based on either personality, psychological symptoms, eating patterns, metabolic health or genetics (Claes, Vandereycken, Vandeputte, & Braet, 2013; Field et al., 2018; Gagnon-Girouard, Bégin, Provencher, Tremblay, Boivin, et al., 2010a; Gullo, Lo; Jansen, Havermans, Nederkoorn, & Roefs, 2008; Müller, Claes, Wilderjans, & de Zwaan, 2014; Wang et al., 2009). For instance, evidence suggests that individuals with high depressive symptoms profile would display more body concerns and problematic eating behaviors despite having no significant difference in BMI with their non-depressive counterparts (Gagnon-Girouard et al., 2010a; Jansen et al., 2008). Heterogeneity in psychological functioning could also reflect on distinctive metabolic profiles, as some recent studies have suggested the existence of a mood-metabolic syndrome that could represent a more severe subtype of obesity (Mansur, Brietzke, & Mcintyre, 2015; Phillips & Perry, 2015). These findings suggest the unlikelihood of achieving similar outcomes from all in response to a given treatment. A better understanding of individual characteristics could certainly improve outcomes by enhancing personalized healthcare (Brownell & Wadden, 1991; Field et al., 2013; Kiernan, 2012; van der Merwe, 2007; Yang et al., 2013).

Nevertheless, little attention has been given to exploring the different types of responses among participants. Up until now, most research emphasizes on assessing the efficacy or effectiveness of interventions by comparing group means over time, thus adopting a one-size-fits-all philosophy in which interventions are compared to each other using strictly mean group outcomes (Field et al., 2013; Yang et al., 2013). Since this can lead to biased conclusions – either underestimating or overestimating effect sizes, it is thus important not only to examine the overall outcomes of an intervention, but also to look how individuals respond to it.

Notwithstanding this need for more individual-oriented research, it should be mentioned that many researchers have put efforts to identify pre-treatment predictors of outcomes with the intent of matching individuals to different treatments (Teixeira, Going, Sardinha, & Lohman, 2005). However, these studies have several statistical and methodological shortcomings that have been raised (Stubbs et al., 2011; Teixeira et al., 2005). Among them, the potentially large number of variables involved that interact with each other in complex ways and the use of too simple statistical analyses have been suggested as potential limitations (Teixeira, Going, Sardihna & Lohman, 2005). In this regard, the use of mixture modeling (e.g. latent class analysis, latent profile analysis, latent class growth modeling) comes in handy for identifying subtypes with similar characteristics within a heterogeneous population.

So far, very few studies have proceeded to both identify subtypes and associate them with treatment outcomes. To our knowledge, only one large-scale study in the field of obesity was conducted using such an analytic approach with a broad range of variables, although it was performed in a context of weight loss after a bariatric surgery (Field et al., 2018). In her commentary article, Gordon-larsen (2018) clearly points out the need to address heterogeneity in a variety of obesity-related treatments. Along the same lines, it has been suggested that outcomes other than weight loss should be examined in obesity management, as many variables account for health-related effects (Teixeira, Going, Sardihna & Lohman, 2005). In the more specific case of the HAES® approach, no study has yet explored these research questions and it remains unknown whether these types of intervention produce benefits in a similar manner across all treatment-seeking overweight/obese individuals or not. As well, no study has yet provided a thorough characterization of these individuals in such a way that could help better understand for whom the HAES® approach could be most appropriate.

This study first aims to unravel heterogeneity in response to core outcomes of a community-based HAES® intervention for women by identifying several types of responders using group-based trajectories. As its effectiveness has been assessed several times by comparing outcomes of the experimental group to a control group, both in controlled and natural environments (specifically in the cohort of women we are studying here),(Bégin et al., 2018; Carbonneau et al., 2016; Gagnon-Girouard, Bégin, Provencher, Tremblay, Mongeau, et al., 2010b; Provencher et al., 2007; Provencher et al., 2009), this research focuses exclusively on exploring types of responses within the intervention group. Secondly, it aims to profile the given types of responders on sociodemographic, weight, health, psychological, eating and lifestyle variables. It is expected that women would neither respond nor maintain their gains in the same way. Similarly, we also expected different starting points on outcomes that reflects on the heterogeneity of the obese/overweight population. Thus, the existence of multiple trajectories differing in intercepts and slope magnitude or direction was hypothesized. More particularly, we expected to identify at least two types of responders, comprising a more responsive and a less responsive subtype. Given the literature revealing the existence of more severe profiles with poorer psychological functioning and more disordered eating (Gagnon-Girouard et al., 2010a; Jansen et al., 2008), it is expected that such profiles would be associated with a less responsive type of responders to the non-diet intervention.

Section snippets

Research design and overview of procedures

Data were extracted from a multicentric quasi-experimental study aiming to evaluate a non-diet intervention “Choisir de maigrir?” (CdM?) (What about losing weight?) compared to a waiting-list control group. Conducted in a real-world setting, twenty-one Health and Social Services Centres (HSSC) from different areas of the province of Quebec (Canada) were entirely in charge of recruitment and data collection. Participants were recruited during the fall or winter sessions of the program in 2010

Descriptive statistics

Participants were aged from 21 to 83 years-old (M = 51.26, SD = 11.17). The sample consisted of individuals presenting a wide range of socioeconomic status, from which a considerable percentage came from a lower socioeconomic environment. Participants’ sociodemographic characteristics at baseline are presented in Table 1. Differences between completers and dropouts were found only on education levels.

Group-based trajectory modeling on outcomes

The LCGM revealed a significant 4-trajectory model (p < .001). The 4 trajectories were

Discussion

This study investigated the heterogeneity of responses among women participating in a HAES® intervention by modeling group-based trajectories, thus allowing the categorization of participants into types of responders, which were then profiled on sociodemographic, weight, health, lifestyle, psychological and eating variables. To our knowledge, this is the first study to investigate for whom the non-dieting framework of intervention works best when considering individual characteristics, using an

Conclusions

The current study is, to our knowledge, the first to examine for whom a community-based non-diet intervention fits. This research has successfully identified several trajectories that were categorized as four types of responders. While most participants displayed some improvement on either or both intuitive eating and body esteem, a non-responder group was identified. The latter displayed the more negative profile of all on psychological, eating and health variables, being markedly more

Submission declaration

The manuscript is original and not under any consideration for publication at another journal. Its publication is approved by all contributing authors.

Sources of funding

This research project was funded by the Canadian Institutes of Health Research (CIHR) GIR 99712 operating grant (Population Health Intervention Research) and by the Heart and Stroke Foundation.

Conflicts of interest

The authors of this manuscript do not have any conflicts of interest to declare.

References (60)

  • A.J. Stunkard et al.

    The three-factor eating questionnaire to measure dietary restraint, disinhibition and hunger

    Journal of Psychosomatic Research

    (1985)
  • J. Westenhoefer

    Dietary restraint and disinhibition: Is restraint a homogeneous Construct ?

    Appetite

    (1991)
  • K.G.M.M. Alberti et al.

    Metabolic syndrome - a new world-wide definition. A consensus statement from the International Diabetes Federation

    Diabetic Medicine

    (2006)
  • H. Andruff et al.

    Latent class growth modelling: A tutorial

    Tutorials in Quantitative Methods for Psychology

    (2009)
  • L. Bacon et al.

    Bacon, L., & Aphramor, L. (2011). Weight science: evaluating the evidence for a paradigm shift. Nutrition journal, 10(1), 9.ISO 690

    Nutrition journal

    (2011)
  • A.T. Beck

    Cognitive models of depression

    Clinical advances in cognitive psychotherapy: Theory and application

    (2002)
  • A.T. Beck et al.

    Manual for the Beck Depression Inventory-II

    (1996)
  • C. Bégin et al.

    Eating-related and psychological outcomes of health at every size intervention in health and social Services Centers across the province of québec

    American Journal of Health Promotion

    (2018)
  • M.I. Berman et al.

    Health at every size and acceptance and commitment therapy for obese, depressed women: Treatment development and clinical application

    Clinical Social Work Journal

    (2015)
  • A. Bombak

    Obesity, health at every size, and public health policy

    American Journal of Public Health

    (2014)
  • A.J. Bradshaw et al.

    Non-dieting group interventions for overweight and obese women: What predicts non-completion and does completion improve outcomes?

    Public Health Nutrition

    (2010)
  • D. Burgard

    What is “health at every size?

  • E. Carbonneau et al.

    A Health at Every Size intervention improves intuitive eating and diet quality in Canadian women

    Clinical nutrition

    (2016)
  • L. Claes et al.

    Personality subtypes in female pre-bariatric obese patients: Do they differ in eating disorder symptoms, psychological complaints and coping behaviour?

    European Eating Disorders Review

    (2013)
  • D. Clifford et al.

    Impact of non-diet approaches on attitudes, behaviors, and health outcomes: A systematic review

    Journal of Nutrition Education and Behavior

    (2015)
  • P.J. Curran et al.

    Twelve frequently asked questions about growth curve modeling

    Journal of Cognitive Development

    (2010)
  • L. Dubois et al.

    The choice of a diet quality indicator to evaluate the nutritional health of populations

    Public Health Nutrition

    (2000)
  • A.E. Field et al.

    The merits of subtyping obesity one size does not fit all

    Journal of the American Medical Association

    (2013)
  • A.E. Field et al.

    Association of obesity subtypes in the longitudinal assessment of bariatric surgery study and 3‐year postoperative weight change

    Obesity

    (2018)
  • G.A. Gaesser et al.

    Exercise and diet, independent of weight loss, improve cardiometabolic risk profile in overweight and obese individuals

    The Physician and sports medicine

    (2011)
  • Cited by (6)

    • Latent profile analysis as a method for process evaluations: Discovering response subgroups in a mindfulness intervention

      2022, Social Science and Medicine
      Citation Excerpt :

      Finally, the sample of other-language speakers could have been larger to make the findings more robust. Using classes for tailoring interventions represents a step towards more targeted intervention delivery, which is a method conducive to personalizing healthcare (Cloutier-Bergeron et al., 2019). They may be particularly suitable when an individualized approach is not feasible, which may very well be the case in many school-based mindfulness interventions.

    • The weight of words: Discursive constructions of health in weight-neutral peer-reviewed journal articles

      2022, Body Image
      Citation Excerpt :

      The term ‘obesity management’ is paradoxical in a weight-neutral context, as the focus of these approaches is to shift focus away from weight-loss as a direct (rather than incidental) health outcome. Cloutier-Bergeron et al. (2019, p. 1) also use paradoxical language by suggesting that “(HAES®) interventions have been increasingly recognized as a sustainable strategy in obesity management” and Carrol et al. (2007, p. 125) describe the HAES® approach as a “paradigm for obesity treatment” (Carroll et al., 2007, p. 125). This contradictory language either indicates a misunderstanding of the original purpose of HAES®, unfamiliarity with the HAES® model or a desire to integrate the practices of opposing weight-focused and weight-neutral approaches to health.

    • Intuitive eating buffers the link between internalized weight stigma and body mass index in stressed adults

      2022, Appetite
      Citation Excerpt :

      Additionally, it is important to note that different approaches will likely be needed for different individuals. Within the context of HAES and intuitive eating interventions some individuals are less responsive (e.g., Cloutier-Bergeron et al., 2019). Recent EMA findings also suggest that dietary restriction may be most challenging for those who experience heightened stress or emotional eating (Reichenberger et al., 2019).

    View full text