Predicting program attendance and weight loss in obesity interventions: Do triggering events help?

Abstract

Medical events that “trigger” motivation to lose weight may improve treatment outcomes compared to non-medical or no triggering events. However, previous findings include only long-term successful participants, not those initiating treatment. The current study compared those with medical triggering events or non-medical triggering events to no triggering events on attendance and weight loss during a weight management program. Medical-triggering-event participants lost 1.8 percent less weight (p = 0.03) than no-triggering-event participants. Non-medical-triggering-event participants attended 1.45 more sessions (p = 0.04) and were 1.83 times more likely to complete the program (p = 0.03) than no-triggering-event participants. These findings fail to support the benefit of medical triggering events when beginning treatment for obesity.

Introduction

For patients seeking treatment for obesity, 50 to 72 percent of individuals listed health as their primary motivation for losing weight (Holley et al., 2016; O’Brien et al., 2007; Yoong et al., 2013). For some, this motivation is triggered by a medical event or health crisis that spurs behavior change (Epiphaniou and Ogden, 2010; Gorin et al., 2004; Ogden and Hills, 2008). Medical triggering events (MTEs) for weight loss could include acute events (e.g. myocardial infarction, diagnosis of diabetes) and/or ongoing chronic conditions or disease progression (e.g. consistently elevated glucose values, increased medication dosage) that prompt a desire for change.

Findings from the National Weight Control Registry (NWCR), a cohort of individuals who successfully lost and maintained ⩾10 percent of their body weight for ⩾1 year(s), suggest that those with MTEs lost more and regained less weight at follow-up than those with non-medical triggering events (NMTEs) or no triggering events (NTEs) (Gorin et al., 2004). Thus, MTEs may provide a “teachable moment” that leads to successful initial and long-term weight control.

While the NWCR provides longitudinal data from a large national cohort of individuals who achieved and maintained clinically significant weight loss, data are based on retrospective self-reports from a homogeneous (e.g. 94% Caucasian) and uniquely successful weight loss group (Gorin et al., 2004; LaRose et al., 2013). Given that the vast majority of those who initiate behavioral weight loss treatments are unsuccessful (or significantly less successful) than the NWCR cohort, this limits the generalizability of previous findings regarding MTEs. In contrast to these findings, medical reasons are a prominent factor in participant dropout (Zizzi et al., 2016), and treatments emphasizing medical or health reasons over appearance reasons for weight loss have actually been less successful in short-term weight loss (Kalarchian et al., 2011). Thus, it remains unknown whether MTEs confer benefit for those initially engaging in weight loss efforts.

The current study sought to evaluate whether triggering events, specifically MTEs, result in different initial treatment outcomes in comparison with NMTEs or NTEs among a racially diverse sample enrolled in a behavioral weight loss program. Prior to treatment, participants reported the presence of different triggering events motivating their current weight loss efforts and categorized these events as MTEs or NMTEs. Treatment outcomes included attendance and weight loss. It was hypothesized that those with MTEs would show better rates of attendance and greater weight loss compared to those with NMTEs or NTEs.

Methods

Participants and procedure

This study was conducted as part of the “Improving Weight Loss” study and approved by the academic health center’s institutional review board and in accordance with the 1964 Helsinki Declaration and subsequent amendments. Primary outcomes and methods for the study have been described elsewhere (Dutton et al., 2017). Importantly, the previous study evaluated weight loss maintenance only for those who had achieved >5 percent weight loss, whereas this study utilizes all participant data. The study recruited and consented 305 adults with obesity who participated in a 16-week, group-based behavioral weight loss intervention modeled after the Diabetes Prevention Program (DPP, 2002). Participants were excluded if they were below 21 years of age, had lost >4.5 kg or used weight loss medications in the past 6 months, planned to relocate from the area in the next 18 months, were unwilling to attend treatment sessions, or had a body mass index (BMI, kg/m²) <28 or >45. For their safety, individuals with severe medical events, such as myocardial infarction or stroke, were also excluded from participation.

Measures

Primary outcomes included percentage weight change and attendance (defined as the number of treatment sessions attended). Triggering events were assessed on a questionnaire identical to the NWCR survey asking participants whether they had experienced “a specific incident, or triggering event, that prompted you to begin this weight loss program?” Individuals responding “yes” then selected from seven events (e.g. “Saw image or picture of myself,” “inspiration from another person”) and described the specifics of the event (e.g. “to inspire my friend,” “recent tornado”). One response choice included a “medical event,” and participants were also asked to describe specifics of this event (“high cholesterol,” “recent breast cancer and heart trouble in my family”). Like the NWCR survey, participants were not required to specify the exact nature of any triggering event, and participants chose which category best described their triggering event. Any participants endorsing multiple triggering events (12.5% of individuals) were categorized in the MTE group if any of their triggers were medical; otherwise, they were categorized as having a NMTE. Covariates of interest included race, gender, education, income, marital status, baseline weight status, self-reported perceived health (assessed on a 5-point Likert-type scale, poorer health indicated by higher score: 1 = excellent, 5 = poor), and chronic, weight-related medical conditions. Assessed medical conditions included hypertension, diabetes, chest pain, myocardial infarction, stroke, high cholesterol, asthma or lung disease, dizziness or fainting spells, stroke, sleep apnea, arthritis, and pre-diabetes.

Analytic strategy

Linear regression analyses were conducted to evaluate whether triggering events (i.e. MTE and NMTE) significantly predicted weight loss and treatment session attendance compared to NTE. Participant characteristics (e.g. race, sex) demonstrating significant correlations with these outcomes were included as covariates in subsequent models. Additional models further included perceived health and medical conditions to better understand the relationship between triggering events and treatment outcomes.

Results

Table 1 summarizes characteristics of the sample. Only baseline weight, age, and race were significantly related to treatment outcomes and were therefore retained in regression models. Session attendance was significantly associated with NMTEs (n = 127, 41.8% of participants), such that participants with NMTEs attended 1.4 more sessions on average than those with NTEs (p = 0.0476), while attendance of participants with MTEs (n = 77, 25.33% of participants) did not differ significantly from those reporting NTE (p = 0.0541). In contrast, participants with MTEs lost 1.8 percent less weight on average compared to participants with NTEs (p = 0.039). Participants with NMTEs did not differ from those with NTEs with regard to weight loss (p > 0.05, see Table 2).

Table 1.

Participant characteristics.

Characteristics	Total sample (N = 304)
Baseline BMI, kg/m2	35.82 (4.57)
4-month BMI, kg/m2	33.33 (4.60)
Age, years	47.97 (12.44)
Perceived health scorea	2.54 (0.78)
Comorbid medical conditionsb	1.46 (1.42)
Gender, female (%)	277 (91.12)
Race, non-white (%)	185 (61.06)
Education (%)
Associates or lower degree	128 (42.11)
Bachelors or higher degree	176 (57.89)
Income (%)
⩽US$40,000	104 (34.32)
US$40,000-US$80,000	122 (40.26)
⩾US$80,000	77 (25.41)

BMI: body mass index.

Data are given as mean (SD) unless otherwise indicated. All demographics were collected at baseline.

^aPerceived health was rated on a 5-point Likert-type scale (1 = excellent, 5 = poor).

^bComorbid medical conditions included blood pressure, high cholesterol, stroke, dizziness, pre-diabetes, diabetes, gestational diabetes, asthma, heart attack, sleep apnea, and chest pain.

Table 2.

Predictors of percentage weight loss (Model 1).

Variables	Regression coefficient	Standard error	Pr > |t|
Intercept	−0.071	0.029	0.018
Medical triggers	0.018	0.009	0.039
Non-medical triggers	0.004	0.008	0.571
Baseline weight	0.000	0.000	0.303
Age	0.000	0.000	0.135
Race (non-white vs white)	0.009	0.007	0.186

Medical triggering events and non-medical triggering events are compared to no triggering events. Bold values indicate statistical significance.

Given the unexpected finding that MTEs were predictive of poorer weight loss, additional analyses explored other factors potentially relevant to this association. In particular, participants with MTEs reported significantly more chronic medical conditions (M = 2.2, SD = 1.4) than those with NTEs (M = 1.2, SD = 1.2) or NMTEs (M = 1.2, SD = 1.4) (p < 0.0001). Also, participants with MTEs (M = 2.8, SD = 0.7) had worse perceived health compared to those with NMTEs (M = 2.4, SD = 0.8) and NTEs (M = 2.4, SD = 0.8, p = 0.0003). In light of these findings, medical conditions and perceived health were included in subsequent models. Medical conditions were associated with 0.534 percent less weight loss per reported condition (p = 0.039), and each unit increase in perceived health score (indicating worse perceived health) was associated with 1.2 percent greater weight loss (p = 0.0072); MTEs were no longer significantly associated with weight loss after adjustment for these variables, although the association still trended in the same direction (0.0164, p = 0.0662) (Table 3). Attendance remained associated with NMTEs after accounting for perceived health and medical conditions, such that participants with NMTEs attended 1.4 more sessions on average compared to those with NTE (p = 0.0482).

Table 3.

Predictors of percentage weight loss (Model 2).

Variables	Regression coefficient	Standard error	Pr > |t|
Intercept	−0.030	0.032	0.348
Medical triggers	0.016	0.009	0.066
Non-medical triggers	0.002	0.008	0.833
Baseline weight	0.000	0.000	0.352
Age	−0.001	0.000	0.017
Race (non-white vs white)	0.009	0.007	0.174
Medical conditions	0.005	0.003	0.039
Perceived health	0.012	0.005	0.007

Medical triggering events and non-medical triggering events are compared to no triggering events. Bold values indicate statistical significance.

Discussion

In contrast to previous findings from the NWCR (Gorin et al., 2004), current results indicate that MTEs were not beneficially associated with initial weight loss. In fact, current results suggest that MTEs may modestly hinder weight loss efforts among individuals initially seeking treatment for obesity. It may be that findings from a highly successful cohort such as the NWCR do not generalize well to other samples, or perhaps MTEs confer benefit for weight loss maintenance more so than initial weight loss. Regardless, these findings suggest that medical triggers may not provide an ideal “teachable moment” to initiate weight loss efforts for some individuals.

While MTEs were associated with slightly less weight loss, they were not related to program attendance. In addition, while MTEs were related to attenuated weight loss, this association was no longer significant after accounting for participants’ perceived health status and reported weight-related medical comorbidities. Interestingly, while medical conditions predicted less weight loss, perception of poorer health predicted greater weight loss. These findings suggest that one’s overall perception of health and the number of chronic conditions requiring management by the individual may be more salient factors in one’s weight loss efforts than an acute or chronic medical event. This is consistent with recent findings indicating that a greater number of chronic medical conditions were associated with unhealthy behaviors and more barriers to behavior change among individuals with obesity (Bastin et al., 2019). In addition, one’s perceived health may serve to produce behavior change in a way that actual medical events and conditions do not. However, this assertion warrants further research.

This study has several limitations. Importantly, individuals with severe medical events, such as myocardial infarction or stroke, were excluded from participation in the behavioral intervention due to safety concerns, which may have artificially deflated the impact of MTEs. In addition, several individuals (approximately 12%) listed multiple triggering events, making it difficult to determine the most relevant or salient event (medical or non-medical) that may have prompted weight loss efforts.

Strengths of this study include a racially diverse sample and a prospective design evaluating the impact of triggering events identified at baseline on objectively measured treatment engagement and weight loss outcomes. In addition, these findings add to previous research focused on MTEs in relation to weight loss maintenance by examining events for initial weight loss efforts. Future research should evaluate whether perceived health and comorbid conditions may influence perceptions of triggering events, and whether MTE severity affects weight loss motivation and success.

Conclusion

These results provide evidence that MTEs may not confer benefit for initial weight loss efforts as previously suggested. Other variables, such as perceived health and burden from comorbid conditions, may be more salient than MTEs in impacting initial obesity treatment response. It may be that one’s perception of an event’s severity or relevance in relation to overall health, rather than the event itself, activates motivation for change that is necessary for initial weight loss success.