Characterizing Patterns of Adherence to Physical Activity Goals in Behavioral Weight Control

Introduction: Many individuals fall short of the physical activity (PA) goals provided during lifestyle programs, and there is considerable variability in weight loss response and adherence to recommendations for behavior change. Weight loss response within the early weeks of program initiation predicts future weight loss success; however, little is known about what distinguishes those who respond well to treatment from those who fail to respond. Exposing the behavioral patterns which underlie weight loss response may point towards targets for treatment optimization. The purpose of this study was to (1) characterize individuals with distinct patterns of adherence to PA goals during the first 2 months of an online behavioral weight loss intervention and determine whether these patterns were associated with 6-month weight loss and (2) describe weekly transitions in adherence to steps-based and minutes-based PA goals and self-monitoring behavior across 6 months of intervention. Methods: Individuals participating in an online behavioral weight loss intervention were prescribed progressive goals for 7,000 to 10,000 daily steps and 50 to 200 minutes of weekly moderate-to-vigorous PA (MVPA). Goals increased incrementally during the initial two months (graded goal phase) and then stayed constant for the remainder of the program (fixed goal phase). Participants were instructed to self-monitor their total steps and minutes of MVPA daily on the study website. Individuals were classified each week as meeting the program goals for steps (yes/no) and minutes of MVPA (yes/no) based on self-monitoring records. A repeated-measures latent class analysis was conducted to identify distinct subgroups during the initial 2 months. Associations between latent classes and sociodemographic characteristics, weight loss outcomes, and treatment engagement parameters were then examined. Further, individuals were categorized as (1) “adherent”, (2) “suboptimally adherent”, or (3) “nonadherent” to the steps-based goals and separately to the minutes-based goals, based on whether they met the weekly goal (yes/no) and self-monitored their physical activity (yes/no). Weekly transition probabilities were examined across 6 months using multinomial logistic regression. Results: Participants (N=212) were predominantly female (91.5%), middle-aged (mean = 47.9 ± 11.1 years) with an average BMI of 35.8 ± 5.9 kg/m2, and 31.6% self-identified as a racial/ethnic minority. Three latent classes describing patterns of physical activity goal adherence were revealed: 27.3% had a high probability of consistently meeting both steps-based and minutes-based targets (“Both PA Goals”), 28.4% were estimated to meet the weekly MVPA goals but not the step goals (“MVPA Goals Only”), and the largest class (44.3%) had a low probability of attaining either weekly goal (“Neither PA Goal”). Weight losses at 6 months were significantly greater for members of the “Both PA Goals” class (estimated mean weight loss [95% CI]: -9.4% [7.4, 11.5]) compared to “MVPA Goals Only” (-4.8% [3.4, 6.1]) and “Neither PA Goal” (-2.5% [1.4, 3.6]). Factors differentiating these classes included age, gender, race/ethnicity, geographic region, and BMI at study entry, as well as engagement in other treatment components, such as group session attendance and self-monitoring of weight, dietary intake, and physical activity. The most consistent status among both types of goals was “nonadherent,” while “suboptimally adherent” was most vulnerable to movement, particularly during the graded goal phase. Few individuals were initially “adherent” to steps goals and had a high probability of either remaining “suboptimally adherent” from one week to the next or transitioning to a “nonadherent” status. Conversely, most individuals started out “adherent” to MVPA goals but were more likely to become less adherent for MVPA goals than steps goals during the graded goal phase (p=.0042). Across the fixed goal phase, the pattern of weekly adherence status did not differ significantly between steps-based versus minutes-based goals (p=.6530). Conclusion: Distinct patterns of adherence to weekly program PA goals underscore the early period of behavioral weight control as a critical window of opportunity to detect individuals at risk of poor treatment outcomes. This study provides novel insights about behavioral typologies in obesity treatment as they relate to physical activity adherence and contributes a preliminary framework for identifying individuals in need of additional or different support for successfully adopting physical activity during lifestyle modification