key: cord-0857982-626xcbdc authors: Bullard, Tiffany; Medcalf, Adam; Rethorst, Chad; Foster, Gary D. title: Impact of the COVID‐19 pandemic on initial weight loss in a digital weight management program: A natural experiment date: 2021-07-30 journal: Obesity (Silver Spring) DOI: 10.1002/oby.23233 sha: 0028c368fcdb71bcffcc44f72d83108cf0f78091 doc_id: 857982 cord_uid: 626xcbdc OBJECTIVE: The aim of this study was to assess the impact of the coronavirus disease 2019 (COVID‐19) pandemic on initial weight loss during a digital weight management program. METHODS: Participants (n = 866,192; BMI 33.6 [SD 7.4] kg/m(2)) who joined a digital weight management program (WW) in the first 30 weeks of 2020 (COVID‐19 cohort) were compared with participants (n = 624,043; BMI 33.1 [SD 7.2] kg/m(2)) who joined the same program during the same time period in 2019 (control cohort). Weight change (percentage) and self‐monitoring over the first 4 weeks of enrollment were compared between the cohorts. Significance was defined as meeting the criteria for a small effect (d ≥ 0.2). RESULTS: Over the 30‐week enrollment period, the COVID‐19 cohort experienced significantly less weight loss than the control cohort but only for 7 weeks of enrollments. The COVID‐19 cohort also had fewer days of food tracking but only for 3 weeks of enrollments. There were no differences in the self‐monitoring of weight and activity at any time between the two cohorts. CONCLUSIONS: Over a 30‐week enrollment period, COVID‐19 had negative effects on both weight loss and food self‐monitoring, but the effects were short‐lived. Those participating in evidence‐based weight management programs can expect similar levels of initial weight loss as those experienced before the pandemic. Participants were WW digital members in the United States who joined during the first 30 weeks of 2020 as well as those who joined during the same time period in 2019. Participant characteristics are described in Table 1 . Weight change (percentage) and selfmonitoring (in-app tracking of weight, food, and activity) data were assessed during the first 4 weeks of WW membership. Four-week weight loss was chosen as the outcome because weight loss during the first month of a behavioral weight loss program is predictive of long-term weight loss (6) (7) (8) (9) . All data were self-reported as entered into the WW app. Data from weekly cohorts were analyzed independently (i.e., weight loss for members joining in Week 1 of 2020 was compared with weight loss for members joining in Week 1 of 2019). Members who logged their weight during their first week of their membership and entered a subsequent weight in weeks 3-5 were included in the analysis. Weight from week 4 was used as the follow-up weight, when available. If a member had missing weight data for week 4, weight data from week 5 were used. If neither were available, weight data from week 3 were used. If weight was not reported in weeks 3 to 5, then the member was excluded from the analysis. The data included in the analyses were normally distributed. Weight loss outliers were identified and removed using thresholds calculated on weight loss percentage: low (25th weight change % percentile − 1.5*interquartile range [IQR]) and high (75th weight change % percentile + 1.5*IQR). Using these criteria, only 2.5% of the data were considered weight loss outliers, equating to approximately ≤−9.3% weight loss and ≥3.3% weight gain in 4 weeks. We compared the excluded outliers sample to the remaining sample and found some differences, including outlying members on average were 3 years younger than nonoutliers (42.5 years old vs. 45.5), more likely to be male (14.5% vs. 11.0%), and had higher start weights and BMIs, likely due to misentering weights (11.9 kg heavier at first weigh in and 3.5kg/m 2 higher BMI). Because of the large sample sizes, p values were not calculated. Instead, significant differences were defined as meeting the criteria for Cohen's convention for a small effect (10) (d ≥ 0.2). Over the first 30 weeks of enrollment, the COVID-19 cohort had In this natural experiment of almost 1.5 million enrollees in a digital WW program, we examined the effects of COVID-19 on initial weight loss and self-monitoring behavior during 30 weeks of enrollment in 2020 (COVID-19 cohort) and 2019 (control cohort). There were several principal findings. The first was that COVID-19 produced significantly less weight loss but its effects were short-lived. Weight loss was less among the ► These data suggest that initial weight loss success was compromised only during the initial period of the pandemic. They also suggest that those participating in evidence-based weight management programs can expect similar levels of initial weight loss as those experienced before the pandemic. precise reasons for the decrease relative to prepandemic weight losses are difficult to identify. However, it is likely that the conditions associated with the earlier stages of the COVID-19 outbreak (stress, disruption of routines, more proximal access to food, social isolation, closure of in-person gyms) served as barriers to effective weight management. In support of that hypothesis, recent studies examining changes in weight (5,11) after shelter-in-place orders report that on average, people gained ~1.5 kg per month during this time period. The weight loss patterns observed in our study do show that by the end of April, initial weight loss returned to 2019 prepandemic levels. These patterns suggest that at the start of the strict stay-at-home order time period, members may have paused their health-related routines and over time, began to adapt to the "new normal," focusing on their health and weight-related goals again. They also suggest that participants in evidence-based weight management programs can expect similar levels of initial weight loss as those experienced before the pandemic. We also examined the effects of COVID-19 on participants' selfmonitoring of weight, food, and activity. Research examining the effects of the pandemic on obesity, eating behaviors, and physical activity during the COVID-19 lockdown found a large number of people reporting snacking more frequently and having more difficulties with excess food consumption, compared with before lockdown (12) , but little has been reported on the tracking of health behaviors during this period. We found that the COVID-19 cohort had fewer days of food self-monitoring than the control cohort, but it was limited to 3 weeks of enrollment (enrollment weeks of March 8 through March 22), suggesting that at the start of the COVID-19 pandemic, members were tracking their food less often but returned to pre-COVID-19 tracking levels after that initial 3-week period. Among those beginning treatment in the end of March and beyond, differences in food tracking between cohorts no longer existed. It is not surprising that periods of decreased weight loss and decreased self-monitoring intake overlapped to some degree, given the strong relationship between dietary self-monitoring and weight (13) . It was surprising that there were no effects of COVID-19 on weight and activity tracking. It may be because they are, in general, less Year Among a large sample of those in a digital weight management program over 30 weeks of enrollments, the COVID-19 cohort experienced significantly less weight loss than the control cohort but only for 7 weeks of enrollments. The COVID-19 cohort also had fewer days of food tracking but only for 3 weeks of enrollments. There were no differences in the self-monitoring of weight and activity between the two cohorts. These data suggest that COVID-19 had negative effects on both weight loss and food self-monitoring, but the effects were short-lived. These data also suggest that those participating in evidence-based digital weight management programs can expect similar levels of initial weight loss as those experienced before the pandemic.O The impact of COVID-19 stay-at-home orders on health behaviors in adults Health behavior changes during COVID-19 and the potential consequences: a mini-review Impact of COVID-19 stayat-home orders on weight-related behaviors among patients with obesity Self-quarantine and weight gain related risk factors during the COVID-19 pandemic Cohen's d effect size by week for tracking weight, food, and activity in COVID-19 versus control cohorts Body weight changes during pandemic-related shelter-in-place in a longitudinal cohort study Initial weight loss response as an indicator for providing early rescue efforts to improve long-term treatment outcomes Early-treatment weight loss predicts 6-month weight loss in women with obesity and depression: implications for stepped care Evaluation of early weight loss thresholds for identifying nonresponders to an intensive lifestyle intervention Early weight loss in behavioral treatment predicts later rate of weight loss and response to pharmacotherapy A power primer Changes in weight and nutritional habits in adults with obesity during the "lockdown" period caused by the COVID-19 virus emergency Obesity, eating behavior and physical activity during COVID-19 lockdown: a study of UK adults Self-monitoring in weight loss: a systematic review of the literature Impact of the COVID-19 Pandemic on Initial Weight Loss in a Digital Weight Management Program: A Natural Experiment