Identification and Prediction of High-Risk Periods for Regain After Weight Loss

 描述（由申请人提供）：肥胖仍然是美国的一个重大公共卫生问题。尽管行为减肥疗法最初有效，但这些治疗结束后减肥的长期维持效果仍然很差。目前改善长期减肥维持的干预措施具有统计显着性，但对体重恢复的影响有限，并且没有一项能够改变治疗结束后通常观察到的体重恢复模式。这些结果表明需要新的方法来维持长期减肥。适应参与者进展并在参与者体重反弹高风险时提供有针对性的干预措施的治疗方法可能是有益的，并且可能有助于适应通常在干预反应中观察到的巨大个体差异。基础的 然而，开发这些治疗方法所需的证据，包括个体何时体重增加以及哪些因素预测体重增加，尚不可用。我们建议对丰富的纵向数据集进行二次分析，包括从 3 个月的基于互联网的行为减肥干预和随后的 9 个月的非接触“维持”期收集的数据。 75 名参与者（平均年龄 = 50.8 岁，BMI = 31.2，69% 女性，84% 白种人）提供了每日体重测量（使用“智能秤”，通过蜂窝网络将体重直接发送到我们的研究中心）和每周自我报告的行为依从性（体重、热量摄入和体力活动的自我监测）以及被假设影响减肥维持的心理因素（例如压力、饥饿、对与目标相关的饮食和活动行为、竞争优先事项感到厌倦）。在干预期间，参与者平均 (±SD) 降低了基线体重的 -6.26±4.88%，随后从第 3 个月到第 12 个月，平均体重恢复了 2.98±4.55%（51% 达到了 12 个月时体重显着恢复的临床临界值）。对每日和每周数据收集的遵守程度很高。权重数据集目前包括大约 23,500 个权重值，代表所测量的潜在日期的 86%。 52 周内，每周自我报告措施的平均遵守率为 72%。这些数据提供了一个独特的机会来识别个体的体重轨迹何时从体重减轻转向体重反弹（由“拐点”识别），并描述这些点之前的高风险期（预测这些拐点的前一周的体重变化模式、行为坚持和心理因素），使我们能够开发算法来预测个体何时接近体重反弹的高风险时间。这些结果具有明确的治疗意义，因为它们将为未来维持减肥的适应性干预提供经验基础，这可以在个体体重反弹高风险时提供有针对性的行为干预策略。

 DESCRIPTION (provided by applicant): Obesity remains a substantial public health problem in the United States. Despite the initial effectiveness of behavioral treatments for weight loss, long-term maintenance of weight loss after the end of these treatments remains poor. Current interventions to improve long-term weight loss maintenance have statistically significant but only modest impact on weight regain, and none have been able to change the pattern of weight regain typically observed after treatment ends. These results suggest that new approaches to long-term weight loss maintenance are needed. Treatment approaches that adapt to participants' progress and that deliver targeted intervention components at times when participants are at high-risk for weight regain may be beneficial, and may help accommodate the substantial individual variability typically observed in response to intervention. The foundational evidence needed to develop these treatments, however, including when individuals regain weight and what factors predict this regain, is not available. We propose to conduct secondary analyses on a rich longitudinal dataset including data collected from a 3- month Internet-based behavioral weight loss intervention followed by a 9-month no-contact "maintenance" period. Seventy-five participants (mean age=50.8y, BMI=31.2, 69 percent Female, 84 percent Caucasian) provided daily weight measurements (using "smart scales" that sent weights directly to our research center via the cellular network) and weekly self-report of behavioral adherence (self-monitoring of weight, caloric intake, and physical activity) and psychological factors that have been hypothesized to affect weight loss maintenance (e.g., stress, hunger, boredom with goal-related eating and activity behaviors, competing priorities). During intervention, participant lost an average (±SD) of -6.26±4.88 percent of baseline weight, followed by a mean weight regain of 2.98±4.55 percent from Month 3 to 12 (51 percent met clinical cut-offs for significant weight regain at 12 months). Adherence to daily and weekly data collection was high. The weight dataset currently includes approximately 23,500 weight values representing 86 percent of potential dates measured. Average adherence to weekly self-report measures was 72 percent over the 52 weeks. These data provide a unique opportunity to identify when an individual's weight trajectory shifts from weight loss to regain (identified by an "inflection poin") and to characterize the high-risk periods prior to these points (the pattern of weight change, behavioral adherence and psychological factors from the week prior that predict these inflection points), allowing us to develop algorithms to predict when an individual is approaching a high-risk time for weight regain. These results have clear treatment implications as they will provide the empirical foundation for a future adaptive intervention for weight loss maintenance, which could provide targeted, behavioral intervention strategies at times when individuals are at high risk for weight regain.