Date of Completion


Embargo Period



Tania Huedo-Medina and Michael Copenhaver

Field of Study

Health Promotion


Master of Science

Open Access

Open Access


Individuals with lower income have higher rates of obesity and are underrepresented in weight loss treatment. Moreover, when individuals with lower income do enroll in face-to-face interventions, their adherence and weight loss outcomes are worse due to known treatment barriers, including unreliable transportation, lack of childcare, and scheduling constraints associated with multiple jobs and unconventional work hours. Because online interventions are asynchronous and can be accessed any time of day, they hold significant promise for reducing treatment barriers and improving adherence and weight loss outcomes in lower income populations. However, one drawback of online interventions is that they are self-directed, which, in the context of multiple life stressors, may be particularly challenging. This study examined whether income is in fact associated with engagement and weight loss outcomes in an online weight loss program, and whether perceived stress mediates these effects. We hypothesized that lower income would be associated with less engagement with the weight loss program and poorer weight losses and that stress would mediate these effects.

Participants (N=260, 79% female, 71.9% White/Non-Hispanic/Latino, age 50.7±11.9 years, 331 ± 129.6% above the poverty line, BMI 35±5.7kg/m2) received a 16-week DPP-based online behavioral weight loss intervention. Income was assessed via self-reported data. Engagement (weeks lesson viewed and weight entered) was obtained from the intervention website. Weight loss was calculated using objective measurement when available, or, when unavailable, last weight reported on the website. Stress was measured with the Perceived Stress Scale.

Our mediation analyses indicated perceived stress mediated the relationship between income on engagement [β = 0.001, 95% CI = 0.02-0.3]. Through the mediator of stress, lower income was associated with lower stress scores (β = -0.37, p = 0.01, 95% CI = -0.66-(-0.08)) and lower engagement was associated with higher stress scores (β = -0.369, p = 0.001, 95% CI = -0.58-(-0.16)). Additionally, higher levels of stress partially mediated the relationship between income and percent weight loss [β = -0.1, 95% Cl = -0.3-(-0.03)]. Through the mediator of perceived stress, lower income was associated with less weight loss (β = -0.352, p = 0.001, 95% CI = 0.14-0.56).

These findings demonstrate that perceived stress mediates, in part, the relationship between income and obesity treatment outcomes in an online intervention. Thus, stress reduction interventions may be critical to effectively engage individuals with lower income and improve weight loss in obesity treatments.

Major Advisor

Tricia Leahey Ph.D