Conference Coverage

For weight-loss apps, the evidence base is still small



Digital apps for weight loss are ubiquitous, but do they work as a standalone intervention? So far, the evidence is scant, according to a new systematic review.

Beginning with a pool of 1,380 publications, Christina Hopkins and her colleagues at Duke University, Durham, N.C., eventually identified just nine trials of all-digital interventions for weight loss that met their inclusion criteria.

Presenting their findings at a late-breaking poster session during Obesity Week, presented by the Obesity Society and the American Society for Metabolic and Bariatric Surgery, Ms. Hopkins, a clinical psychology graduate student at Duke, and her colleagues found that three of the nine studies showed statistically significant weight loss, compared with a control state. Absolute weight loss in these three trials ranged from 3 kg to about 7 kg (between-group differences, P less than .001 for all).

Participants in another trial didn’t lose a statistically significant amount of weight, compared with the control arm of the study. However, the mean 5 kg lost by those in the intervention arm was enough to be clinically significant, so Ms. Hopkins and her colleagues included this study in a subanalysis that looked for effective modalities and interventions among the studies with significant results.

The duration of the studies ranged from 6 to 24 months, though five of the trials were less than 1 year long. Women made up the majority of participants in all but one trial.

“There is limited evidence that standalone digital weight-loss interventions produce clinically meaningful outcomes,” wrote Ms. Hopkins and her coauthors. “Absolute magnitude of weight loss was low, and the short intervention lengths call into question the sustainability of these weight losses.”

The systematic review cast a broad net to include digital modalities such as wireless scales, text messaging, email, and web-based interventions, as well as the use of smartphone apps and tracking devices. All interventions used multiple digital modalities.

The most frequently used technologies were the use of a website, used in six (67%) of the trials, followed by text messaging and smartphone apps, each used in five (56%) of the trials. Tracking devices, email, message boards, and gamification of some sort were all used in three (33%) of the trials.

In terms of the specific interventions used in the trials, weight, diet, and activity were all tracked in eight trials (89%). Similarly, all but one trial gave feedback and weight and health education to participants. Behavior change education, as well as calorie goals, were each used in six trials (67%).

Ms. Hopkins and her colleagues looked at which trials incorporated which modalities and interventions, finding that “trials that integrated components unique to digital interventions, such as gamification, podcasts, or interactive features, yielded significantly greater and more clinically meaningful weight losses.”

To be included in the systematic review, trials had to include adult participants with a body mass index of at least 25 kg/m2 and use a standalone digital intervention of at least 6 months’ duration. The primary outcome of interest in the review was the change in participant weight from baseline to the end of the minimum 6-month follow-up period. Randomized, controlled trials and feasibility trials were included, so long as participants were allocated randomly.

Of the 126 trials reviewed at the full text level, the most frequent reason for exclusion was the inclusion of human coaching. Also, 30 of the trials didn’t report weight change as an outcome, the investigators said.

Future directions should include comparing digital interventions that “utilize features unique to digital delivery” with those that more closely resemble in-person weight-loss management interventions, suggested Ms. Hopkins and her collaborators.

The authors reported no outside sources of funding and no conflicts of interest.

SOURCE: Hopkins C et al. Obesity Week 2018, Abstract T-P-LB-3640.

Next Article: