An Evaluation of a Smartphone–Assisted Behavioral Weight Control Intervention for Adolescents: Pilot Study

Abstract

BackgroundThe efficacy of adolescent weight control treatments is modest, and effective treatments are costly and are not widely available. Smartphones may be an effective method for delivering critical components of behavioral weight control treatment including behavioral self-monitoring.ObjectiveTo examine the efficacy and acceptability of a smartphone assisted adolescent behavioral weight control intervention.MethodsA total of 16 overweight or obese adolescents (mean age=14.29 years, standard deviation=1.12) received 12 weeks of combined treatment that consisted of weekly in-person group behavioral weight control treatment sessions plus smartphone self-monitoring and daily text messaging. Subsequently they received 12 weeks of electronic-only intervention, totaling 24 weeks of intervention.ResultsOn average, participants attained modest but significant reductions in body mass index standard score (zBMI: 0.08 standard deviation units, t (13)=2.22, P=.04, d=0.63) over the in-person plus electronic-only intervention period but did not maintain treatment gains over the electronic-only intervention period. Participants self-monitored on approximately half of combined intervention days but less than 20% of electronic-only intervention days.ConclusionsSmartphones likely hold promise as a component of adolescent weight control interventions but they may be less effective in helping adolescents maintain treatment gains after intensive interventions.

A systematic review of community‐based interventions for the treatment of adolescents with overweight and obesity

“…Although the program with the greatest participant contact ( n = 52 sessions) yielded the largest effect size (9%) when BMIz was reported as an outcome, no consistent pattern was observed between total or frequency of contact and weight outcome measures. Including 1:1 support in addition to group sessions did not appear to improve effectiveness, with reductions in BMIz ranging from 2% to 9% at intervention end for programs including 1:1 support compared with 4% to 9% for those without. Likewise, reductions in BMIz at intervention‐end were similar between studies that had an additional psychological component (3–4% to 9% ) and those that did not (2% to 9% ).…”

“…Studies specifically targeting adolescents reported reductions in BMIz ranging from 3 to 4% to 9% at intervention end which was similar to those studies targeting the whole family (range 2% to 9% ). Online‐only programs which did not involve any face‐to‐face sessions had somewhat smaller reductions in BMI (between 1% and 2–4% ) compared with face‐to‐face only programs (1% to 7% ), and those which involved technology in addition to face‐to‐face sessions (reductions in BMIz; 2% to 4% End‐I, 3–4% to 9–10% FU) did not appear to be more effective than those without technology (reductions in BMIz; 5% to 9% End‐I; 8–12% to 16% at FU). Program length and study quality did not appear to be associated with weight management outcomes.…”

“…For the studies that also reported BMIz after a period of follow‐up, effect size typically increased over time, excluding that by Jensen et al . where effect size was greatest partway through the program at 3 months (6%), reduced at 6 months (4%) and remained stable at 12 months (4%) . Seven programs, ranging in length from 6 weeks to 4 months, reported BMI as an outcome measure and had smaller effect sizes which ranged from 1% to 7% .…”

“…Study, program and participant characteristics are reported in Table . Half of the programs were from the United States (US; n = 11) with the remainder from Europe ( n = 7) , Australia ( n = 2) and the United Kingdom (UK; n = 1) . The programs were evaluated as controlled trials ( n = 15), both randomized ( n = 10) and non‐randomized ( n = 5) , and as pre–post case series ( n = 6) .…”

“…Study quality was assessed to be weak for five programs ; moderate for eight ; and strong for eight . More than half of the programs were rated as being weak for selection bias ( n = 11) , as they involved self‐referrals from community advertisements and public service announcements, as is commonplace for community interventions, and thus participants were not selected at random from the eligible population.…”