Effects of Decentering on Food Craving in Adults: A Systematic Review and Meta-analysis of Controlled Clinical Trials

Food cravings (FCs) are well-established as a significant factor in obesity. Despite their critical importance, mainstream anti-obesity interventions, including dietary programs and anti-obesity medications, often overlook FCs. While several studies have explored the impact of decentering on FCs, a comprehensive review of this literature has been lacking. Therefore, this study systematically reviewed and analyzed the existing research on the application of decentering to food cravings.

We conducted a systematic review and meta-analysis following PRISMA guidelines and registered it with PROSPERO (CRD420250655903). We systematically searched MEDLINE (via PubMed), Scopus, Embase, CENTRAL, Web of Science, and ProQuest databases. We assessed methodological quality and risk of bias in included studies using the Cochrane RoB2 tool. We evaluated the certainty of the synthesized evidence using GRADEpro. Data synthesis and statistical analyses were performed with CMA v3 software, reporting standardized mean differences as Hedges' g with 95% confidence intervals (CI). To clarify interpretation, outcomes favoring decentering interventions are presented as positive Hedges' g values, while effects favoring control conditions are shown as negative values.

We identified twelve eligible studies, encompassing 1,079 participants. Across these studies, participants had a mean age of 20.4 years, a mean BMI of 22.8, and were predominantly female (82.8%). Using the RoB2 tool, one study was rated as low risk of bias, seven as having some concerns, and four as high risk. The overall effect size for FC intensity was g = 0.38 (95% CI, 0.16 to 0.59; P = 0.001; n = 11). Using the GRADEpro tool, we rated the certainty of this result as moderate. The pooled effect size for FC frequency was g = 0.12 (95% CI, - 0.28 to 0.52; P = 0.56; n = 6).

Decentering effectively reduces the intensity of food cravings but not frequency. Because most included studies involved young, normal-BMI women and had short follow-up, the generalizability of our pooled estimates is limited.