Mucundorfeanu · Body image 2026 · scoping review · n=59 studies

Pretty hurts. A scoping review of interventions targeting digitally manipulated images.

Cited 0 times in the scientific literature.

Level 5 - mechanism / opinion, no new human data

Scoping review mapping heterogeneous literature without quantitative meta-analysis (graded by design analogy)

PubMed 41990434 · doi:10.1016/j.bodyim.2026.102088 · record verified 2026-08-29

What was done

Following PRISMA-ScR guidelines, this scoping review synthesized 59 empirical studies examining interventions targeting psychological and behavioral responses to idealized or digitally altered media images. Literature was mapped across intervention types, mediating and moderating variables, and outcome measures.

What was found

The abstract reports no numerical data, effect sizes, or test statistics. Across seven identified intervention categories, disclaimer and warning labels dominated the literature but showed limited or mixed outcomes, occasionally increasing appearance-related concerns. Conversely, authenticity labels, paired images depicting unedited reality ('Instagram vs reality'), and body-positive or self-compassionate messages were more frequently linked to favorable affective and body-related outcomes. Media literacy and psychoeducation showed modest, short-term benefits, and mandatory disclosures raised manipulation awareness without consistent attitudinal improvement. Perceived authenticity and realism frequently mediated effects, while appearance comparison and thin-ideal internalization acted as recurrent moderators.

Why it matters

These findings suggest that standard disclaimer labels on retouched media provide minimal benefit and may backfire, whereas interventions emphasizing authenticity and compassion offer a more promising direction for policy and intervention design.

Limits

The abstract provides no numerical data, pooled effect sizes, or confidence intervals. Participant characteristics, sample sizes, intervention durations, and study-level risk of bias are not reported in the abstract.

Cited by