A platform-by-frequency view of mental-health severity among U.S. adults. Toggle between overall depression (PHQ-9) and the suicidal-ideation item, and between raw means and age-standardized means, to see how much of the "heavy-use" pattern is really about who uses each platform most.
Source: CHIP50 (Civic Health and Institutions Project / COVID States Project), Wave 35. All figures are weighted estimates. Frequency-scale wording matches the fielded Wave 35 instrument. Age-standardized to each platform's user population using age (8-category) as the control. Companion report: Social Media Use and Depression.
If this topic is affecting you personally, support is available — in the U.S. you can call or text 988 (the Suicide & Crisis Lifeline) at any time.
How to read it. Each cell is the weighted mean symptom score among adults who report a given platform at a given frequency. Rows are platforms; columns are usage frequency from occasional (left) to most of the day (right). Color is scaled within each outcome, so the Raw/Age-standardized toggle is directly comparable.
Why age-standardize. Heavy users of most platforms skew young, and younger adults report more depression regardless of screen time. The Raw view therefore mixes a frequency effect with an age-composition effect. The Age-standardized view holds age constant (age-8 categories) and re-centers each platform to its own overall user mean, isolating the within-age gradient. Where the two views differ, the raw pattern was partly about who uses the platform, not the intensity itself.
The recurring signal. Across nearly every platform, the jump concentrates in the heaviest category ("most of the day"); the gradient across the lighter categories is shallow or non-monotonic, and often flattens further after age-standardization. Estimates for smaller cells (e.g., Threads, high-frequency LinkedIn) carry more uncertainty.