Exposure to self-harm-inducing content

Self-harm can take on various forms, from posting mean content about oneself to developing depressive or antisocial views about oneself, to, at another extreme, suicide ideation and an eventual attempt to inflict physical injury. Youth activity in social media has helped accelerate this phenomenon, with recent research in the United States suggesting that between 2019 and 2021 about 10 percent of youth experienced some form of digital self-harm over that time.1

Eating disorders

Eating disorders are an increasingly prevalent challenge facing youth, fuelled in part by the rise of social media and the impact that image inspiration has on developing minds. Between 2002 and 2020 hospitalization rates more than doubled among children and youth in Ontario for eating related disorders.2 Despite the obvious and significant impact that online platforms can have in disseminating content that contributes to "thinspiration" and other forms of self-consciousness about body image, establishing a direct causal connection is not entirely clear in the literature and suffers from significant data gaps in Canadian public health data.3

That being said, given the significant rise in hospitalizations over the period in which social media platforms have gained widespread use, we can infer at least a notional connection. We treat this sub-component as a conservative illustrative upper bound rather than something directly taken from the literature: assuming a 25 percent attribution rate yields approximately 694 hospitalizations. Based on CIHI's patient cost estimator, and excluding any follow-up or pharmaceutical costs, this produces health system costs of between $8 to $14 million annually. Given the acknowledged limitations in establishing direct causality, readers may exclude this sub-component without materially affecting the paper's headline estimate; its inclusion or exclusion changes the overall total by less than 0.5 percent.

Self-harm-inducing content is the category where quality-of-life (QALY) losses dominate: direct health-system costs total $72–103 million annually, but welfare losses reach $554 million to $1.2 billion.

Self injury and suicide

In a more extreme sense we are now seeing a sustained rise in self-harm and mental health presentations4 and despite declining hospital admissions overall mental health burden remains above pre-pandemic levels. Mood and anxiety medication dispensing to youth increased 18 percent between 2018–19 and 2023–24 — an increase that reflects both population growth and rising per-youth prescription rates. A major 2026 review led by researchers at The Hospital for Sick Children (SickKids) and published in JAMA Pediatrics examined 42 international studies spanning 25 years and documented a 3.5 percent average annual increase in youth self-injury presentations to health facilities globally.5

In Canada, approximately 103 per 100,000 (crude rate) children and youth aged 10 to 24 were hospitalized as a result of an attempted self-injury or harm in 2024-25. This represents approximately 7,490 hospitalizations per year among this age group.6 Direct attribution to online harms is difficult to establish, however a variety of studies internationally have tended to show that about 1 in 4 children and youth who are admitted for self-injury report accessing online sources of information in order to help facilitate their attempt.7 In the U.K., for example, about a quarter of suicides among children and youth were associated with online search activity prior to, and of which a further quarter had a history of being bullied online in the preceding period.8

Based on this prevalence, and in light of previous cost for hospital admission and follow-up, we estimate that self-harm admissions produce annualized health system costs of between $64 to $89 million per year assuming a range of attribution between 18 to 25 percent associated to online harms. A portion of this includes costs previously captured as part of cyber-bullying which we remove in duplication as the final step in our costing analysis.

Quality-of-life losses

Self-harm-inducing content exposure imposes substantial welfare losses beyond direct health system utilization. We estimate these losses using Quality-Adjusted Life Years (QALYs), the standard health economics methodology used by Canada's Drug Agency, NICE in the United Kingdom, and equivalent regulators globally to value health states in dollar terms. One QALY equals one year of life in perfect health; a year lived in a degraded health state is worth proportionally less, with the decrement representing the welfare loss. We apply the $50,000 per QALY threshold implicit in CDA's drug reimbursement decisions.9 For context, at a 0 percent discount rate the implied value per attributable youth death rises to approximately $3.15 million (63 undiscounted QALYs × $50,000), which would increase the welfare loss estimate for attributable deaths by approximately 110 percent relative to the 3 percent discounted figure used here. Some recent Canadian literature suggests the revealed-preference threshold is closer to $100,000 per QALY, which would double our estimates; we use the lower figure to ensure a more defensible lower bound.

We decompose welfare losses into three severity tiers:

• Attributable youth suicide. Triangulating PHAC and Statistics Canada age-specific suicide rates across the 10–24 cohort yields a population-weighted rate of approximately 7.5 per 100,000, implying roughly 545 youth suicide deaths annually. Applying an 18 to 25% attribution fraction yields between 98 and 136 attributable deaths. Each death represents approximately 30 discounted QALYs lost (mean age 18, remaining life expectancy 63 years, 3% discount rate), valued at $147 million to $204 million annually.
• Attributable self-harm hospitalization. Applying CIHI's self-harm hospitalization rate of 103 per 100,000 (crude rate, 2024–25) to the 10–24 cohort implies approximately 7,490 hospitalizations annually, against a similar 18 to 25 percent attribution range. We apply a QALY decrement of 0.25 for the year of the event — consistent with EQ-5D utility weights for severe depression with self-harm features — yielding a welfare loss of approximately $17 million to $23 million annually.
• Sub-clinical to mild clinical mental health impacts on the broader exposed population — depressive symptoms, anxiety, disordered eating cognitions, or general psychological distress without hospitalization. We estimate prevalence at 3 to 4 percent of Canadian adolescents aged 13–19, equivalent to 97,000 to 129,000 affected youth annually. Applying a QALY decrement of 0.08 to 0.15 per affected youth-year yields a welfare loss of $390 million to $970 million annually.

Combined, this produces welfare losses of between $554 million to $1.2 billion annually. This figure is conceptually distinct from, and non-overlapping with, the direct health system costs presented earlier — the latter captures public resources expended on acute care, while welfare losses capture diminished wellbeing experienced by affected youth and their families. Caregiver burden — including hours of parental and family labour time diverted to supporting youth with mental health challenges attributable to online harms — is not captured in these estimates. If included, it would increase the total potentially substantially.