Loneliness Has a Cost Estimate. It Does Not Have a Cost Study.

The claim that loneliness costs a health system, an employer, or a national economy a specific sum of money appears constantly in coverage of this topic. It rarely appears with a method attached. That gap is worth examining directly, because it is the weakest link between otherwise solid epidemiological findings and the policy urgency built on top of them.

What the mortality evidence actually supports

The strong part of this literature is genuinely strong. Julianne Holt-Lunstad’s 2010 meta-analysis, pooling 148 studies and 308,849 participants, found that stronger social relationships were associated with a 50% increased likelihood of survival. Her 2015 follow-up put the mortality risk of social isolation at an odds ratio of 1.29, loneliness at 1.26, and living alone at 1.32, comparable in scale to well-established risk factors. The American Heart Association’s 2022 scientific statement, led by Crystal Cene, found isolation and loneliness associated with roughly a 29% increased risk of heart attack or death from heart disease and a 32% increased risk of stroke. The 2023 Surgeon General advisory summarized the mortality comparison memorably: a risk on the order of smoking up to 15 cigarettes a day.

These are association studies, not causal proof, and the authors are generally careful to say so. But they are large, replicated across populations, and consistent in direction. This is not where the cost problem lives.

Where it breaks down

The cost problem starts the moment someone tries to convert a hazard ratio into a dollar or pound figure. To do that legitimately requires knowing several things the mortality literature was never designed to measure: the incremental health care utilization attributable to loneliness specifically (as opposed to isolation, living alone, or the underlying conditions that cause both), the employer-side costs of absenteeism and turnover attributable to loneliness specifically, and some accepted method for monetizing years of life or quality of life lost. Each step compounds uncertainty. None of the widely cited economic figures documents this chain in a way that would let another researcher reproduce it.

Consider what does exist. The Cigna 2020 workplace report found that lonely workers miss work twice as often due to illness and five times as often due to stress, and that 61% of U.S. adults reported feeling lonely at least sometimes, up seven points year over year. Those are real, useful workplace metrics: absenteeism ratios drawn from a large employee survey. But a ratio of missed days is not a cost estimate. Getting from “lonely workers miss work five times as often due to stress” to “loneliness costs employers $X billion annually” requires assumptions about wage rates, replacement costs, and the counterfactual absence rate for otherwise-identical non-lonely workers that the survey itself does not supply. Gallup’s 2024 workplace data adds a comparable descriptive figure – one in five employees worldwide report feeling lonely a lot of the previous day, with fully remote workers at 25% against 16% for fully on-site staff – but again stops at prevalence and correlation, not dollars.

The National Academies’ 2020 consensus report on older adults, one of the more methodologically careful documents in this field, is notable for what it does not do. It estimates that roughly a quarter of adults 65 and older are socially isolated and calls for health systems to screen for isolation routinely. It does not attach a national cost figure to that quarter, and its recommendations are framed around clinical practice and screening infrastructure rather than a return-on-investment case. That restraint is itself informative: the body most invested in getting health systems to act declined to lean on a cost number it could not defend.

The intervention evidence is thinner still

If the cost-of-inaction side is weak, the cost-effectiveness side of interventions is weaker. Social prescribing – referring patients to community activities, groups, or programs rather than, or alongside, medical treatment – is the most widely tried policy response to loneliness, embedded in the UK’s 2018 national loneliness strategy. Systematic reviews of social prescribing report consistently positive qualitative outcomes: participants describe restored self-esteem and a sense of purpose, and a 2021 systematic review found three of nine included studies reported reductions in GP, emergency, social worker, or inpatient service use. A 2022 qualitative meta-synthesis found participants perceived benefits extending well beyond social contact itself.

None of this amounts to a cost-effectiveness study in the sense health economists use the term. Reduced service utilization in three small studies is suggestive, not quantified against the cost of running the referral program. The 2021 review is explicit that trial evidence is limited and programs are too heterogeneous to pool. The AHA’s 2022 statement goes further than most sources in naming the actual gap: it identifies the absence of intervention evidence as the central research problem in this entire field, not a footnote to it. A 2023 review in BMC Public Health mapping the state of loneliness research reaches a similar conclusion from a different angle, flagging inconsistent measurement across studies as a barrier to comparing anything, cost included.

What would actually settle this

A defensible cost estimate would need a longitudinal design that separates loneliness from isolation and from confounding health conditions, tracks health care utilization and workplace outcomes for the same individuals over time, and applies a transparent, disclosed method for monetizing outcomes – the kind of costing exercise that exists for smoking, obesity, or alcohol harm, built on decades of attributable-risk modeling. Nothing in the current loneliness literature does this. What exists instead is a set of large, credible prevalence and mortality-risk studies, bridged to policy urgency by cost figures that, when traced to their source, generally turn out to rest on assumption rather than measurement.

That does not mean the underlying problem is overstated. The mortality and cardiovascular associations are among the more robust findings in social epidemiology. It means the dollar figure attached to a news story about loneliness is doing more rhetorical work than evidentiary work, and should be read accordingly.