How the AHA Statement Handled Its Own Uncertainty

The American Heart Association’s 2022 scientific statement on social isolation and cardiovascular and brain health reports a 29% increased risk of heart attack or death from heart disease, and a 32% increased risk of stroke, associated with social isolation and loneliness. Those numbers have circulated widely since, usually stripped of the sentence that follows them in the original document: the authors state, in the same statement, that there is no adequate trial evidence showing that intervening on isolation changes cardiovascular outcomes. That is an unusual thing for a scientific statement to say about its own headline finding, and it is worth examining closely, because it demonstrates both what careful uncertainty disclosure looks like and where even a careful document loses its nerve.

What the statement actually claims

The AHA document, led by Crystal W. Cene on behalf of several of the association’s councils, synthesizes observational evidence linking social isolation and loneliness to cardiovascular and cerebrovascular outcomes. The headline figures are associational: isolated or lonely people have measurably worse cardiovascular outcomes than people who are not, after adjustment for the usual confounders. The statement is explicit that these are risk associations drawn from cohort studies, not effects demonstrated by trials. It also draws a further distinction that a great deal of secondary reporting drops: between objective isolation, a structural fact about the size and contact frequency of a person’s network, and perceived loneliness, a subjective appraisal of that network as insufficient. The statement treats these as related but non-identical exposures, each independently associated with worse prognosis in people who already have coronary disease or a prior stroke, including elevated recurrence and mortality.

Where the statement earns its keep, methodologically, is in the section that most such documents bury or omit: it says outright that the absence of intervention evidence is the central gap in the field, not a footnote to it. Most of the underlying literature establishes that isolated and lonely people fare worse. Almost none of it establishes that reducing isolation improves cardiovascular outcomes, because the trials that would show this mostly do not exist. The American Journal of Lifestyle Medicine review by Julianne Holt-Lunstad, published the year before, makes a parallel point in a different register: it argues for treating social connection as a modifiable preventive factor alongside diet and exercise, while acknowledging that the evidence base for that comparison rests far more on observational risk association than on demonstrated modifiability. Naming the gap does not close it.

Why this matters more than it sounds

A reader who takes only the 29% and 32% figures away from the AHA statement could reasonably conclude that a cardiology intervention program built around reducing isolation would lower heart attack and stroke rates by a comparable margin. The statement does not support that conclusion, and says so. The risk figures come from comparing people who already differ in their level of isolation; nothing in that design tells anyone what happens if isolation is changed for people who did not choose their level of it. This is the standard association-versus-causation problem, but the AHA statement is worth citing specifically because it states the limitation as a matter of formal record within a document that is otherwise built to be quoted for its numbers. Most producers of a striking statistic let the caveat trail off into methods-section prose that nobody reads. This one puts the caveat in the abstract-adjacent material, where the newsroom summary can find it. The American Heart Association’s own newsroom release does reproduce the risk figures prominently and does not foreground the missing-trial-evidence point with the same weight, which is a reminder that the press office and the scientific statement are not always calibrated to the same audience.

Where the candor runs out

The statement is less careful about a second kind of uncertainty: measurement heterogeneity. The 29% and 32% figures aggregate studies that used different instruments to define isolation and loneliness, some relying on a single self-report item (“how often do you feel lonely”), others using multi-item scales such as the UCLA Loneliness Scale used in the 2015 Holt-Lunstad meta-analysis in Perspectives on Psychological Science, which found separate odds ratios of 1.29 for isolation, 1.26 for loneliness, and 1.32 for living alone. A pooled hazard ratio drawn from studies using different instruments is not wrong, but it is doing more work than a single number implies, since it is averaging across measurement choices that do not agree on what they are measuring. The National Academies’ 2020 consensus report on isolation in older adults makes essentially the same aggregation, estimating that roughly a quarter of adults 65 and older are socially isolated, while relying on cohort studies that vary in how isolation was operationalized. That report, like the AHA statement, calls for routine clinical assessment of isolation despite acknowledging that no consensus instrument exists for that assessment to use. Recommending routine screening before settling on what to screen with is a genuine tension, and neither report resolves it; both simply proceed past it into a policy recommendation.

The cigarette comparison, by contrast

The 2023 Surgeon General’s advisory on loneliness and isolation offers a useful point of contrast, because it makes a much bolder-sounding claim with less internal hedging. The advisory states that the mortality risk of social disconnection is comparable to smoking up to fifteen cigarettes a day. That comparison, drawn ultimately from the same Holt-Lunstad body of work, is arresting and has been widely repeated. It is also, on its face, a comparison between an association (disconnection and mortality risk) and a well-established causal exposure (smoking and mortality). The advisory does not spend comparable space distinguishing what is known about disconnection’s association with mortality from what would be needed to establish it as a cause on the same evidentiary footing as tobacco. The AHA statement’s more granular treatment of its own gap, by comparison, reads as the more disciplined document, even though it is the less quotable one. A comparison built for public communication and a scientific statement built for a clinical and research audience are different genres with different obligations, and it may be unfair to hold the advisory to the standard of the statement. But the asymmetry is real: the document more likely to be quoted verbatim in a news article is also the one that does the least to prevent an inferential leap from association to cause.

What would close the gap

A statement that names the absence of trial evidence as the central limitation, as the AHA’s does, is implicitly specifying what would strengthen the field: randomized or quasi-experimental studies that manipulate social contact or perceived connectedness and follow cardiovascular endpoints over years, not weeks. That is an expensive and slow study to run, which is presumably why it does not yet exist in a form the statement can cite. Short of that, cohort studies that use a single validated instrument across sites, distinguish isolation from loneliness at the design stage rather than the analysis stage, and report effect sizes separately by instrument would let a future reviewer do what this one cannot: compare like with like, rather than averaging across studies that were never asking quite the same question.