What the AHA's Redefinition of Isolation Left Out

On August 4, 2022, the American Heart Association published a scientific statement on social isolation and cardiovascular and brain health, written on behalf of several AHA councils with Crystal W. Cene as lead author. The statement’s most consequential move was not a new number. It was a definitional split: it separated “objective social isolation” — the structural fact of a small network and infrequent contact — from “perceived isolation,” which is loneliness, the subjective experience of lacking connection. Cardiovascular research had not always made this distinction cleanly, and the statement’s authors treat doing so as a prerequisite for comparable future research.

What the revision actually changed

Before this statement, much of the cardiovascular literature on isolation used a single composite score, or a proxy like living alone, and reported one risk estimate. The AHA statement instead reports isolation and loneliness as related but separable exposures, each with its own risk profile: a roughly 30% increased risk of heart attack, stroke, or death from either, with the statement giving 29% for heart attack or death from heart disease and 32% for stroke specifically. It also reports that isolation and loneliness worsen prognosis in people who already have coronary disease or a prior stroke, including risk of recurrence.

This tracks Julianne Holt-Lunstad’s 2015 meta-analysis, which found separate odds ratios for isolation (1.29), loneliness (1.26), and living alone (1.32) as predictors of early mortality, and her earlier 2010 meta-analysis, which found that weak social relationships carried a mortality risk comparable to smoking. The AHA statement’s contribution was applying that same conceptual separation specifically to cardiovascular and cerebrovascular outcomes, where the two constructs had often been blurred together in a field that was, until recently, more focused on physiology than on measurement.

The National Academies’ 2020 consensus report had already flagged the underlying problem: roughly one quarter of adults 65 and older meet criteria for social isolation by structural measures, a figure derived from network size and contact frequency rather than from how isolated people say they feel. Isolation and loneliness are not interchangeable, and a person can score high on one and low on the other. Cacioppo’s earlier account of loneliness as an evolved aversive signal, distinct from the objective fact of being alone, gave the psychological grounding for a distinction that cardiovascular medicine took over a decade to formalize into a scientific statement.

What the new definition still does not settle

Splitting the construct clarifies what is being measured, but it does not resolve what counts as contact in the first place. “Objective social isolation” is typically operationalized through network size and frequency of contact — how many people, how often. What the AHA statement does not specify, and what the underlying instruments it draws on generally do not standardize, is whether a phone call or a video call counts the same as an in-person visit toward reducing that score.

This is not a minor gap. During the period this statement’s evidence base was built, much of the contact available to isolated people, particularly older adults with mobility limits, was mediated by phone or screen. If a structural isolation measure counts a weekly video call from a grandchild the same as a weekly in-person visit, two people with identical scores could have very different actual risk, if the mode of contact matters physiologically and not just numerically. If the measure does not count remote contact at all, then any population that shifted heavily toward phone and video contact would appear to have deteriorating objective isolation scores even where the total amount of social contact held steady. Neither convention is obviously correct, and the statement does not adjudicate between them because the underlying studies it synthesizes were not designed to.

The statement is candid about a related and larger gap: it explicitly identifies the absence of intervention evidence as the central problem facing this research area. Knowing that isolation and loneliness predict a 30% increase in cardiovascular risk says nothing about whether any intervention — in-person, remote, or some combination — reduces that risk. The AHA statement calls for that evidence rather than claiming to have found it.

Why the split matters for policy, not just research

Separating objective isolation from perceived loneliness has a direct implication for how health systems screen for either. A structural isolation score can be estimated from administrative or self-reported network data without asking a patient how they feel. A loneliness score requires asking, typically with an instrument like the UCLA Loneliness Scale. These are different clinical tasks with different burdens, and a health system that adopts one screening approach may miss patients who are captured only by the other. A patient who lives with a large family but feels disconnected from all of them would screen as low-risk on a structural measure and high-risk on a subjective one.

The AHA statement does not recommend a specific screening protocol, and in the newsroom summary accompanying the statement the Association’s own framing is limited to noting that older adults and socially vulnerable groups face elevated risk, without prescribing how clinicians should act on that. That restraint is appropriate given the evidence gap the statement itself names.

What a better study would look like

A study that closed this gap would measure contact by both frequency and modality, tracking in-person, phone, and video contact as separate variables against the same cardiovascular outcomes the AHA statement examined, rather than folding all three into one “objective isolation” score. It would also need to be a genuine intervention trial rather than another observational cohort. The mortality and cardiovascular associations are now well established across multiple large meta-analyses. What remains unestablished, and what the AHA statement says so plainly, is whether changing a person’s isolation status — by any means — changes their risk.