In July 2010, three American psychology researchers — Julianne Holt-Lunstad and Timothy B. Smith of the Brigham Young University Department of Psychology in Provo, Utah, and J. Bradley Layton of the University of North Carolina at Chapel Hill Department of Epidemiology — published in the open-access journal PLoS Medicine the substantial 21-page meta-analytic paper “Social Relationships and Mortality Risk: A Meta-analytic Review,” which had, across the preceding approximately three years, systematically aggregated the mortality outcome data from 148 previously-published empirical studies of the relationship between social connectedness and premature death, encompassing a combined sample size of 308,849 human research subjects across substantially every developed country in which comparable epidemiological research had been conducted, with an average follow-up interval per subject of approximately 7.5 years. The paper’s substantive finding, benchmarked against the specific mortality-risk data for the other major behavioural and physiological risk factors that contemporary public health research had established as substantive predictors of premature death, was that the specific magnitude of the mortality-risk elevation associated with substantial deficits in social connection was approximately equivalent to the specific mortality-risk elevation associated with smoking 15 cigarettes per day.
The specific comparison to cigarette smoking was, by every subsequent account of the paper’s international press reception, substantially the single specific research finding that produced the paper’s substantial subsequent global cultural impact. The comparison had, in the specific 2010 public health discourse in which the paper was published, essentially no precedent. Contemporary public health research had, across the preceding half-century, systematically documented the specific mortality effects of tobacco consumption (well-established as one of the substantially most consequential single behavioural risk factors for premature death), of obesity, of hypertension, of physical inactivity, of excessive alcohol consumption, of air pollution exposure, and of a substantial range of other well-documented environmental and lifestyle factors. The specific mortality effects of social isolation — of the substantive absence of friends, family members, or community connections in an individual’s life — had been substantially documented in isolated empirical studies since approximately the 1970s but had not previously been systematically quantified against the standard reference risk factors that public health researchers used to benchmark comparative mortality risks. The Holt-Lunstad-Smith-Layton 2010 paper was the first substantive academic publication in the recorded history of Anglophone public health research to place the specific risk of social disconnection on the same quantitative axis as the specific risk of tobacco consumption. And it found them approximately equivalent.
The specific methodology by which the Holt-Lunstad team arrived at the specific 15-cigarette comparison is worth reconstructing in detail. As detailed in the original 2010 PLoS Medicine paper published in the National Center for Biotechnology Information’s PubMed Central archive, the researchers began with a systematic literature search of essentially every major English-language psychology, sociology, epidemiology, and medical research database, initially identifying 11,124 potentially relevant published studies. Successive rounds of methodological screening (excluding studies that did not measure social connectedness as an independent variable, that did not track mortality as an outcome variable, that used insufficient statistical methods to extract effect sizes, that studied only marital status rather than broader social connectedness, or that included suicides as a cause of death) eliminated approximately 10,976 of the initial studies, leaving the 148 that provided the specific data used in the meta-analysis. The specific measure the researchers extracted from each study was the odds ratio for premature mortality among individuals with weaker versus stronger social connections. The specific aggregated odds ratio across all 148 studies was 1.50 (95 percent confidence interval 1.42 to 1.59), meaning that individuals with substantially weaker social connections were, across the pooled sample, approximately 50 percent more likely to die during the study follow-up period than individuals with substantially stronger social connections. The specific magnitude of that 50 percent mortality elevation was, when translated onto the standard comparative-risk axis used for tobacco research, of similar magnitude to the mortality risk of smoking approximately 15 cigarettes per day.
What the 2010 paper actually measured
The specific technical scope of the 2010 paper is worth being precise about, since essentially every subsequent popular translation of its findings has substantially conflated two distinct concepts that the research literature does not, in fact, treat as identical. As Holt-Lunstad has subsequently clarified on her own official academic website’s specific explanation of the 15-cigarette comparison and its data source, the 2010 paper measured “social relationships” broadly — a composite construct that included both objective social integration (the number and frequency of an individual’s social contacts) and subjective perceived social support (the individual’s sense of being cared for by others). The specifically subjective psychological state of “loneliness” — the individual’s perceived deficit between the social connection they have and the social connection they want — was not the specific isolated variable that the 2010 paper measured. That specific psychological state was subsequently studied more directly in Holt-Lunstad’s 2015 follow-up meta-analytic paper “Loneliness and Social Isolation as Risk Factors for Mortality: A Meta-Analytic Review,” which examined the mortality effects of loneliness, social isolation, and living alone as three distinct measurable variables. The 2015 follow-up produced substantially similar (though slightly smaller) mortality risk estimates: 26 percent increased mortality risk for chronic loneliness, 29 percent increased mortality risk for objective social isolation, and 32 percent increased mortality risk for living alone. The specific popular framing that has, across the intervening 15 years, presented the “loneliness equals 15 cigarettes per day” claim as the specific finding of the original 2010 paper is, in a strict technical sense, a substantial simplification of what the actual research literature says. In a substantive practical sense, however, the simplification captures the specific direction and approximate magnitude of what Holt-Lunstad’s cumulative research programme has established.
What happened to the finding
The specific policy consequences of the Holt-Lunstad 2010 finding across the subsequent 15 years have been, by essentially every measure of contemporary public health institutional response, substantially non-trivial. As reported in the Holt-Lunstad team’s 2015 follow-up meta-analytic review published in Perspectives on Psychological Science, the accumulated subsequent research literature has substantially confirmed the 2010 finding’s core magnitude estimate while progressively refining its specific psychological components. In January 2018, the government of the United Kingdom under Prime Minister Theresa May formally appointed the world’s first Minister for Loneliness (Tracey Crouch), a substantive institutional response that referenced the Holt-Lunstad research programme directly in its founding public communications. In February 2021, the government of Japan under Prime Minister Yoshihide Suga appointed the world’s second Minister of Loneliness (Tetsushi Sakamoto), similarly citing the Holt-Lunstad research programme. In May 2023, the United States Surgeon General Vivek Murthy formally declared an American national loneliness epidemic in an 81-page advisory publication that cited the specific Holt-Lunstad research as the substantive empirical foundation of the advisory’s central public health claims. The World Health Organization established a Commission on Social Connection in the same year, with Julianne Holt-Lunstad appointed as one of its founding commissioners. Per the Campaign to End Loneliness’s archival documentation of Holt-Lunstad’s specific research trajectory across the 2010-2020 period, the specific 148-study meta-analysis that Julianne Holt-Lunstad and her Brigham Young University colleagues published in July 2010 has, in the 15 years since, become one of the substantially most consequential single academic psychology papers in the recorded history of the modern public health literature — measured not by academic citation count (though the paper has accumulated approximately 15,000 academic citations across the subsequent decade and a half, itself substantially exceeding the citation counts of essentially every other paper published in the same PLoS Medicine issue) but by the specific number of national government ministries, international health organisations, and public policy initiatives that the specific finding has directly produced. The specific comparison that Holt-Lunstad established between the mortality risk of social disconnection and the mortality risk of smoking 15 cigarettes per day has, across the 15 years since its 2010 publication, progressively become one of the substantially most-cited single public health statistics of the early 21st century.
