This study examined if loneliness at age 70 was prospectively associated with cardiovascular- and all-cause mortality over a 12-year follow-up period, after controlling for initial health status, health behaviors, depression, and other social factors. We found that loneliness was an independent predictor of cardiovascular mortality in women, while there was no significant association with all-cause mortality. Among men, we found no evidence of an association between loneliness and subsequent death whether cardiovascular- or all-cause both in unadjusted or multivariable-adjusted models.
The strengths of the study included the population-based samples of both men and women and the comprehensive examinations data on socio-demographic factors, medical history, and clinical- and physical measurements, including loneliness status. However, this study also had a number of limitations. First, although the response rate at baseline was 70%, which is higher than in many other studies [25], we cannot exclude the possibility that participants were healthier than non-participants. Although a short health survey in non-participants revealed that participants and non-participants were similar in terms of health history and health related behaviors, married men were overrepresented among participating men [17]. This may imply that lonely men are underrepresented in the sample, as unmarried status is strongly associated with loneliness [19], which may have caused selection bias in our study. Another limitation was the definition of loneliness, which was assessed with a single-item question. This may result in underreporting due to the stigma associated with being identified as lonely [2, 26]. Therefore, this single measurement may not capture the overall influence of loneliness on mortality. This single item question of loneliness, however, is most common and a widely used measure[3], which previously has been shown to predict mortality [5, 7, 9].
As expected, feeling lonely was more common in women than in men. A recent review concluded that women are more likely to report loneliness than their male counterparts regardless of country studied and the classification of loneliness used [19]. One possible explanation for the greater loneliness experienced by older women may be that they are more willing to admit socially unacceptable feelings than men [1], and that disclosing loneliness may be more socially accepted in women than in men [27].
Findings across previous studies on loneliness and all-cause mortality are inconsistent as to whether loneliness independently predicts mortality risk after adjusting for initial health status, health behaviors, depression, and other social factors. The lack of association between loneliness and all-cause mortality found in our present study is similar to several previous studies conducted in Amsterdam [11], France [7], England [10], and China [5]. Two studies conducted in the USA, however, reported that loneliness was an independent predictor of all-cause mortality after adjusting for prior health and health behavior and depression [4, 8]. Similarly, a recent meta-analysis reported that loneliness is associated with all-cause mortality in both gender but this effect is slightly stronger in men than in women [28]. On the other hand, the Amsterdam study of the elderly (AMSTEL) on 4004 older men and women aged 65-84 years have shown that loneliness is an independent risk factor for all-cause mortality in men but not in women [29]. These inconsistent findings across studies may be related to varying study designs, sample size, methods and follow-up periods, but also differences in cultural settings.
Studies concerning long-term longitudinal associations between loneliness and cardiovascular death are scarce. To our knowledge, our study is the 2nd to examine loneliness in relationship to cardiovascular death in a population-based sample of men and women. The first study was conducted using UK biobank data on 466 901 men and women by Elovainio M, et. al. and reported that loneliness was not independently associated with cardiovascular mortality for both genders in multivariable adjusted models [6]. Our study result partly contradicts the study by Elovainio M, et. al., as we observed an independent association between loneliness and cardiovascular mortality in women but not in men. Possible mechanisms by which loneliness contribute to cardiovascular mortality, which is observed in our study, may be that loneliness affects cardiovascular health by altering biomarkers and shaping health behavior that are associated with increased CVD risks. For example, loneliness has been associated with elevated blood pressure [18], elevated triglycerides level [20], CHD [30], smoking, and physical inactivity [19]. In our study, no associations were found in women between loneliness and baseline SBP, DBP, CHD, triglyceride level or most of the prior health variables including smoking and physical activity. However, women with loneliness more often reported poor economic status, chronic bronchitis, and living alone, and were more often diagnosed with depression, compared to those without loneliness. Loneliness predicted cardiovascular mortality in women after adjusting for the associated factors suggesting that loneliness alters physiology at a more fundamental level. Future research should include efforts to examine how physiological processes contribute to the effect of loneliness on mortality.
In conclusion, we found that loneliness was an independent predictor of cardiovascular mortality in women, while there was no evidence to indicate that loneliness was associated with an increased risk of either cardiovascular- or all-cause mortality in men. Our results emphasizes the importance of considering women with loneliness as a high-risk group to target for public health and medical care efforts in reducing cardiovascular mortality.