Cardiometabolic multimorbidity and incident dementia: the Swedish twin registry

Abstract

Abstract Aims Cardiometabolic diseases (CMDs), including diabetes, heart disease, and stroke, are established risk factors for dementia, but their combined impact has been investigated only recently. This study aimed to examine the association between mid- and late-life cardiometabolic multimorbidity and dementia and explore the role of genetic background in this association. Methods and results Within the Swedish Twin Registry, 17 913 dementia-free individuals aged ≥60 were followed for 18 years. CMDs [including age of onset in mid (60) or late (≥60) life] and dementia were ascertained from medical records. Cardiometabolic multimorbidity was defined as having ≥2 CMDs. Cox regression was used to estimate the CMD–dementia association in (i) a classical cohort study design and (ii) a co-twin study design involving 356 monozygotic and dizygotic pairs. By comparing the strength of the association in the two designs, the contribution of genetic background was estimated. At baseline, 3,312 (18.5%) participants had 1 CMD and 839 (4.7%) had ≥2 CMDs. Over the follow-up period, 3,020 participants developed dementia. In the classic cohort design, the hazard ratio (95% confidence interval) of dementia was 1.42 (1.27–1.58) for 1 CMD and 2.10 (1.73–2.57) for ≥2 CMDs. Dementia risk was stronger with mid-life as opposed to late-life CMDs. In the co-twin design, the CMD–dementia association was attenuated among monozygotic [0.99 (0.50–1.98)] but not dizygotic [1.55 (1.15–2.09)] twins, suggesting that the association was in part due to genetic factors common to both CMDs and dementia. Conclusion Cardiometabolic multimorbidity, particularly in mid-life, is associated with an increased risk of dementia. Genetic background may underpin this association.