Chronic diseases (e.g., hypertension, diabetes, and heart diseases) have been proposed as marked predictors of incident dementia. However, synthesised evidence on the effect of multimorbidity on dementia is still lacking. We aim to summarise the association between multimorbidity and risk of dementia in longitudinal cohorts.

In this systematic review and meta-analysis, we conducted a systematic search in PubMed, Web of Science and Embase from inception to Dec 14, 2024, to identify longitudinal cohort studies reporting the association between multimorbidity or multimorbidity patterns and risk of dementia. Information of included studies were extracted by three reviewers (YaZ, YY and YuZ), and the quality assessment was conducted using the Newcastle-Ottawa Scale. The inverse-variance weighted random effects meta-analysis was performed to obtain the pooled hazard ratios (HRs) and 95 % confidence intervals (CIs) for dementia associated with multimorbidity and cardiometabolic multimorbidity (CMM). Cochran's Q test and the I² statistic were used to indicate heterogeneity among the studies. Meta-regression analysis, subgroup analysis and sensitivity analysis were conducted to determine any valid sources of heterogeneity. This study was registered with PROSPERO (CRD42023403684).

We included 17 longitudinal cohort studies (2262,885 middle-aged and older participants) in the systematic review, of which seven were included in meta-analysis. All studies presented moderate to high methodological quality. Meta-analysis showed a positive association between multimorbidity and incident dementia (HR=1.53, 95 % CI=1.12 to 2.09), with substantial heterogeneity (I²=95.2 %). Studies using health records to measure dementia tend to find a stronger positive relationship between multimorbidity and risk of dementia than those using self-report (HR_{health records}=1.94, 95 % CI=1.35 to 2.78, I²=94 %; HR_self-report=1.17, 95 % CI=1.07 to 1.28, I²=0 %). The impacts of CMM were also observed, and the HRs for dementia ranged from 2.49 (combination of heart diseases and stroke: 95 % CI=1.64 to 3.78) to 3.77 (combination of diabetes, heart diseases and stroke: 95 % CI=2.02 to 7.02). The heterogeneity was moderate, with I² ranging from 46.9 % (p for heterogeneity=0.152) to 84.1 % (p for heterogeneity=0.002). The impacts of number of diseases, multimorbidity clusters, and multimorbidity trajectory on risk of dementia were narratively summarised due to lacking comparable studies. Limited evidence (only one study) precluded quantitative synthesis for the association of physical and psychological multimorbidity with dementia.

Multimorbidity and CMM pattern were significantly associated with risk of dementia, while the effect of physical and psychological multimorbidity remain inconclusive. Individuals affected by multimorbidity should be prioritised in risk factor modification and dementia prevention. Preventing the development of multimorbidity is also crucial—particularly those who already have one chronic disease—in order to maintain cognitive health.