Table 3.

Summary of findings for influence of sunlight exposure on risk of dementia

Population: General population aged between 54 and 85.3 years (median/mean) Settings: Health examinations, geriatric hospitals Exposure/Risk: Vitamin D deficiency Comparison: No vitamin D deficiency (reference group)
Outcomes	Comparative risk (95% CI) and narrative results		Relative effect (95% CI)	Number of participants (number of studies)	Quality of evidence (GRADE)	Comments
	No exposure (no vitamin D deficiency)	Exposure (vitamin D deficiency)
Incidence of dementia (Results meta-analysis) Follow-up:18.03 years (weighted mean)	Study population (≥50 nmol/L or ≥ 54–159 nmol/L)	(<25 nmol/L or 7–28 nmol/L)	Point (raw) 1.54 (1.19 to 1.99)	18 639 subjects (5 studies)	⊕⊝⊝⊝ Very lowa,b
Incidence of dementia (Narrative results) Follow-up: 7 years	Study population (≥25 nmol/L)	(<25 nmol/L)	OR 19.57 (1.11 to 343.69)	40 subjects (1 study)

Abbreviations: CI confidence interval, nmol/L nanomoles per Litre, OR odds ratio

^aRisk of bias: Adjustment for confounders varies across study groups, Apo E ε4 genoytpe only considered in one study

^bIndirectness: no study investigated the direct relationship between solar radiation and dementia, vitamin D is a surrogate parameter