Cerebral Microbleeds: Different Prevalence, Topography, and Risk Factors Depending on Dementia Diagnosis—The Karolinska Imaging Dementia Study

Article Figures & Data

Prevalence and odds ratios of CMBs depending on number of risk factors^{^a}

Diagnostic Group	0 RF^{^b} Prevalence	1 RF		2 RF		3 RF		≥4 RF
Diagnostic Group	0 RF^{^b} Prevalence	Prevalence	OR	Prevalence	OR	Prevalence	OR	Prevalence	OR
All patients (n = 1504)	10 (31)	18 (92)	2.1 (1.4–3.3)^{^d}	28 (105)	3.8 (2.5–5.9)^{^c}	35 (69)	5.3 (3.3–8.5)^{^c}	33 (35)	4.8 (2.8–8.4)^{^c}
Subjective cognitive impairment (n = 385)	8 (13)	10 (13)	1.2 (0.5–2.8)	11 (6)	1.3 (0.5–3.8)	24 (7)	3.6 (1.3–10.2)^{^e}	20 (2)	3.2 (0.6–17.2)
Alzheimer disease (n = 423)	7 (4)	26 (37)	4.9 (1.7–14.6)^{^d}	37 (47)	8.2 (2.8–24.4)^{^c}	33 (23)	7.1 (2.3–22.2)^{^d}	23 (7)	3.4 (0.9–13.1)
Mild cognitive impairment (n = 418)	7 (5)	17 (24)	2.8 (1.0–7.7)^{^e}	22 (25)	3.7 (1.3–10.3)^{^e}	37 (21)	7.8 (2.7–22.9)^{^c}	35 (14)	7.2 (2.2–21.3)^{^d}
Vascular dementia (n = 54)	67 (4)	50 (3)	0.5 (0.04–6.1)	58 (11)	0.6 (0.1–4.7)	58 (7)	0.5 (0.1–4.8)	64 (7)	0.8 (0.1–7.1)

Note:—RF indicates risk factors; OR, odds ratio of CMBs in all groups; Prevalence, prevalence of CMBs in all groups.
↵a Numbers are % (No.), odds ratio (95% CI). Logistic regression analysis was performed with CMBs (present/absent) as a dependent variable and number of risk factors as an independent variable. Risk factors were defined as hypertension, hyperlipidemia, diabetes, male sex, and age 65 years and older. The model was corrected for MRI field strength (1.5T/3T) and CMB sequence (T2*/SWI). In the risk factor groups, prevalence of CMBs (number of patients) is stated to the left and odds ratios for CMBs, to the right.
↵b Patients with zero risk factors were used as a reference.
↵c P < .001, significant after Bonferroni correction.
↵d P < .05, significant after Bonferroni correction.
↵e P < .05, significant before Bonferroni correction.