Table 2.

Risk for dementia associated with mid- and late-life adiposity.

	N dementia/N	Minimally adjusted modela		Fully adjusted modelb
		HR (95% CI)	P	HR (95% CI)	P
Adiposity measured at 50–64 years	296/9 222
BMI
Underweight	12/437	0.95 (0.52 to 1.72)	0.86	0.92 (0.51 to 1.67)	0.79
Normal weight	129/4729	Ref	-	Ref	-
Overweight	106/2933	1.41 (1.08 to 1.82)	0.01	1.41 (1.09 to 1.83)	0.01
Obesity	49/1123	1.76 (1.26 to 2.46)	<0.01	1.77 (1.26 to 2.47)	<0.01
Waist circumference
No abdominal obesity	138/5111	Ref	-	Ref	-
Abdominal obesity	158/4111	1.54 (1.21 to 1.95)	<0.01	1.56 (1.23 to 1.98)	<0.01
Waist-to-height ratio
No abdominal obesity	60/2521	Ref	-	Ref	-
Abdominal obesity	236/6701	1.55 (1.15 to 2.07)	<0.01	1.55 (1.16 to 2.08)	<0.01
Adiposity measured at ≥ 65 years	327/3806
BMI
Underweight	44/468	1.21 (0.87 to 1.69)	0.25	1.22 (0.88 to 1.70)	0.24
Normal weight	176/2278	Ref	-	Ref	-
Overweight	82/1046	1.07 (0.82 to 1.40)	0.60	1.07 (0.82 to 1.40)	0.62
Obesity	25/341	1.14 (0.74 to 1.75)	0.55	1.14 (0.74 to 1.74)	0.56
Waist circumference
No abdominal obesity	178/2022	Ref	-	Ref	-
Abdominal obesity	149/1784	1.12 (0.89 to 1.41)	0.34	1.11 (0.88 to 1.40)	0.36
Waist-to-height ratio
No abdominal obesity	88/775	Ref	-	Ref	-
Abdominal obesity	239/3031	0.97 (0.75 to 1.25)	0.82	0.97 (0.75 to 1.25)	0.80

BMI body mass index group.

Cox model was used to separately test the age-specific association between each adiposity measure and subsequent dementia. Adiposity was measured at baseline for each participant. The distribution of age at adiposity measurement and dementia onset is shown in Supplementary Fig. 1.

^aMinimally adjusted model: baseline age, sex, and education level were adjusted.

^bFully adjusted model: baseline age, sex, education level, smoking frequency, and drinking frequency were adjusted.