Table 2.

Age- and Sex-Adjusted Distribution of Demographic and Health-Related Characteristics by Brain MR Imaging Pattern^a

Characteristic	No Lesions	Focal Lesions	Diffuse WMLs
Follow-up time to HF or censoring, y	274±6.5	140±6.3	406±6.4
Age, y	70.7±5.5	72.6±6.5b	75.5±6.9c,d
Age <80 y	92	84	71c,d
Male sex	45	46	38c,d
Education, y	4.6 ±2.1	4.6±2.5	4.3± 2.1c
Body mass indexe	27.9±4.1	27.4±4.3	26.9±4.3
Smoking status
Never	62	55	60d
Former	31	35	29
Current	7	10	11
Physical activity
Sedentary	8	11	15c
Intermediate	61	67	61
Active	31	23	25
Blood pressure, mm Hg
Systolic	144.7 ± 18.6	146.9 ± 19.9	150.5 ± 19.9c
Diastolic	78.8 ± 11.2	78.7 ± 11.9	79.2 ± 11.8
Orthostatic hypotension	18	15	18
Ever fell during past year	28	24	30
No. of falls during past year	0.5±1.0	0.44±0.95d	0.7±1.7c,d
Disease status
Previous HF	2.5	2.8	2.7
Other fractures after age 50 y	29	33	28
Stroke	1	6b	4c
Osteoporosis	46	45	50c
Diabetes mellitus	9	7	8
Any urinary incontinence	17	25	24c
Physical and cognitive function
Activities of daily living disability	6	12	18c,d
Short physical performance battery score	9.4±2.3	9.0±2.6d	8.1±2.8c,d
Mini-Mental State Examination score	25.8±3.3	24.8±3.9b	23.6±4.6c
No. of medications	2.8±2.3	3.1±2.6	3.5±2.5c,d
Taking high blood pressure medication	29	39b	28d

Abbreviations: HF, hip fracture; MR, magnetic resonance; WMLs, white matter lesions.

^aData are given as mean±SD or as percentages unless otherwise indicated. Percentages may not total 100% owing to rounding.

^bP<.05 comparing focal lesions with no lesions.

^cP<.05 comparing diffuse WMLs with no lesions.

^dP<.05 comparing diffuse WMLs with focal lesions. From linear regression models for continuous variables, from logistic regression models for dichotomous variables, and from polynomial logistic regression models for multilevel variables.

^eCalculated as weight in kilograms divided by height in meters squared.