Table 1. Participants' characteristics in the Women's Health Initiative Clinical Trial and Observational Study by statin use (current vs never)a at time of cancer diagnosis (N=17 285b).

	Current statin use (N=4025)		Never used statins (N=13 260)
	N	%	N	%	P-valuec
Age at screening (years)					<0.001
50–59	1123	27.9	3973	30.0
60–69	2089	51.9	6228	47.0
70–79	813	20.2	3059	23.1
Age at incident cancer					<0.001
<70	1338	33.2	6706	50.6
70 to <80	2030	50.4	5295	39.9
⩾80	657	16.3	1259	9.5
Tumour stage					0.001
In situ	613	15.9	1732	13.8
Local	1924	49.8	6203	49.4
Regional	702	18.2	2552	20.3
Distant	623	16.1	2079	16.5
Race/ethnicity					0.15
White	3517	87.4	11 701	88.2
Black	261	6.5	832	6.3
Hispanic	86	2.1	310	2.3
American Indian	16	0.4	38	0.3
Asian/Pacific Islander	92	2.3	234	1.8
Unknown	53	1.3	145	1.1
Education					<0.001
High school/GED or less	899	22.5	2617	19.9
School after high school	1534	38.3	4812	36.5
College degree or higher	1570	39.2	5746	43.6
BMI, baseline (kg m−2)					<0.001
<25	1065	26.7	4745	36.1
25 to <30	1474	37.0	4468	33.9
30 to <35	882	22.1	2427	18.4
⩾35	568	14.2	1522	11.6
Smoking status					0.75
Never	1829	46.0	6115	46.7
Past	1817	45.7	5904	45.1
Current	326	8.2	1074	8.2
Vitamin D intake (IU)					0.97
<200	1418	36.2	4690	36.1
200 to <400	737	18.8	2483	19.1
400 to <600	981	25.0	3229	24.9
⩾600	784	20.0	2577	19.9
Alcohol intake					<0.001
Non/past drinker	1121	28.1	3252	24.7
<1 drink per week	1376	34.5	4367	33.2
1–<7 drinks per week	986	24.7	3656	27.8
⩾7 drinks per week	511	12.8	1890	14.4
Physical activity (MET—min)					0.03
<100	824	21.7	2683	21.6
100 to <500	1127	29.7	3420	27.5
500 to <1200	1081	28.4	3620	29.1
⩾1200	768	20.2	2709	21.8
Has current health-care provider	3845	96.1	12 401	94.3	<0.001
Mammogram within last 2 years	3465	88.1	10 822	84.1	<0.001
Hysterectomy at randomisation	1526	37.9	4702	35.5	0.005
Unopposed oestrogen use status					0.64
Never used	2668	66.4	8752	66.1
Past user	509	12.7	1752	13.2
Current user	843	21.0	2746	20.7
Oestrogen+progesterone use status					<0.001
Never used	2934	72.9	9166	69.2
Past user	340	8.5	1123	8.5
Current user	748	18.6	2964	22.4
Age at menarche					0.31
<12	894	22.3	2900	22.0
12–13	2262	56.4	7336	55.6
⩾14	856	21.3	2970	22.5
Oral contraceptive use ever	1672	41.5	5451	41.1	0.63
CHD before cancer diagnosis	680	17.0	756	5.7	<0.001
Diabetes before cancer diagnosis	725	18.0	865	6.5	<0.001
Family history of cancer	2639	68.4	8888	69.8	0.12
Aspirin use	1136	28.2	2573	19.4	<0.001
NSAIDs	1684	41.8	4478	33.8	<0.001
DM trial					0.03
Comparison	1011	62.8	2985	59.6
Intervention	600	37.2	2020	40.4
CEE+MPA trial					0.007
Comparison	283	52.7	819	46.1
Intervention	254	47.3	957	53.9
CEE trial					0.04
Comparison	179	56.5	423	49.6
Intervention	138	43.5	430	50.4
CaD trial					0.15
Comparison	638	53.1	1856	50.7
Intervention	563	46.9	1803	49.3
Clinical trial participant	2218	55.1	6830	51.5	<0.001
	Mean	(s.d.)	Mean	(s.d.)	P-value
Age (years)	63.6	(6.5)	63.7	(7.0)	0.37
Fruit and vegetable consumptiond	4.0	(4.0)	4.1	(2.1)	0.004
Red meat consumptiond	0.7	(0.5)	0.7	(0.6)	0.27
General health (0 worst–100 best)	73.6	(17.2)	75.9	(16.7)	<0.001

Abbrviations: BMI=body mass index; CaD=calcium + Vitamin D; CEE=conjugated equine estrogen; CHD=coronary heart disease; DM=dietary Modification; GED=general education development; MET=metabolic equivalent; MPA=medroxyprogesterone acetate; NSAID=nonsteroidal anti-inflammatory drugs.

^aStatin (current vs never) were the exposure groups of interest for the primary analysis.

^bAt the time of incident cancer, 23 067 participants were at risk for death and included in the primary analyses. Of these, 4025 participants were currently taking statins; 13 260 participants never used statins; 397 participants had reported using statins at baseline or follow-up, but were not currently taking statins; and 5385 participants had medication inventories that were out of date, so their follow-up was censored. Exposure groups were modelled as a time-dependent exposure, so a participants' group status may change during follow-up (e.g., participants with an out-of-date inventory were allowed to re-enter the model when a current medications inventory was collected).

^cOn the basis of χ²-test of association for categorical variables and t-test for continuous variables.

^dMedium servings per day.