Table 3. Adjusted geometric mean hormone levels by estimated number of aspirin tablets per month in 1988 among postmenopausal women in the Nurses' Health Study^*.

		Estimated aspirin tablets/month†					Percent difference‡	p -value for trend§
	N	0	1-10	11-30	31-50	50+
Maximum N	716	276	247	100	25	68
Estradiol, pg/mL	694	7.0	6.9	7.2	6.7	6.6	-4.5	0.45
Free estradiol, pg/mL	663	0.10	0.10	0.10	0.10	0.09	-5.8	0.39
Estrone, pg/mL	545	24.7	25.2	24.2	23.4	22.7	-7.9	0.04
Estrone sulfate, pg/mL	687	204	216	207	213	179	-12.1	0.03
Testosterone, ng/dL	697	21.3	21.9	21.1	19.3	20.5	-3.7	0.08
Free testosterone, ng/dL	681	0.20	0.22	0.20	0.19	0.19	-6.0	0.12
Androstenedione, ng/dL	545	55.4	56.8	55.5	53.3	53.0	-4.4	0.07
Ratio of estrone/androstenedione	522	0.47	0.48	0.45	0.45	0.47	-0.02	0.68
Ratio of estradiol/testosterone	659	0.35	0.35	0.36	0.40	0.34	-1.0	0.72

^*Adjusted for age at blood draw, laboratory batch, fasting status at blood draw, date and time of blood draw, parity, age at first birth, BMI at blood draw, physical activity, smoking history, duration of postmenopausal hormone use among past users, age at menopause, and alcohol intake

^†Calculated as the midpoint of each frequency category (0, 2.5, 9.5, 18, or 26 days/month) multiplied by the midpoint of each category of usual number of tablets taken on each day of use (0, 1, 2, 3.5, 5.5, or 7.5 tablets)

^‡Percent difference for highest versus lowest category of aspirin use (50+ vs. 0 tablets/month), calculated using (e^β-1)*100; no statistically significant differences

^§Weighted by the midpoint of each frequency category (0, 2.5, 9.5, 18, or 26 days/month) multiplied by the midpoint of each category of usual number of tablets taken on each day of use (0, 1, 2, 3.5, 5.5, or 7.5 tablets); calculated using the Wald test