Table 4.

Serum PSA concentration (geometric mean and 95% confidence intervals) in men 40 years and older by cholesterol-lowering drug use and stratified by health status, NHANES 2001 – 2004

	No Co-morbiditiesc		Co-morbiditiesc		p interactionb
	N	Multivariable Adjusteda	N	Multivariable Adjusteda
Current Use of Cholesterol-lowering Drugsd
No	1,488	0.99 (0.95 – 1.04)	588	0.83 (0.75 – 0.91)	0.003
Yes	192	0.86 (0.77 – 0.96)	306	0.91 (0.82 – 1.01)
p-valueb		0.02		0.14
Cholesterol-lowering Drug Use in the Past Monthe
No	1,503	0.99 (0.94 – 1.04)	541	0.83 (0.76 – 0.90)	0.14
Yes	177	0.91 (0.80 – 1.02)	353	0.90 (0.80 – 1.01)
p-value‡		0.19		0.25
<1 year	49	0.83 (0.71 – 0.98)	84	0.91 (0.72 – 1.17)	0.09
1 – <5 years	93	0.99 (0.82 – 1.19)	167	1.01 (0.83 – 1.23)
≥ 5 years	31	0.80 (0.52 – 1.21)	96	0.73 (0.65 – 0.83)
p trendb		0.32		0.32
Statin Drug Use in the Past Monthe
No	1,518	0.99 (0.94 – 1.04)	573	0.84 (0.77 – 0.91)	0.16
Yes	162	0.90 (0.80 – 1.01)	321	0.89 (0.79 – 1.01)
p-value‡		0.14		0.41
<1 year	44	0.83 (0.70 – 1.00)	73	0.92 (0.71 – 1.18)	0.16
1 – <5 years	86	0.96 (0.80 – 1.15)	150	1.00 (0.80 – 1.24)
≥ 5 years	29	0.82 (0.54 – 1.25)	93	0.72 (0.64 – 0.81)
p trendb		0.34		0.26

^aAdjusted for age in years (continuous), race/ethnicity, body mass index, and cigarette smoking status

^bP-values calculated modeling the natural logarithm transformed PSA

^cMen with co-morbidities have one or more of the following: cancer, myocardial infarction, stroke, angina, coronary heart disease, congestive heart failure, or diabetes.

^dParticipants were asked if they were currently taking medications to lower cholesterol; 91% of users took a statin.

^eParticipants were asked whether they had taken any prescription medications in the previous 30 days. Interviewers recorded type of medication as well as how long participants had been taking each medication.