The impact of direct-to-consumer personal genomic testing on perceived risk of breast, prostate, colorectal, and lung cancer: findings from the PGen study

Table 4 Linear regression and generalized estimating equation models for effect of genetic risk estimate on change in perceived risk of cancer

	Breast	Prostate	Colorectal	Lung^c
Change in perceived risk: baseline to 2W^a
Elevated risk result: LS mean (95 % CI)	0.61 (0.42, 0.79)	0.77 (0.58, 0.95)	0.50 (0.41, 0.60)	0.62 (0.49, 0.75)
Average risk result: LS mean (95 % CI)	−0.20 (−0.27, −0.14)	−0.21 (−0.31, −0.11)	−0.05 (−0.10, 0.01)	0.18 (0.10, 0.27)
LS mean difference (95 % CI)	0.81 (0.62, 1.00)	0.97 (0.76, 1.19)	0.55 (0.44, 0.66)	0.44 (0.31, 0.57)
p-value_Difference	<0.0001	<0.0001	<0.0001	<0.0001
Change in perceived risk: baseline to 6M^a
Elevated risk result: LS mean (95 % CI)	0.53 (0.32, 0.74)	0.51 (0.32, 0.69)	0.33 (0.24, 0.43)	0.58 (0.44, 0.72)
Average risk result: LS mean (95 % CI)	−0.15 (−0.22, −0.07)	−0.14 (−0.24, −0.04)	−0.04 (−0.09, 0.02)	0.22 (0.13, 0.31)
LS mean difference (95 % CI)	0.68 (0.46, 0.90)	0.65 (0.43, 0.86)	0.37 (0.26, 0.48)	0.36 (0.22, 0.49)
p-value_Difference	<0.0001	<0.0001	<0.0001	<0.0001
GEE Model: ΔPR^b
Elevated risk result: β ± SE (p-value)	0.68 ± 0.29 (0.02)	1.06 ± 0.20 (<0.0001)	0.79 ± 0.11 (<0.0001)	0.59 ± 0.13 (<0.0001)
Elevated risk result*survey: β ± SE (p-value)	0.06 ± 0.19 (0.76)	−0.15 ± 0.11 (0.17)	−0.20 ± 0.07 (0.0027)	−0.11 ± 0.08 (0.17)

2 W 2-week follow-up, 6 M 6-month follow-up, LS least squares adjusted, CI confidence interval, SE standard error
^aAdjusted for baseline perceived risk, age, gender, race/ethnicity, education, and company
^bAdjusted for baseline perceived risk, age, gender, race/ethnicity, education, and company, with result*survey interaction term
^cMultivariate lung cancer analyses additionally adjusted for smoking status