Table 4

Multivariable analyses of protective behaviors

VariableProtective behaviorspc
Self-careavs. nothingConventional careavs. nothing
ORb95% CIbORb95% CIb
Study0.311
 Study 11.00Reference1.00Reference
 Study 20.400.06–2.450.280.06–1.41
Age (yrs)0.373
 40–491.060.41–2.710.540.24–1.19
 50–591.00Reference1.00Reference
 60–690.940.36–2.460.900.42–1.91
Race/ethnicity0.193
 Nonwhite1.00Reference1.00Reference
 White0.310.07–1.330.800.24–2.67
Place of birth0.930
 Philadelphia1.00Reference1.00Reference
 Outside of Philadelphia1.110.53–2.331.120.61–2.07
Education (yrs)0.002
 ≤121.00Reference1.00Reference
 >121.470.66–3.312.921.50–5.70
Marital status0.308
 Not married1.00Reference1.00Reference
 Married0.950.45–2.011.440.76–2.72
Salience and coherence (1-point increase)d0.008
 Study 1e0.430.14–1.320.660.23–1.88
 Study 2e1.270.25–6.436.871.86–25.5
Worry and concern (1-point increase)d0.050
1.930.92–4.060.920.50–1.68
Baseline survey mode0.002
 Telephone1.00Reference1.00Reference
 Mail4.201.63–10.831.310.58–2.95
  • a “Self-care” includes men with self-care only. “Conventional care” includes men with conventional care only and men with conventional care and self-care.

  • b OR, odds ratio; CI, confidence interval.

  • c p-values test the difference of protective behaviors across levels of a variable, adjusting for all other variables in the model (likelihood ratio test).

  • d Scales used as continuous (scored from 1 to 4). Scoring for the worry and concern scale was reverse-coded. Thus, a higher scale score reflects lower worry and concern.

  • e The association of salience and coherence with protective behaviors was significantly different across the two studies (p = 0.012).