Table 4

Respondent and contact characteristics predictive of discussions relating to sex

Talked about sexDiscussed HIV preventionReceived
sexually transmitted infection (STI) advice
Received partner advice
Model A
  Older generation relative1.001.001.001.00
  Same-generation relative2.37 (1.24 to 4.54)1.77 (0.96 to 3.27)1.62 (0.83 to 3.17)1.19 (0.40 to 3.54)
  Romantic partner115 (26.8 to 501)20.0 (7.20 to 55.4)10.0 (3.85 to 26.1)4.29 (0.71 to 26.1)
  Non-romantic non-relative5.69 (2.76 to 11.7)4.13 (2.07 to 8.23)3.34 (1.62 to 6.87)2.61 (0.91 to 7.43)
  Respondent aged 25–34 vs 18–241.54 (0.84 to 2.83)2.22 (1.25 to 3.96)2.59 (1.36 to 4.94)1.22 (0.48 to 3.08)
  Other vs same-gender contact0.36 (0.18 to 0.71)0.30 (0.16 to 0.57)0.30 (0.15 to 0.62)0.13 (0.03 to 0.60)
  Female vs male respondent0.65 (0.35 to 1.19)0.86 (0.49 to 1.51)0.78 (0.41 to 1.46)0.55 (0.22 to 1.38)
  N383383381373
  Intraclass correlation coefficient0.1700.1320.2330.343
Model B
 Main effect
  Older generation relative1.001.001.001.00
  Same-generation relative2.21 (0.91 to 5.34)1.54 (0.67 to 3.56)1.30 (0.51 to 3.32)1.37 (0.26 to 7.21)
  Romantic partner42.0 (9.08 to 194)10.81 (3.16 to 37.0)7.97 (2.26 to 28.1)11.95 (1.39 to 102)
  Non-romantic non-relative6.25 (2.66 to 14.7)5.23 (2.31 to 11.8)3.84 (1.63 to 9.08)4.53 (1.08 to 19.0)
 Respondent aged 25–34 vs 18–24
  Older generation relative1.37 (0.52 to 3.63)2.14 (0.87 to 5.28)2.44 (0.89 to 6.71)2.82 (0.55 to 14.6)
  Same-generation relative1.21 (0.33 to 4.43)1.33 (0.38 to 4.62)1.52 (0.40 to 5.80)0.68 (0.08 to 5.80)
  Romantic partner4.77 (0.61 to 37.0)1.60 (0.30 to 8.44)0.09 (0.00 to 1.89)
  Non-romantic non-relative0.72 (0.17 to 2.98)0.39 (0.10 to 1.57)0.54 (0.13 to 2.26)0.27 (0.03 to 2.08)
  Other vs same-gender contact0.35 (0.17 to 0.69)0.28 (0.15 to 0.54)0.29 (0.14 to 0.60)0.13 (0.03 to 0.60)
  Female vs male respondent0.63 (0.34 to 1.18)0.84 (0.47 to 1.51)0.78 (0.41 to 1.46)0.53 (0.21 to 1.37)
  N351383381373
  Intraclass correlation coefficient0.1920.1540.2370.349
  • Values are from two-level hierarchical logistic regressions containing varying numbers of contacts nested within 118 respondents. All models exclude five 'younger generation relative' and two 'unspecified relative' contacts, four contacts who declined to answer all these questions and one respondent with no contacts. Models for receipt of STI and partner advice have additional missing observations for question-specific non-response. Each regression for Model B contains main effects and age interaction effects for each type of relative. Model B for ‘talked about sex’ dropped observations for romantic partners of 25–34 year olds since all 32 had discussed sex with the respondent.