Table 3. Associations of cotinine and smoking behaviours in a series of logistic regressions.
| Model | 1 | 2 | 3 | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Smoking behaviour | OR | 95% CI | p | OR | 95% CI | p | OR | 95% CI | p |
| Ever smoked a cigarette at age 16 (n = 1794) | |||||||||
| No cotinine exposure (reference) | 1.00 | - | - | 1.00 | - | - | 1.00 | - | - |
| Passive cotinine exposure | 1.06 | 0.86, 1.30 | .60 | 0.95 | 0.76, 1.19 | .68 | 0.93 | 0.74, 1.17 | .53 |
| Active cotinine smoking | 94.97 | 13.07, 689.89 | < .001 | 70.29 | 9.55, 517.40 | < .001 | 64.31 | 8.71, 475.06 | < .001 |
| Number of cigarettes smoked by age 16 (n = 1786) | |||||||||
| No cotinine exposure (reference) | 1.00 | - | - | 1.00 | - | - | 1.00 | - | - |
| Passive cotinine exposure | 1.13 | 0.93, 1.38 | .21 | 1.01 | 0.82, 1.24 | .91 | 0.99 | 0.81, 1.22 | .96 |
| Active cotinine smoking | 70.68 | 36.62, 136.43 | < .001 | 53.46 | 27.20, 105.06 | < .001 | 49.01 | 24.84, 96.73 | < .001 |
| Active (daily/weekly) smoking at age 16 (n = 1829) | |||||||||
| No cotinine exposure (reference) | 1.00 | - | - | 1.00 | - | - | 1.00 | - | - |
| Passive cotinine exposure | 1.48 | 0.98, 2.23 | .064 | 1.27 | 0.84, 1.94 | .26 | 1.24 | 0.81, 1.89 | .33 |
| Active cotinine smoking | 43.21 | 23.12, 80.77 | < .001 | 32.34 | 16.94, 61.72 | < .001 | 28.70 | 14.93, 55.18 | < .001 |
| Ever smoked a cigarette at age 22 (n = 1307)* | |||||||||
| No cotinine exposure (reference) | 1.00 | - | - | 1.00 | - | - | 1.00 | - | - |
| Passive cotinine exposure | 1.12 | 0.88, 1.43 | .35 | 1.07 | 0.83, 1.38 | .62 | 1.05 | 0.81, 1.36 | .72 |
| Number of cigarettes smoked by age 22 (n = 1345) | |||||||||
| No cotinine exposure (reference) | 1.00 | - | - | 1.00 | - | - | 1.00 | - | - |
| Passive cotinine exposure | 1.17 | 0.95, 1.46 | .14 | 1.14 | 0.91, 1.42 | .24 | 1.12 | 0.90, 1.40 | .32 |
| Active cotinine smoking | 30.50 | 11.89, 78.24 | < .001 | 22.21 | 8.55, 57.68 | < .001 | 20.99 | 8.06, 54.67 | < .001 |
| Current (past 30 day) smoker at age 22 (n = 1345) | |||||||||
| No cotinine exposure (reference) | 1.00 | - | - | 1.00 | - | - | 1.00 | - | - |
| Passive cotinine exposure | 1.23 | 0.92, 1.64 | .17 | 1.19 | 0.88, 1.61 | 0.247 | 1.18 | 0.87, 1.59 | .29 |
| Active cotinine smoking | 8.90 | 4.61, 17.20 | < .001 | 7.20 | 3.64, 14.25 | < .001 | 6.87 | 3.46, 13.66 | < .001 |
OR = odds ratio; 95% CI = 95% confidence interval. Cotinine was treated as a categorical variable in these analyses so ORs reflect the odds of each smoking behaviour for each cotinine level. Passive exposure is defined as levels exceeding 1 ng/ml up to 9 ng/ml in blood samples. Active smoking is defined as cotinine levels exceeding 10 ng/ml in blood samples. The basic model (model 1) was adjusted for age and sex. Model 2 was additionally adjusted for socioeconomic status, BMI and alcohol. Model 3 was additionally adjusted for passive smoke exposure (maternal smoking at 12 years).
*Cotinine levels indicating active smoking at 15 are not included in the table as they perfectly predicted whether young people had ever smoked at age 22.