Table 3

Final logistic regression model for all ulcers. Helicobacter pylori infection, lifestyle practices, and medication at study entry‡ (odds ratios and 95% confidence intervals)*

Baseline exposure (reference category)Multivariate adjusted
Odds ratio95% CI
*Different logistic regression models were constructed as some variables were inter-related and could not be fitted into the same model.
†In former and current smokers.
‡Medication: daily, weekly, or monthly previous or current use.
Anti-Hp IgG sero-status
Anti-Hp changes within the observation period
    (Persistently seronegative)1.0
    Persistently borderline2.20.9;5.3
    Persistently seropositive3.61.8;7.4
Cumulative tobacco consumption
    Former smoker2.00.7;5.9
    1–14.99 g/day4.01.6;10.4
    15–24.99 g/day4.51.6;11.0
    ≥25 g/day3.00.8;11.9
Cumulative number of cigarettes smoked daily†
    1–9 cigarettes1.50.6;3.5
    10–14 cigarettes1.30.5;3.1
    15–20 cigarettes2.01.0;4.1
    ≥21 cigarettes2.50.9;6.9
Wine (weekly consumption)
    (Never drink wine)1.0
    1–3 glasses of wine0.50.3;1.2
    ≥4 glasses of wine0.60.2;1.3
Spirits (weekly consumption)
    (Never drink spirits)1.0
    1–2 glasses of spirits1.80.9;3.6
    ≥3 glasses of spirits1.80.8;4.3
Cumulated number of drinks (weekly consumption)
    (0–2 drinks)1.0
    3–5 drinks1.20.6;2.5
    6–12 drinks0.50.2;1.1
    ≥13 drinks0.90.4;1.8
Leisure time energy expenditure
Uspecified gastrointestinal drugs at study entry in 1982
    (Never-infrequent use)1.0
    Previous use1.20.4;4.3
    Current use3.21.7;6.2
NSAID use at study entry in 1982
    (Never-infrequent use)1.0
    Previous use0.80.3;2.3
    Current use0.40.1;2.3
Minor tranquillisers use at study entry in 1982
    (Never-infrequent use)1.0
    Previous use1.90.2;15.0
    Current use3.01.4;6.6