Table 3.

Odds ratio (95% confidence interval) for incident ischaemic heart disease and mortality comparing men with and without detectable IgA antibodies to Chlamydia pneumoniae, before and after adjustment for cardiovascular risk factors

Model No	Risk factors adjusted for:	All incident ischaemic heart disease	Fatal ischaemic heart disease	Total mortality
1	No adjustments	1.20 (0.87 to 1.64)	1.76 (1.19 to 2.60)**	1.56 (1.18 to 2.07)**
2	Laboratory batch	1.13 (0.80 to 1.58)	1.84 (1.21 to 2.78)**	1.59 (1.19 to 2.13)**
3	Batch, age, smoking, body mass index, total cholesterol concentration, systolic blood pressure, current social class†, father’s social class‡, height, and FEV1§	1.07 (0.75 to 1.53)	1.83 (1.17 to 2.85)**	1.50 (1.10 to 2.04)*
4	Model 2 plus past or prevalent ischaemic heart disease at entry	1.01 (0.70 to 1.44)	1.74 (1.11 to 2.73)*	1.41 (1.03 to 1.93)*
5	Model 3 plus plasma viscosity at entry	0.98 (0.68 to 1.41)	1.69 (1.06 to 2.68)*	1.36 (0.99 to 1.87)

Significance tests for detectable v undetectable IgA antibodies: ^*P<0.05, ^**P<0.01. 

Trend in risk of total mortality with increasing IgA antibody titre (zero, trace, ⩾1 in 16) was not significant at 5% level in models 3-5. 

Trend in risk of fatal ischaemic heart disease with increasing IgA antibody titre was not significant in model 5. 

^†Modelled as six levels: I and II; III non-manual; III manual; IV; V; missing. 

^‡Modelled as five levels: I and II; III non-manual; III manual; IV and V; missing or unemployed. 

^§Subjects with missing FEV₁ retained in model by use of dummy variable representing missing data.