TABLE 5.

Association between personality measures and mortality from external causes (accidents and suicides), 1993–2003-Cox regression

		RII (95% CI) for mortality from external causes
Personality measures		Model 1†	Model 2‡	Model 3§
N event/N total††
Bortner rating scale
Type- A behaviour pattern	50/14 196	0.81 (0.31–2.12)	0.88 (0.33–2.32)
Buss-Durkee hostility Inventory
Total hostility	46/13 922	2.76 (0.99–7.69)	2.65 (0.95–7.42)
Neurotic hostility	49/14 019	3.97 (1.44–10.94)**	3.67 (1.31–10.22)**	3.24 (1.03–10.18)*
Reactive hostility	49/14 014	1.65 (0.62–4.39)	1.63 (0.61–4.32)
Grossarth-Maticek and Eysenck’s Personality-Stress Inventory
Cancer-prone-Type 1	47/14 045	1.04 (0.38–2.85)	1.00 (0.37–2.75)
CHD-prone-Type 2	47/13 984	0.97 (0.35–2.68)	0.92 (0.33–2.52)
Ambivalent-Type 3	46/13 999	2.33 (0.83–6.53)	2.13 (0.76–5.95)
Healthy-Type 4	46/14 000	0.50 (0.17–1.42)	0.50 (0.18–1.40)
Rational-Type 5	47/13 999	1.18 (0.43–3.25)	1.18 (0.43–3.25)
Anti social-Type 6	47/14 040	3.48 (1.23–9.85)**	3.40 (1.19–9.69)**	2.13 (0.69–6.58)

^*p ≤ 0.05,

^**p ≤ 0.01,

^***p ≤ 0.001

^†Model 1= RII adjusted for age and sex.

^‡Model 2= Model 1 + additionally adjusted for educational level, and marital status.

^§Model 3=Model 2 + additionally mutually adjusted for personality measures predicting mortality in Model 2

^††N events in table differ with each personality and models 1 and 2 are complete case analysis.