Table 2. Adjusted hazard ratios from final multivariable Cox regression model predicting 10-year mortality in the training cohort (n = 376,971).

Predictor Variables	Hazard Ratio	P-Value	95% Confidence Interval
			Lower Confidence Limit	Upper Confidence Limit
Gender
Female	Ref	—	—	—
Male	2.17	< 0.001	2.08	2.27
Log age	44.00	< 0.001	36.35	53.27
Qualifications
None	Ref	—	—	—
College/University	0.75	< 0.001	0.71	0.80
A/AS Levels	0.83	< 0.001	0.77	0.89
O Levels/GCSEs	0.81	< 0.001	0.76	0.85
CSEs	0.88	0.020	0.79	0.98
NVQ/HND/HNC	0.80	< 0.001	0.74	0.87
Other professional qualifications	0.79	< 0.001	0.73	0.87
Unknown	0.97	0.590	0.86	1.09
Ethnicity
White	Ref	—	—	—
South Asian	0.59	< 0.001	0.49	0.70
East Asian	0.67	0.110	0.41	1.09
Black	0.63	< 0.001	0.51	0.78
Other/Mixed	0.81	< 0.030	0.66	0.98
Unknown	1.00	0.990	0.78	1.27
Previous diagnosis of cancer
No	Ref	—	—	—
Yes	2.58	< 0.001	2.47	2.71
Previous diagnosis of CHD
No	Ref	—	—	—
Yes	1.58	< 0.001	1.49	1.68
Previous diagnoses of T2DM
No	Ref	—	—	—
Yes	1.72	< 0.001	1.62	1.83
Previous diagnosis of COPD
No	Ref	—	—	—
Yes	1.87	< 0.001	1.71	2.05
Smoking
Non-smoker	Ref	—	—	—
Current smoker	2.01	< 0.001	1.91	2.11
Log diastolic blood pressure	0.73	< 0.001	0.60	0.89
Log systolic blood pressure	1.14	0.200	0.94	1.38
Log Townsend deprivation index	1.13	< 0.001	1.10	1.17
Log body mass index	1.12	0.070	0.99	1.27
Log MET-min per week	0.94	< 0.001	0.93	0.95
Log forced expiratory volume 1	0.53	< 0.001	0.51	0.56