Application of GUHA data mining method in cohort data to explore paths associated with premature death: a 29-year follow-up study

Table 2 Single predicates of premature death in total population and both sexes

Variable	“a” value			Frequency coefficient ¹			Variable	“a” value			Frequency coefficient ¹
Variable	Total	Male	Female	Total	Male	Female	Variable	Total	Male	Female	Total	Male	Female
Smoking less than 20 cigarettes a day	236	-	-	0.54		-	Incompetence to work in two years’ time	97	56	41	0.65	0.57	0.68
Smoking more than 20 cigarettes a day	287	-	-	1.00		-	Diseases affecting work which cause a sense of being unable to work	85	50	35	0.98	0.61	1.19
Alcohol use a couple of times a month or once a week	249	-	-	0.51	-	-	Feeling dizzy often	-	82	-	-	0.61	-
Worse fitness strength compared to others at the same age	115	70	45	0.71	0.64	0.69

¹ Frequency coefficient (p) expresses the prevalence of that association, the magnitude of the dependence compared to the mean. For example, if ‘frequency coefficient (p)’ is 2, then premature death is (p + 1) 3 times more common among those with the characteristic under consideration than in the entire study population

ISSN: 1471-2288