Table 4A

Whitehall II: predicting SRH using stepwise regression†

†Stepping criteria: p = 0.001 for entry, p = 0.01 for exit and r² change associated with the variable to be >0.01.
Men
	Overall model	r²_adjusted = 0.330	F_{5, 4846} = 478.86; p<0.0001
		r² change	Change statistics ( F)
1	Symptom score	0.154	884.72; p<0.0001
2	Sickness absence	0.096	621.88; p<0.0001
3	Longstanding illness	0.040	272.73; p<0.0001
4	Affect balance score	0.029	205.18; p<0.0001
5	Number of recurring health problems	0.012	83.53; p<0.0001
Women
	Overall model	r²_adjusted = 0.350	F_{5, 2190} = 237.59; p<0.0001
		r² change	Change statistics ( F)
1	Symptom score	0.197	537.29; p<0.0001
2	Sickness absence	0.081	244.29; p<0.0001
3	Longstanding illness	0.038	120.38; p<0.0001
4	Minor psychiatric morbidity: GHQ	0.025	81.31; p<0.0001
5	Number of recurring health problems	0.012	41.55; p<0.0001
Men and women
	Overall model	r²_adjusted = 0.347	F_{5, 7038} = 747.94; p<0.0001
		r² change	Change statistics ( F)
1	Symptom score	0.181	1554.21; p<0.0001
2	Sickness absence	0.091	883.45; p<0.0001
3	Longstanding illness	0.039	398.02; p<0.0001
4	Minor psychiatric morbidity: GHQ	0.024	250.39; p<0.0001
5	Number of recurring health problems	0.012	131.94; p<0.0001