Table 3C

Men	Difference between low and high educated in absolute mortality decline (deaths per 100 000)
Men	All causes	All cardio- vascular	Ischaemic heart disease	Cerebro- vascular disease	All cancer	Lung cancer	All other diseases	Liver cirrhosis	All external	Road traffic accidents
(C) Absolute mortality decline, men
North
Finland	75.3	61.3	61.8	12.7	13.8	17.3	−3.7	−14.5	−5.5	−1.9
Sweden	48.4	52.7	42.3	5.2	2.9	0.2	−5.1	−2.3	−11.1	−1.9
Norway	70.8	98.2	77.7	15.1	−12.7	−6.1	−7.3	−0.4	−12.7	−0.9
Denmark	−21.3	14.8	25.9	−1.9	−14.4	4.8	−18.7	−8.5	7.1	0.7
West
England & W	108.4	59.0	48.9	9.7	16.0	18.6	31.2	−3.3	8.0	−0.7
Belgium	21.6	17.4	7.2	5.0	14.7	22.8	6.6	−1.1	−8.9	−3.7
France	62.9	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
Switzerland	170.2	66.6	32.6	9.3	44.1	25.7	26.1	3.8	22.8	−0.9
South
Barcelona	50.8	−7.8	−6.3	−2.5	25.4	13.1	35.1	−1.2	−1.9	0.6
Basque C	4.4	−13.7	−7.9	−9.6	8.9	−1.5	3.8	0.2	5.3	−1.3
Madrid	33.0	−7.9	−10.2	−9.5	11.9	−5.4	35.1	−1.7	−6.1	−3.5
Turin	109.3	42.0	19.8	22.1	10.5	−1.3	55.3	−0.2	6.6	3.2
East
Hungary	−309.5	−157.5	−110.6	−12.6	−179.6	−64.6	−39.6	−42.8	67.2	17.0
Lithuania	−802.6	−223.4	−120.1	−26.4	−81.3	−20.3	−186.8	n.a.	−311.1	−20.4
Estonia	−821.2	−304.4	−161.8	−48.6	−87.9	−14.5	−208.0	n.a.	−220.9	−2.9

More mortality decline (or less mortality increase) among low than among high educated.
Less mortality decline among low than among high educated.
Mortality increase among low educated, mortality decline (or less mortality increase) among high educated.
Difference between low and high educated in absolute mortality decline was calculated by subtracting the absolute mortality decline among the high educated from the absolute mortality decline among the low educated. In algebraic form: (R_1990,L − R_2000,L) − (R_1990,H − R_2000,H) in which R=age-standardized mortality rate, 1990=1990s, 2000=2000s, L=low educated, and H=high educated. For example, in the case of Finnish men, mortality declined by 361.9 and 286.6 deaths per 100,000 among the low and the high educated, respectively, and thus by 75.3 deaths per 100,000 more among the low educated. Grey shading indicates a disadvantage for the low educated; no shading indicates a disadvantage for the high educated. Difference in mortality decline could not be calculated for liver cirrhosis among women in Turin because of 0 deaths among high educated women in this population in the 2000s. Please note that the sum of the cause-specific changes does not always exactly equal the change for all-cause mortality, because of various adjustments and rounding.