Environmental Geochemistry and Health 1993 15(4) page 201
Water hardness and cardiovascular mortality in Finland Risto Piispanen University of Oulu, Linnanmaa, SF-90570 Oulu, Finland
Abstract In Finland, where mortality rates of cardiovascular diseases are as a rule among the highest in the world, significant areal differences in these figures are encountered especially between the eastern and western parts of the country. To test the hypothesis that these differences,( the reason of which has long been a subject for a lively debate in Finland), may result from geochemical factors such as variations in the hardness or calcium or magnesium content of well water a correlation analysis was carried out in which the geochemical properties in each of the country's individual administrative districts, or communes (444 in all, excluding Aland) were compared with the percentage of heart diseases among the causes of death from diseases in 1991. The results show a poor or extremely low correlation nation-wide (r ranging from -0.039 to +0.045), indicating a virtual absence of any causal relationship. On the other hand, taken pairwise, the extreme western and eastern provinces, Vaasa and Northern Karelia, show marked differences both in their geochemical and mortality data, the percentage of deaths caused by cardiovascular diseases being markedly lower in the province of Vaasa, where the hardness and the magnesium content of well water (but not the calcium content) are significantly higher (p < 0.001 and 0.002 respectively). This inverse relationship is in line with the supposition that in addition to other factors, geochemical differences may have an effect on mortality.
Introduction
Extremely high age-adjusted rates of mortality from cardiovascular diseases have been typical of Finland, especially among the male population, for many decades. According to Karvonen et aI. (1991), Finland ranked third in the global mortality statistics for these diseases in the 1980s, after having had the questionable honour of ranking first during the 1970s. Recently, however, the situation seems to have started to improve, presumably because of the vigorous campaign against the most severe risk factors, i.e. cigarette smoking and the use of saturated animal fats in food, and in favour of healthier ways of life in general, propagated by the health authorities, the life assurance companies, the media, the schools and public opinion (Nikkari et al., 1983; Salo, 1987; Karvonen etal., 1991). Despite the gradually emerging positive results of the c a m p a i g n and the i m p r o v e d situation, age-adjusted mortality from cardiovascular diseases among the male population aged 35-64 years in Finland is still four times higher than in Portugal and France and ten times higher than in Japan (Karvonen et aI., 1991). Even t h o u g h the m o r t a l i t y rate for cardiovascular diseases in Finland is high in general terms, there is great regional variation, some areas deviating markedly from others (Karvonen, 1977). For instance, Vaasa, an administrative province in western Finland (Figure 1), characteristically has
significantly lower age-adjusted mortality rates for cardiovascular diseases (about 250/100 000; 1990) than the provinces in the east, an example of which is Northern Karelia (Figure 1), where the corresponding figure is around 400/100 000. Past figures in the same provinces in the 1970s were as high as around 400 and 700/100 000 respectively (Karvonen, 1977; Puska et al., 1977; Py6r~il~iet al., 1977; Punsar and Karvonen, 1979; Py6r~il~i and Valkonen, 1981; Karvonen et al., 1991). The high rates and the wide regional variation, and particularly the strong east-west contrast in Finland have been felt by many investigators to pose a tempting scientific challenge to find explanation and raise the interesting question concerning the reasons lying behind these relationships. Despite the long tradition and the i n t e n s i v e c l i n i c a l , epidemiological and geomedical research dealing with the problem, which has become known worldwide (Kobayashi, 1957; Morris et aI., 1961; Crawford etal., 1971; Anderson etal., 1975; Anderson, 1977; de Faire et al., 1977; Linderholm and Westlund, 1977; Neri and Johansen, 1978; National Academy of Sciences, 1979; Masironi, 1979; Comstock, 1980; National Public Health Institution, 1982; Bloom and Peric-Golia, 1983; Nikkari etal., 1983; Lacey and Shaper, 1984, Shaper 1984; Pocock et al., 1986; Masironi, 1987; Dzik, 1989), no consensus seems to have been reached on the issue and no single universally acceptable solution has been discovered. Instead,
202
Water hardness and cardiovascular mortality
several hypotheses differing widely in detail but following two main lines of thought have been offered as answers. The first group of theories emphasises the role of genetic factors and the other environmental ones. In spite of the relatively frequent familial aggregation, which could hint at a decisive role played by hereditary factors, cardiovascular diseases are commonly not regarded as clearly genetically controlled, since only a few of the k n o w n risk factors are transmitted as straightforward Mendelian traits (Nikkil~i and Rissanen, 1977), and in Finland in particular, the genetic differences and the differences in familial aggregation between east and west are relatively small (Nikkil/~ and Rissanen, 1977). Instead, differences in environmental conditions, such as socio-economic ones (Joensuu, 1989 in Finland; Rose and Marmot, 1981 in Great Britain) or regional differences in habits concerning animal and other fats in food (Puska et al., 1977; Nikkari et al., 1983;
Salo, 1987), differences in blood pressure (hypertension; Aromaa et al., 1977; Witteman and Grobbee, 1991), differences in serum cholesterol content (National Public Health Institution, 1982), smoking habits (Puska et aL, 1977; Reunanen et al., 1983), amount of bodily exercise (Gyntelberg, 1977; Py6r~il/i etal., 1977) or the level of triglycerides or post-load plasma glucose in the blood (Reunanen et al., 1983) have been favoured recently as the basis for explanations. In addition to the factors listed above, obesity (Reunanen et al., 1983), cold climate (Valkonen and Notkola, 1977; Gyllerup et al., 1991) and unemployment (Joensuu, 1989) have also been suggested as important
g
0 e
I 28OE
Q
I
Ca > 45mg/I I-r-I-J-n 3 0 - 4 5
II I I II
I I
20-30 I
18-20
9-18
I
I,:9
Q
°
°
I
I
O g
".[5 ~ "
I
Q
--64ON
Vaasa Karelia
24°E
I
Figure 1 Index map of Finland showing the location of the administrative provinces of Vaasa and Northern Karelia.
Figure 2 Incidence of calcium (mg L -1) in Finnish well water. The map is a simplified black and white reproduction of the original map presented in colour in the Geochemical Atlas of Finland, Part 1. (Lahermo et al., 1990).
R. Piispanen
environmental conditions leading to enhanced incidences of cardiovascular diseases. The claim that environmental conditions affect cardiovascular health also finds support in the observation of de Faire et aL (1977) that the incidence of ischaemic heart diseases was lower among Finnish immigrants in Sweden than among the population remaining in Finland. Similarly, the incidence of coronary heart disease among policemen moving from Northern Karelia to Helsinki was found to be lower than among the male population remaining in their home province (Py6r~il/i et aL, 1977). On the other hand, the role of genetic and ethnic factors is reflected in the fact that the i n c i d e n c e among Finnish immigrants in Sweden is nevertheless higher than among the Swedish host population (Py6r~il~iet al., 1977).
Mfl
Figure 3 Magnesium in Finnish well water.
203
Among the explanations that emphasise the role of environmental conditions, proposals based on geochemical differences between the areas under comparison as solutions to the problem have also been put forward and discussed (Morris et aL, 1961; Crawford etal., 1971; Anderson etal., 1975; Anderson, 1977; Neri and Johansen, 1978; Comstock, 1980; Shaper, 1984; Pocock et al., 1986; Masironi, 1987; Dzik 1989). Magnesium, the chief factor responsible for water hardness alongside with calcium, has been shown clinically to be beneficial for the prevention and cure of cardiovascular problems (Turlapaty et al., 1980; Lowik etal., 1982; Leary et al., 1983; Luoma et al., 1983; Altura et al., 1984; Altura and Altura, 1986; Eisenberg, 1986; Gunther 1986; Hopps and Feder, 1986; Karppanen, 1990; Schechter, 1990; Singh, 1990; Whang and Ryder, 1990; Rylander etal., 1991; Schechter and Hod, 1991; Woods et al., 1992), and Altura and Altura (1986), Karppanen (1990) and Whang and Ryder (1990) have proved in clinical tests that a magnesium deficit in the diet can lead to cardiac arrythmias. The role of calcium, however, seems to be much more controversial (for a discussion, see Masironi, 1987). Clinical experience referred to above seems to suggest that magnesium may have a beneficial effect on cardiac health while present in the drinking water, and in fact a number of epidemiological investigations have demonstrated that this is true at least in some areas, and that an inverse relationship prevails between water hardness and the incidence of heart diseases in many places (Kobayashi, 1957; Crawford etal., 1971; Anderson, 1977; Valkonen and Notkola, 1977; Neri and Johansen, 1978; Punsar and Karvonen, 1979; Comstock, 1980; Leary et al., 1983; Luoma et aL, 1983; Hopps and Feder, 1986; Leary, 1986). Similarly, a statistically significant inverse relationship between water hardness and mortality from cardiovascular diseases for both sexes was reported recently by Rylander et al. (1991) in Sweden, while Bloom and Peric-Golia (1989) c o n c l u d e d that the higher level of magnesium in drinking water may be the reason why death from myocardial infarction is rarer in Salt Lake City than in Washington. Even if the overall global picture (summarised above) favours a positive effect of water hardness on cardiac health, the etiology of cardiovascular diseases is, much more complicated, the diseases being influenced by many factors. There are significant exceptions from the overall global pattern outlined above. For example, the correlation between mortality from cerebrovascular diseases and water hardness in North Dakota (Dzik, 1989) was found to be only very weak at its best and, as an extreme, the drinking water in Greenland, known for its low incidence of ischaemic heart deaths, is extremely soft (Jeppesen, 1987). As far as Finland is concerned, a study of the possible correlation between the hardness of water and the incidence of cardiovascular deaths is specifically warranted in view of the high mortality
204
Water hardness and cardiovascular mortality
figures and the marked regional variation mentioned above. Also, the general conformity of water hardness (Figure 2) and magnesium and calcium content maps (Figure 3) with the general picture of east-west differences (Vaasa versus Northern Karelia) strongly suggests that at least one of the causative factors may be geochemical. It is because of the impulse obtained from the maps that the present study was launched. Material and Methods
The main purpose was to attempt to establish a correlation between regional hydrogeochemical variables (hardness of well water and their calcium and magnesium content) and mortality from cardiovascular diseases in Jthe administrative districts (communes) of Finland. Based on the assumption that hydrogeochemical maps give
sufficiently reliable information as to the quality of the water used in an area, the geochemical data necessary were read off from the geochemical maps published recently by the National Board of Survey and the Geological Survey of Finland. The former is the publisher of the Atlas of Finland, which includes a map presenting well-water hardness in Finland (Alalammi, 1992), while the latter has now published two volumes of The GeochemicalAtlas of Finland, the first of which contains multicolour maps which include such features as the calcium and magnesium content of well water (Lahermo et al., 1990). For methods of sampling and analysis the reader is referred to the original publication of
• .°." • °...
.-. ,
.
. Water
hardness
~
-50
~
-60
]
-70
]
-80
~
>80
E E E
Figure 5 Deaths caused by cardiovascular diseases in Figure 4 Hardness of Finnish well water as expressed in German degrees. (Alalammi, 1992).
1991 as a percentage of all deaths caused by diseases. in the administrative districts (communes) of Finland; male population.
R. Piispanen
Table 1 Arithmetic means (x) and lowest (min) and highest (max) values and standard deviations (s) for water hardness (H; °dH), and calcium (Ca; mg L -1) and magnesium (Mg; mg L -1) content in well water and deaths (T-F)from cardiovascular diseases as a percentage of all deaths caused by diseases in the 444 administrative districts population (communes) of Finland (excluding ,~land) in 1991.
H
Ca' Mg T M F
x
min.
3.2
1
19.3 4.4 56.0 55.2 56.9
10 1 22.2 0.1 0.1
max. 9
40 9 84.6 99.9 99.9
s 1.4
5.4 2.2 8.7 11.9 12.7
T = total, M = male and F = female population.
Lahermo et al. (1990). The maps are redrawn in simplified form in Figures 2-4. The necessary mortality data for the study was obtained from the archives of Statistics Finland (formerly the Central Statistical Office of Finland). The original data were in the form of numbers of deaths caused by cardiovascular diseases in each of the administrative districts (communes) of mainland Finland (444 in all, with Aland excluded). The most recent year for which data was available, (1991), was chosen as the year to be studied, and the data were converted to deaths from cardiovascular diseases as percentages of all deaths caused by d i s e a s e s . The c a l c u l a t i o n s w e r e p e r f o r m e d separately for males and females and for the whole p o p u l a t i o n c o m b i n e d . T o g e t h e r w i t h the geochemical variables, the data were subjected to c o m p u t e r i s e d statistical p r o c e s s i n g in which S p e a r m a n rank c o r r e l a t i o n c o e f f i c i e n t s and arithmetical means and standard deviations for the variables were calculated, followed by a t test based on Student's t statistic in order to evaluate the statistical significance of the differences between the arithmetical means for the most lively debated provinces (Vaasa and Northern Karelia) which show contrasting geochemical and mortality values. The statistical calculations were performed using the BMDP computer program package (Dixon, 1981) and an IBM 7171 computer. Results
The results are given in Tables 1-3 and Figures 5 and 6. Table 1 gives the arithmetical means, standard deviations and lowest and highest values for the geochemical variables and the mortality percentages. The overall means for magnesium (4.4 mg L-l; Table 1) and calcium (19.3 mg L- l) can be seen to be significantly higher than the average values recorded by Matthess (1982) for groundwater
205
Table 2 Matrix of Spearman's rank correlation coefficients between well water hardness (H), calcium (Ca) and magnesium (Mg) and the percentage of heart diseases as the cause of death in female (F), male (M) and total (T) population in 444 communes of Finland in 1991.
H Ca Mg
F
M
T
0.002 0.013 -0.006
0.008 0.020 -0.008
-0.005 0.045 -0.039
Table 3 Arithmetical means (x), standard deviations (s), t values (t) and the associated risk levels (p) for the differences between the geochemical variables of water hardness (H; °dH), and calcium (Ca," mg L -1) • and magnestum (Mg; mg L - - 1 ) content, and deaths caused by cardiovascular diseases as a percentage of all deaths caused by diseases in the administrative districts (communes) (T-F) between the province of Vaasa (n = 57) and Northern Karelia in 1991.
x H
Ca Mg T M F
Vaasa s
3.4
1.3
19.0 4.6 54.7 54.5 55.1
3.5 2.3 8.8 12.6 12.9
Northern Karelia x s t
p
2.0 17.5 3.4 60.1 60.4 59.6
Water hardness and cardiovascular mortality in Finland Risto Piispanen University of Oulu, Linnanmaa, SF-90570 Oulu, Finland
Abstract In Finland, where mortality rates of cardiovascular diseases are as a rule among the highest in the world, significant areal differences in these figures are encountered especially between the eastern and western parts of the country. To test the hypothesis that these differences,( the reason of which has long been a subject for a lively debate in Finland), may result from geochemical factors such as variations in the hardness or calcium or magnesium content of well water a correlation analysis was carried out in which the geochemical properties in each of the country's individual administrative districts, or communes (444 in all, excluding Aland) were compared with the percentage of heart diseases among the causes of death from diseases in 1991. The results show a poor or extremely low correlation nation-wide (r ranging from -0.039 to +0.045), indicating a virtual absence of any causal relationship. On the other hand, taken pairwise, the extreme western and eastern provinces, Vaasa and Northern Karelia, show marked differences both in their geochemical and mortality data, the percentage of deaths caused by cardiovascular diseases being markedly lower in the province of Vaasa, where the hardness and the magnesium content of well water (but not the calcium content) are significantly higher (p < 0.001 and 0.002 respectively). This inverse relationship is in line with the supposition that in addition to other factors, geochemical differences may have an effect on mortality.
Introduction
Extremely high age-adjusted rates of mortality from cardiovascular diseases have been typical of Finland, especially among the male population, for many decades. According to Karvonen et aI. (1991), Finland ranked third in the global mortality statistics for these diseases in the 1980s, after having had the questionable honour of ranking first during the 1970s. Recently, however, the situation seems to have started to improve, presumably because of the vigorous campaign against the most severe risk factors, i.e. cigarette smoking and the use of saturated animal fats in food, and in favour of healthier ways of life in general, propagated by the health authorities, the life assurance companies, the media, the schools and public opinion (Nikkari et al., 1983; Salo, 1987; Karvonen etal., 1991). Despite the gradually emerging positive results of the c a m p a i g n and the i m p r o v e d situation, age-adjusted mortality from cardiovascular diseases among the male population aged 35-64 years in Finland is still four times higher than in Portugal and France and ten times higher than in Japan (Karvonen et aI., 1991). Even t h o u g h the m o r t a l i t y rate for cardiovascular diseases in Finland is high in general terms, there is great regional variation, some areas deviating markedly from others (Karvonen, 1977). For instance, Vaasa, an administrative province in western Finland (Figure 1), characteristically has
significantly lower age-adjusted mortality rates for cardiovascular diseases (about 250/100 000; 1990) than the provinces in the east, an example of which is Northern Karelia (Figure 1), where the corresponding figure is around 400/100 000. Past figures in the same provinces in the 1970s were as high as around 400 and 700/100 000 respectively (Karvonen, 1977; Puska et al., 1977; Py6r~il~iet al., 1977; Punsar and Karvonen, 1979; Py6r~il~i and Valkonen, 1981; Karvonen et al., 1991). The high rates and the wide regional variation, and particularly the strong east-west contrast in Finland have been felt by many investigators to pose a tempting scientific challenge to find explanation and raise the interesting question concerning the reasons lying behind these relationships. Despite the long tradition and the i n t e n s i v e c l i n i c a l , epidemiological and geomedical research dealing with the problem, which has become known worldwide (Kobayashi, 1957; Morris et aI., 1961; Crawford etal., 1971; Anderson etal., 1975; Anderson, 1977; de Faire et al., 1977; Linderholm and Westlund, 1977; Neri and Johansen, 1978; National Academy of Sciences, 1979; Masironi, 1979; Comstock, 1980; National Public Health Institution, 1982; Bloom and Peric-Golia, 1983; Nikkari etal., 1983; Lacey and Shaper, 1984, Shaper 1984; Pocock et al., 1986; Masironi, 1987; Dzik, 1989), no consensus seems to have been reached on the issue and no single universally acceptable solution has been discovered. Instead,
202
Water hardness and cardiovascular mortality
several hypotheses differing widely in detail but following two main lines of thought have been offered as answers. The first group of theories emphasises the role of genetic factors and the other environmental ones. In spite of the relatively frequent familial aggregation, which could hint at a decisive role played by hereditary factors, cardiovascular diseases are commonly not regarded as clearly genetically controlled, since only a few of the k n o w n risk factors are transmitted as straightforward Mendelian traits (Nikkil~i and Rissanen, 1977), and in Finland in particular, the genetic differences and the differences in familial aggregation between east and west are relatively small (Nikkil/~ and Rissanen, 1977). Instead, differences in environmental conditions, such as socio-economic ones (Joensuu, 1989 in Finland; Rose and Marmot, 1981 in Great Britain) or regional differences in habits concerning animal and other fats in food (Puska et al., 1977; Nikkari et al., 1983;
Salo, 1987), differences in blood pressure (hypertension; Aromaa et al., 1977; Witteman and Grobbee, 1991), differences in serum cholesterol content (National Public Health Institution, 1982), smoking habits (Puska et aL, 1977; Reunanen et al., 1983), amount of bodily exercise (Gyntelberg, 1977; Py6r~il/i etal., 1977) or the level of triglycerides or post-load plasma glucose in the blood (Reunanen et al., 1983) have been favoured recently as the basis for explanations. In addition to the factors listed above, obesity (Reunanen et al., 1983), cold climate (Valkonen and Notkola, 1977; Gyllerup et al., 1991) and unemployment (Joensuu, 1989) have also been suggested as important
g
0 e
I 28OE
Q
I
Ca > 45mg/I I-r-I-J-n 3 0 - 4 5
II I I II
I I
20-30 I
18-20
9-18
I
I,:9
Q
°
°
I
I
O g
".[5 ~ "
I
Q
--64ON
Vaasa Karelia
24°E
I
Figure 1 Index map of Finland showing the location of the administrative provinces of Vaasa and Northern Karelia.
Figure 2 Incidence of calcium (mg L -1) in Finnish well water. The map is a simplified black and white reproduction of the original map presented in colour in the Geochemical Atlas of Finland, Part 1. (Lahermo et al., 1990).
R. Piispanen
environmental conditions leading to enhanced incidences of cardiovascular diseases. The claim that environmental conditions affect cardiovascular health also finds support in the observation of de Faire et aL (1977) that the incidence of ischaemic heart diseases was lower among Finnish immigrants in Sweden than among the population remaining in Finland. Similarly, the incidence of coronary heart disease among policemen moving from Northern Karelia to Helsinki was found to be lower than among the male population remaining in their home province (Py6r~il/i et aL, 1977). On the other hand, the role of genetic and ethnic factors is reflected in the fact that the i n c i d e n c e among Finnish immigrants in Sweden is nevertheless higher than among the Swedish host population (Py6r~il~iet al., 1977).
Mfl
Figure 3 Magnesium in Finnish well water.
203
Among the explanations that emphasise the role of environmental conditions, proposals based on geochemical differences between the areas under comparison as solutions to the problem have also been put forward and discussed (Morris et aL, 1961; Crawford etal., 1971; Anderson etal., 1975; Anderson, 1977; Neri and Johansen, 1978; Comstock, 1980; Shaper, 1984; Pocock et al., 1986; Masironi, 1987; Dzik 1989). Magnesium, the chief factor responsible for water hardness alongside with calcium, has been shown clinically to be beneficial for the prevention and cure of cardiovascular problems (Turlapaty et al., 1980; Lowik etal., 1982; Leary et al., 1983; Luoma et al., 1983; Altura et al., 1984; Altura and Altura, 1986; Eisenberg, 1986; Gunther 1986; Hopps and Feder, 1986; Karppanen, 1990; Schechter, 1990; Singh, 1990; Whang and Ryder, 1990; Rylander etal., 1991; Schechter and Hod, 1991; Woods et al., 1992), and Altura and Altura (1986), Karppanen (1990) and Whang and Ryder (1990) have proved in clinical tests that a magnesium deficit in the diet can lead to cardiac arrythmias. The role of calcium, however, seems to be much more controversial (for a discussion, see Masironi, 1987). Clinical experience referred to above seems to suggest that magnesium may have a beneficial effect on cardiac health while present in the drinking water, and in fact a number of epidemiological investigations have demonstrated that this is true at least in some areas, and that an inverse relationship prevails between water hardness and the incidence of heart diseases in many places (Kobayashi, 1957; Crawford etal., 1971; Anderson, 1977; Valkonen and Notkola, 1977; Neri and Johansen, 1978; Punsar and Karvonen, 1979; Comstock, 1980; Leary et al., 1983; Luoma et aL, 1983; Hopps and Feder, 1986; Leary, 1986). Similarly, a statistically significant inverse relationship between water hardness and mortality from cardiovascular diseases for both sexes was reported recently by Rylander et al. (1991) in Sweden, while Bloom and Peric-Golia (1989) c o n c l u d e d that the higher level of magnesium in drinking water may be the reason why death from myocardial infarction is rarer in Salt Lake City than in Washington. Even if the overall global picture (summarised above) favours a positive effect of water hardness on cardiac health, the etiology of cardiovascular diseases is, much more complicated, the diseases being influenced by many factors. There are significant exceptions from the overall global pattern outlined above. For example, the correlation between mortality from cerebrovascular diseases and water hardness in North Dakota (Dzik, 1989) was found to be only very weak at its best and, as an extreme, the drinking water in Greenland, known for its low incidence of ischaemic heart deaths, is extremely soft (Jeppesen, 1987). As far as Finland is concerned, a study of the possible correlation between the hardness of water and the incidence of cardiovascular deaths is specifically warranted in view of the high mortality
204
Water hardness and cardiovascular mortality
figures and the marked regional variation mentioned above. Also, the general conformity of water hardness (Figure 2) and magnesium and calcium content maps (Figure 3) with the general picture of east-west differences (Vaasa versus Northern Karelia) strongly suggests that at least one of the causative factors may be geochemical. It is because of the impulse obtained from the maps that the present study was launched. Material and Methods
The main purpose was to attempt to establish a correlation between regional hydrogeochemical variables (hardness of well water and their calcium and magnesium content) and mortality from cardiovascular diseases in Jthe administrative districts (communes) of Finland. Based on the assumption that hydrogeochemical maps give
sufficiently reliable information as to the quality of the water used in an area, the geochemical data necessary were read off from the geochemical maps published recently by the National Board of Survey and the Geological Survey of Finland. The former is the publisher of the Atlas of Finland, which includes a map presenting well-water hardness in Finland (Alalammi, 1992), while the latter has now published two volumes of The GeochemicalAtlas of Finland, the first of which contains multicolour maps which include such features as the calcium and magnesium content of well water (Lahermo et al., 1990). For methods of sampling and analysis the reader is referred to the original publication of
• .°." • °...
.-. ,
.
. Water
hardness
~
-50
~
-60
]
-70
]
-80
~
>80
E E E
Figure 5 Deaths caused by cardiovascular diseases in Figure 4 Hardness of Finnish well water as expressed in German degrees. (Alalammi, 1992).
1991 as a percentage of all deaths caused by diseases. in the administrative districts (communes) of Finland; male population.
R. Piispanen
Table 1 Arithmetic means (x) and lowest (min) and highest (max) values and standard deviations (s) for water hardness (H; °dH), and calcium (Ca; mg L -1) and magnesium (Mg; mg L -1) content in well water and deaths (T-F)from cardiovascular diseases as a percentage of all deaths caused by diseases in the 444 administrative districts population (communes) of Finland (excluding ,~land) in 1991.
H
Ca' Mg T M F
x
min.
3.2
1
19.3 4.4 56.0 55.2 56.9
10 1 22.2 0.1 0.1
max. 9
40 9 84.6 99.9 99.9
s 1.4
5.4 2.2 8.7 11.9 12.7
T = total, M = male and F = female population.
Lahermo et al. (1990). The maps are redrawn in simplified form in Figures 2-4. The necessary mortality data for the study was obtained from the archives of Statistics Finland (formerly the Central Statistical Office of Finland). The original data were in the form of numbers of deaths caused by cardiovascular diseases in each of the administrative districts (communes) of mainland Finland (444 in all, with Aland excluded). The most recent year for which data was available, (1991), was chosen as the year to be studied, and the data were converted to deaths from cardiovascular diseases as percentages of all deaths caused by d i s e a s e s . The c a l c u l a t i o n s w e r e p e r f o r m e d separately for males and females and for the whole p o p u l a t i o n c o m b i n e d . T o g e t h e r w i t h the geochemical variables, the data were subjected to c o m p u t e r i s e d statistical p r o c e s s i n g in which S p e a r m a n rank c o r r e l a t i o n c o e f f i c i e n t s and arithmetical means and standard deviations for the variables were calculated, followed by a t test based on Student's t statistic in order to evaluate the statistical significance of the differences between the arithmetical means for the most lively debated provinces (Vaasa and Northern Karelia) which show contrasting geochemical and mortality values. The statistical calculations were performed using the BMDP computer program package (Dixon, 1981) and an IBM 7171 computer. Results
The results are given in Tables 1-3 and Figures 5 and 6. Table 1 gives the arithmetical means, standard deviations and lowest and highest values for the geochemical variables and the mortality percentages. The overall means for magnesium (4.4 mg L-l; Table 1) and calcium (19.3 mg L- l) can be seen to be significantly higher than the average values recorded by Matthess (1982) for groundwater
205
Table 2 Matrix of Spearman's rank correlation coefficients between well water hardness (H), calcium (Ca) and magnesium (Mg) and the percentage of heart diseases as the cause of death in female (F), male (M) and total (T) population in 444 communes of Finland in 1991.
H Ca Mg
F
M
T
0.002 0.013 -0.006
0.008 0.020 -0.008
-0.005 0.045 -0.039
Table 3 Arithmetical means (x), standard deviations (s), t values (t) and the associated risk levels (p) for the differences between the geochemical variables of water hardness (H; °dH), and calcium (Ca," mg L -1) • and magnestum (Mg; mg L - - 1 ) content, and deaths caused by cardiovascular diseases as a percentage of all deaths caused by diseases in the administrative districts (communes) (T-F) between the province of Vaasa (n = 57) and Northern Karelia in 1991.
x H
Ca Mg T M F
Vaasa s
3.4
1.3
19.0 4.6 54.7 54.5 55.1
3.5 2.3 8.8 12.6 12.9
Northern Karelia x s t
p
2.0 17.5 3.4 60.1 60.4 59.6
