Cardiovascular mortality and calcium and magnesium

[Guest guest]

Hi All, The below is PDF-available from: apater@... . ff's and fi's seem

to be ?'s. Too much magnesium in drinking water seems to be inferior to an

intermediate amount, so does that mean that too much magnesium supplement

will do the same thing?

Cheers, Al.

European Journal of Epidemiology 18 :305 –309,2003.

Cardiovascular mortality and calcium and magnesium in

drinking water:An ecological study in elderly people

Sebastien Marque,Hele`ne Jacqmin-Gadda, Jean-Francois Dartigues &

Commenges

Abstract.Background :Previous studies found rela-

tions between cardiovascular mortality and minerals

in drinking water,but the major works considered

water hardness or neglected the di ?erences between

adults and elderly.Drinking water is an important

source of calcium in the elderly particularly because

of increased needs and decreased consumption of

dairy products.Methods :We collected informations

about all deaths (14,311)occurring in 69 parishes of

the South –West of France during 7 years (1990 –

1996).We obtained the causes of deaths from a spe-

cial service of INSERM for each death,with age at

death and sex.The exposure value was supplied by

administrative source (DDASS)and by measurement

surveys.We use an extra-Poisson variation model to

Key words:Calcium,Cardiovascular mortality,Drinking

take into account the heterogeneity of the population

of these parishes.Results :A signi ?cant relationship

was observed between calcium and cardiovascular

mortality with a RR:0.90 for non-cerebrovascular

causes and RR:0.86 for cerebrovascular (when cal-

cium is higher than the second tercile:94 mg/l).We

found a protective e ?ect of magnesium concentra-

tions between 4 and 11 mg/l with a RR:0.92 for non-

cerebrovascular and RR:0.77 for cerebrovascular

mortality,as compared to concentrations lower than

4 mg/l.Conclusions :These ?ndings strongly suggest a

potential protective dose-e ?ect relation between cal-

cium in drinking water and cardiovascular causes.

For magnesium,a U-shape e ?ect is possible,espe-

cially for cerebrovascular mortality.

Introduction

Cardiovascular mortality represents the main cause

of mortality in people over 65 years of age,with an

incidence of 1.9 per 1000 person-years among men

and 1.3 per 1000 among women [1 ]in 1990 in France.

Previous epidemiologic studies have suggested an

inverse relationship between the number of deaths by

cardiovascular pathologies and water hardness [2 ],

mainly determined by concentrations of calcium and

magnesium.

Calcium and magnesium are antagonistic enzy-

matic activators.Calcium is essential to coagulation,

nerve impulse and muscular contraction,in particular

for the cardiac muscle.Magnesium is involved in the

transfer and the release of energy and takes part in

cardiac physiology.The results of epidemiologic

studies have suggested a potential hypotensive action

for calcium and an e ?ect against thrombosis for

magnesium [3 –5 ].

In developed countries,dairy products represent

the most signi ?cant source of calcium.However,it is

recognized that the intestinal absorption of calcium

decreases with aging [6 ],and particularly among

women [7 ].Moreover,elderly are sometimes con-

fronted with a problem of lactose intolerance;for

these subjects drinking water becomes the major

source of calcium [8,9 ].The daily intake of magne-

sium in the developed countries is insu ?cient com-

pared to recommended values [10 ].Moreover,since

the magnesium in drinking water is in ionized form,it

might be more bioavailable than that provided by

solid foods [11,12 ].This may also be the case for

calcium which seems to be at least as bioavailable in

water as in milk [13 ].

The various studies about the role of calcium and

magnesium have often shown a protective e ?ect of

these two elements on cardiovascular mortality.

However these works are essentially based on studies

of correlations between the concentrations of these

minerals in water and the death rates in the various

geographical areas [14,15 ].Moreover,some of these

studies have considered the hardness of water as a

global factor rather than the separate e ?ect of calci-

um and magnesium [16 ].

Recently,Rylander et al.[17 ]found that deaths by

ischaemic cardiopathies were inversely associated

with the concentrations of calcium and magnesium in

drinking water.According to the same authors,cal-

cium may have a protective e ?ect on cerebrovascular

mortality in men.However,this relation was not

found by Yang et al.[18,19 ]who suggested that

magnesium in drinking water may have a protective

e ?ect.This protective e ?ect of magnesium was also

found in other studies on cardiovascular mortality

[20 ].Nevertheless,some doubts persist regarding

these possible relationships because of contradictory

[21 ]or non-signi ?cant [22,23 ]results.Finally,the

relationship between cardiovascular mortality and

the mineral elements of drinking water has rarely

been studied in elderly [24 ].

The major aim of this study was to assess the re-

lationship between cardiovascular mortality in the

elderly and the concentration of calcium and mag-

nesium in drinking water.In e ?ect,the few number of

studies among elderly encourage us to investigate

more precisely this relationship [6,7 ].The study was

based on the population of the 75 parishes from

which the PAQUID sample was drawn.We have also

studied separately cerebrovascular and other cardio-

vascular mortality.The water distribution network

had already been studied for these parishes and mean

concentrations of calcium and magnesium were al-

ready available.Deaths certi ?cates were collected for

a 7-years period and the analysis beared on 14,331

deaths and 777,493 persons-years.............

Results

When we considered global cardiovascular mortality

(See Table 1),we observed a protective e ?ect of

higher calcium concentration (RR:0.90 –Con ?dence

Interval of 95%(CI95)[0.84;0.96 ]–for concentra-

tion >94 mg/l).We also found a protective e ?ect of

magnesium between 4 and 11 mg/l with estimated

relative risk between 0.88 and 0.92 (respectively

among women and men with ,CI95 [0.81;0.96 ]and,

CI95 [0.84;1.00 ]).

We then analysed separately non-cerebrovascular

(cardiovascular)mortality (10,492 deaths)and cere-

brovascular mortality (3819 deaths).Table 2 shows

the results of the extra-variation Poisson models for

cardiovascular mortality,excluding cerebrovascular

origin.These models suggested a protective e ?ect of a

high concentration of calcium on cardiovascular

mortality with a dose –response e ?ect.This e ?ect

seemed to be similar in both sexes,with a signi ?cant

relative risk in the parishes having a water with a

content higher than 94 mg/l (RR:0.90,CI95 [0.84;

0.97 ]).Concentration of magnesium between 4 and

11 mg/l seemed to be slightly protective on cardio-

vascular mortality (RR:0.92,CI95 [0.86;0.99 ]).

Table 3 shows the results of the extra-variation

Poisson models for cerebrovascular mortality.About

calcium,we found roughly similar results on cere-

brovascular mortality than on cardiovascular mor-

tality.We observed a protective relationship for

concentration >94 mg/l.This e ?ect seemed slightly

clearer among women (RR:0.84 vs.0.89,with CI95

respectively [0.74;0.97 ]and [0.74;1.07 ]).However,

we observed a more signi ?cant protective e ?ect for

magnesium for concentration between 4 and 11 mg/l.

This e ?ect seemed similar on both sexes and highly

signi ?cant (RR:0.77,p O :001).

The rural status of parishes did not seem to be

associated with cardiovascular mortality,and we

found this result in all models we performed.

Discussion

These ?ndings suggest a protective e ?ect of calcium

in drinking water with a dose-e ?ect relationship.This

e ?ect is similar in both sexes with a RR:0.90 in

cardiovascular mortality and RR:0.86 in cerebro-

vascular mortality.

Magnesium seems to be protective on cardiovas-

cular mortality for concentrations between 4 and

11 mg/l (RR:0.90).The e ?ect of magnesium seems to

be higher on cerebrovascular mortality (RR:0.77).

Considering two di ?erent causes in CV mortality

among the elderly was the major originality of this

analysis.Few analysis were performed among this

population and all of them considered only cardio-

vascular cause,without distinction between the sub-

causes.Furthermore,we could estimate independently