Vol. 21, No. 1 Printed in Great Britain
International Journal of Epidemiology © International Epidemiological Association 1992
Calcium Intake and 28-Year Cardiovascular and Coronary Heart Disease Mortality in Dutch Civil Servants Van der Vijver L P L (Department of Epidemiology and Public Health, Agricultural University Wageningen, PO Box 238, 6700 AE Wageningen, The Netherlands), van der Waal M A E , Weterings K G C, Dekker J M, Schouten E G and Kok F J. Calcium intake and 28-year cardiovascular and coronary heart disease mortality in Dutch civil servants. International Journal of Epidemiology 1992; 21: 36-39. Data obtained from a general health examination in 1953-1954 of 2605 middle-aged Dutch civil servants were analysed to investigate the relation between dietary calcium and cardiovascular (CVD) and coronary heart disease (CHD) mortality. Calcium intake was assessed at baseline by a 1-week food frequency recall. Multivariate adjusted odds ratios (OR) were calculated using the highest quintile of calcium intake as the reference. No statistically significant associations were observed for low calcium intake in 15 and 28 years of follow-up in both men and women. For men, multivariate adjusted OR for the lowest quintile of calcium intake were 1.3 (95% confidence interval (CD: 0.8-1.9) and 0.9 (95% Cl: 0.6-1.6) for 28-year CVD and CHD mortality, respectively. For women, corresponding OR were 1.1 (95% Cl: 0.6-2.0) and 1.1 (95% Cl: 0.5-2.5). Although an inverse association between calcium intake and CVD and CHD mortality, possibly mediated by blood pressure, might be hypothesized, no clear association was observed. Because dietary patterns in the 1950s were quite stable, and major calcium sources were addressed, misclassification of calcium intake may not be fully responsible for this finding.
POPULATION AND METHODS Population In 1953-1954 a general health survey was held in Amsterdam. Its purpose was to study the feasibility of early detection of chronic diseases. About 54% of 11 700 eligible civil servants volunteered to participate in the study, with or without their spouses. From these volunteers an age- and sex-stratified sample was chosen, with the purpose of obtaining about equal numbers in the 5-year age and sex categories, resulting in participation of 1583 men and 1508 women, aged between 40 and 65 years.8
Several ecological studies have shown that populations living in hard water areas (high calcium content) have lower age-adjusted cardiovascular disease (CVD) mortality than people living in soft water areas.1 Although a low calcium intake in soft water areas might be responsible for these observations, it may be argued that this effect is mediated by other risk factors, including blood pressure.2 This is plausible since calcium intake and blood pressure are inversely related.3"7 From these findings it may be hypothesized that a low calcium intake increases risk of CVD. Such an effect may be entirely or partly attributable to blood pressure. Therefore, we examined the association of calcium intake with CVD and coronary heart disease (CHD) mortality in a prospective cohort study of 3091 Dutch civil servants and their spouses, taking into account a potential intermediate role of blood pressure.
Data Collection The survival of 99.2% of the male and 99.7% of the female population was determined up to 1981. Further details of this survival assessment have been published previously.9 In 1989, the causes of death until 1981 were ascertained. For each deceased the residence at death and the death certificate number were assessed through the Dutch Central Bureau of Genealogy. The combination of date of death, residence at death and death
• Department of Epidemiology and Public Health, Agricultural University Wageningen, PO Box 238, 6700 AE Wageningen, The Netherlands. t Epidemiology section, TNO Toxicology and Nutrition Institute, Zeist, The Netherlands.
36
Downloaded from http://ije.oxfordjournals.org/ at Thomas Jefferson University, Scott Library on March 10, 2015
LUCY P L VAN DER VIJVER.* MARIEKE A E VAN DER WAAL,* KARIN G C WETERINGS,* JACQUELINE M DEKKER,* EVERT G SCHOUTEN* AND FRANS J KOK*t
37
CALCIUM INTAKE AND MORTALITY IN THE NETHERLANDS
RESULTS Table 1 shows the baseline characteristics of 2605 Dutch civil servants and spouses, by quintiles of calcium intake for men and women separately. The overall mean calcium intake for females (660±283 mg) is lower than for males (928 ±413 mg) and is due to lower energy intakes of females compared to males. In Table 2, the age-adjusted cumulative incidences of CVD and CHD mortality, according to quintiles of calcium intake, are presented. The second, third and fourth quintile are combined into one category. In 28 years of follow-up, dietary calcium showed no
Data Analysis Individuals with missing information on survival or cause of death (n = 31) were excluded. Also excluded were people who did not meet the age criteria (n = 36), who had no data on calcium intake (n = 289), and those with extremely low energy intake (n = 130) (for men energy intake < 1500 kcal, women < 1250 kcal). This way the analyses comprised 1340 men and 1265 women. Participants were classified in quintiles of calcium intake for men and women separately. Directly age-standardized cumulative incidences were calcu-
TABLE 1 Baseline characteristics of the Dutch Civil Servants Study in 1953-1954 Calcium (Ca) quintiles in mg/day
No.
Males O85 585
International Journal of Epidemiology © International Epidemiological Association 1992
Calcium Intake and 28-Year Cardiovascular and Coronary Heart Disease Mortality in Dutch Civil Servants Van der Vijver L P L (Department of Epidemiology and Public Health, Agricultural University Wageningen, PO Box 238, 6700 AE Wageningen, The Netherlands), van der Waal M A E , Weterings K G C, Dekker J M, Schouten E G and Kok F J. Calcium intake and 28-year cardiovascular and coronary heart disease mortality in Dutch civil servants. International Journal of Epidemiology 1992; 21: 36-39. Data obtained from a general health examination in 1953-1954 of 2605 middle-aged Dutch civil servants were analysed to investigate the relation between dietary calcium and cardiovascular (CVD) and coronary heart disease (CHD) mortality. Calcium intake was assessed at baseline by a 1-week food frequency recall. Multivariate adjusted odds ratios (OR) were calculated using the highest quintile of calcium intake as the reference. No statistically significant associations were observed for low calcium intake in 15 and 28 years of follow-up in both men and women. For men, multivariate adjusted OR for the lowest quintile of calcium intake were 1.3 (95% confidence interval (CD: 0.8-1.9) and 0.9 (95% Cl: 0.6-1.6) for 28-year CVD and CHD mortality, respectively. For women, corresponding OR were 1.1 (95% Cl: 0.6-2.0) and 1.1 (95% Cl: 0.5-2.5). Although an inverse association between calcium intake and CVD and CHD mortality, possibly mediated by blood pressure, might be hypothesized, no clear association was observed. Because dietary patterns in the 1950s were quite stable, and major calcium sources were addressed, misclassification of calcium intake may not be fully responsible for this finding.
POPULATION AND METHODS Population In 1953-1954 a general health survey was held in Amsterdam. Its purpose was to study the feasibility of early detection of chronic diseases. About 54% of 11 700 eligible civil servants volunteered to participate in the study, with or without their spouses. From these volunteers an age- and sex-stratified sample was chosen, with the purpose of obtaining about equal numbers in the 5-year age and sex categories, resulting in participation of 1583 men and 1508 women, aged between 40 and 65 years.8
Several ecological studies have shown that populations living in hard water areas (high calcium content) have lower age-adjusted cardiovascular disease (CVD) mortality than people living in soft water areas.1 Although a low calcium intake in soft water areas might be responsible for these observations, it may be argued that this effect is mediated by other risk factors, including blood pressure.2 This is plausible since calcium intake and blood pressure are inversely related.3"7 From these findings it may be hypothesized that a low calcium intake increases risk of CVD. Such an effect may be entirely or partly attributable to blood pressure. Therefore, we examined the association of calcium intake with CVD and coronary heart disease (CHD) mortality in a prospective cohort study of 3091 Dutch civil servants and their spouses, taking into account a potential intermediate role of blood pressure.
Data Collection The survival of 99.2% of the male and 99.7% of the female population was determined up to 1981. Further details of this survival assessment have been published previously.9 In 1989, the causes of death until 1981 were ascertained. For each deceased the residence at death and the death certificate number were assessed through the Dutch Central Bureau of Genealogy. The combination of date of death, residence at death and death
• Department of Epidemiology and Public Health, Agricultural University Wageningen, PO Box 238, 6700 AE Wageningen, The Netherlands. t Epidemiology section, TNO Toxicology and Nutrition Institute, Zeist, The Netherlands.
36
Downloaded from http://ije.oxfordjournals.org/ at Thomas Jefferson University, Scott Library on March 10, 2015
LUCY P L VAN DER VIJVER.* MARIEKE A E VAN DER WAAL,* KARIN G C WETERINGS,* JACQUELINE M DEKKER,* EVERT G SCHOUTEN* AND FRANS J KOK*t
37
CALCIUM INTAKE AND MORTALITY IN THE NETHERLANDS
RESULTS Table 1 shows the baseline characteristics of 2605 Dutch civil servants and spouses, by quintiles of calcium intake for men and women separately. The overall mean calcium intake for females (660±283 mg) is lower than for males (928 ±413 mg) and is due to lower energy intakes of females compared to males. In Table 2, the age-adjusted cumulative incidences of CVD and CHD mortality, according to quintiles of calcium intake, are presented. The second, third and fourth quintile are combined into one category. In 28 years of follow-up, dietary calcium showed no
Data Analysis Individuals with missing information on survival or cause of death (n = 31) were excluded. Also excluded were people who did not meet the age criteria (n = 36), who had no data on calcium intake (n = 289), and those with extremely low energy intake (n = 130) (for men energy intake < 1500 kcal, women < 1250 kcal). This way the analyses comprised 1340 men and 1265 women. Participants were classified in quintiles of calcium intake for men and women separately. Directly age-standardized cumulative incidences were calcu-
TABLE 1 Baseline characteristics of the Dutch Civil Servants Study in 1953-1954 Calcium (Ca) quintiles in mg/day
No.
Males O85 585
