Search PubMed⌕ Search

Biomedical subjects

J Pekkanen

Publications and source records attributed to J Pekkanen.

At least 109 records · Page 6Linked to original sources
PubMed1995

Dietary factors determining diabetes and impaired glucose tolerance. A 20-year follow-up of the Finnish and Dutch cohorts of the Seven Countries Study.

OBJECTIVE: To investigate the role of diet as a predictor of glucose intolerance and non-insulin-dependent diabetes mellitus (NIDDM). RESEARCH DESIGN AND METHODS: At the 30-year follow-up survey of the Dutch and Finnish cohorts of the Seven Countries Study, in 1989/1990, men were examined according to a standardized protocol including a 2-h oral glucose tolerance test. Information on habitual food consumption was obtained using the cross-check dietary history method. Those 338 men in whom information on habitual diet was also available 20 years earlier were included in this study. Subjects known as having diabetes in 1989/1990 were excluded from the analyses. RESULTS: Adjusting for age and cohort, the intake of total, saturated, and monounsaturated fatty acids and dietary cholesterol 20 years before diagnosis was higher in men with newly diagnosed diabetes in the survey than in men with normal or impaired glucose tolerance. After adjustment for cohort, age, past body mass index, and past energy intake, the past intake of total fat was positively associated with 2-h postload glucose level (P < 0.05). An independent inverse association with the past intake of vitamin C was observed (P < 0.05). These associations were independent of changes in the intake of fat and vitamin C during the 20-year follow-up. An increase in the consumption of vegetables and legumes, potatoes, and fish during the 20-year follow-up was inversely related with 2-h glucose level (P < 0.05). CONCLUSIONS: Although the regression coefficients were in general not very large, these results indicate that a high intake of fat, especially that of saturated fatty acids, contributes to the risk of glucose intolerance and NIDDM. Foods such as fish, potatoes, vegetables, and legumes may have a protective effect. In addition, the observed inverse association between vitamin C and glucose intolerance suggests that antioxidants may also play a role in the development of derangements in glucose metabolism.

Cohort Studies↗
PubMed1994

Serum cholesterol concentration and mortality from accidents, suicide, and other violent causes.

OBJECTIVE: To study the association of mortality from accidents, suicides, and other violent deaths with serum cholesterol concentration. DESIGN: Baseline measurements in two randomly chosen independent cohorts were carried out in 1972 and 1977. Mortality was monitored over 10-15 years through the national death registry. SETTING: Eastern Finland. SUBJECTS: The two cohorts comprised men (n = 10,898) and women (n = 11,534) born between 1913 and 1947. There were 193 deaths due to accidents, suicides, and violence among men and 43 among women. MAIN OUTCOME MEASURE: Mortality from accidents, suicides, and other violent deaths was used as the end point. Deaths from these causes were pooled together in the analyses. RESULTS: Serum cholesterol concentration was not associated with mortality from accidents, suicides, and other violent deaths in the univariate analyses or in the proportional hazards regression analyses including smoking, systolic blood pressure, alcohol drinking, and education. In both genders smoking was more prevalent among those who died from accidents, suicides, and other violent causes than from other causes. Frequent use of alcohol increased mortality from these causes. CONCLUSION: The risk of accidents, suicides, and other violent deaths was not related to serum cholesterol concentration, whereas such deaths were more prevalent in smokers and alcohol drinkers.

Accidents↗
PubMed1994

Changes in risk factors explain changes in mortality from ischaemic heart disease in Finland.

OBJECTIVES: To estimate the extent to which changes in the main coronary risk factors (serum cholesterol concentration, blood pressure, and smoking) explain the decline in mortality from ischaemic heart disease and to evaluate the relative importance of change in each of these risk factors. DESIGN: Predicted changes in ischaemic heart disease mortality were calculated by a logistic regression model using the risk factor levels assessed by cross sectional population surveys, in 1972, 1977, 1982, 1987, and 1992. These predicted changes were compared with observed changes in mortality statistics. SETTING: North Karelia and Kuopio provinces, Finland. SUBJECTS: 14,257 men and 14,786 women aged 30-59 randomly selected from the national population register. MAIN OUTCOME MEASURES: Levels of the risk factors and predicted and observed changes in mortality from ischaemic heart disease. RESULTS: The observed changes in the risk factors in the population from 1972 to 1992 predicted a decline in mortality from ischaemic heart disease of 44% (95% confidence interval 37% to 50%) in men and 49% (37% to 59%) in women. The observed decline was 55% (51% to 58%) and 68% (61 to 74) respectively. CONCLUSION: An assessment of the data on the risk factors for ischaemic heart disease and mortality suggests that most of the decline in mortality from ischaemic heart disease can be explained by changes in the three main coronary risk factors.

Adult↗
PubMed1994

Changes in serum cholesterol level and mortality: a 30-year follow-up. The Finnish cohorts of the seven countries study.

The association of past changes in serum cholesterol level with cause-specific mortality between 1974 and 1989 was examined in a cohort of 784 Finnish men aged 55-74 years who were free of symptomatic coronary heart disease in 1974. Changes in serum cholesterol level were computed based on measurements made in 1959, 1964, 1969, and 1974. Of the 405 deaths, 202 were due to cardiovascular diseases and 107 due to cancer. Men with a decline in serum cholesterol level between 1959 and 1974 also experienced greater than average declines in body mass index and tended to be more often current smokers in 1974. Among 339 men aged 65-74 years in 1974, men in the lowest tertile of serum cholesterol change, i.e., with greatest declines, had increased cardiovascular (hazard ratio, 1.58; 95% confidence interval 1.00-2.50) and all-cause (hazard ratio, 1.46; 95% confidence interval 1.06-2.02) mortality compared with men in the middle tertile of change, i.e., with least change, in multivariate analysis. Among 445 men aged 55-64 years in 1974, there was a significant U-shaped association between change in serum cholesterol level and coronary and all-cause mortality risk. The authors suggest that both the decline in serum cholesterol level and the associated high mortality may be caused by a third factor, such as increased prevalence of chronic diseases or other changes associated with aging. This would help to explain why several studies have not found an association of serum cholesterol with coronary risk among the elderly.

Aged↗
PubMed1994

Relation of serum homocysteine and lipoprotein(a) concentrations to atherosclerotic disease in a prospective Finnish population based study.

The relation of serum total homocysteine and lipoprotein(a) (Lp(a)) with the incidence of atherosclerotic disease was investigated among 7424 men and women aged 40-64 years free of atherosclerotic disease at baseline in 1977. During the 9-year follow-up, 134 male and 131 female cases with either myocardial infarction or stroke were identified. For each case a control subject was selected belonging to the same sex and 5-year age group. Serum samples collected in 1977 were stored at -20 degrees C and analyzed in 1991. The mean serum homocysteine concentration of male cases and controls was 9.99 mumol/l and 9.82 mumol/l at baseline and that of female cases and controls 9.58 mumol/l and 9.24 mumol/l, respectively. The median serum Lp(a) concentration of male cases and controls was 73 mg/l and 108 mg/l and that of female cases and controls 113 mg/l and 91 mg/l, respectively. The differences between cases and controls were not statistically significant. There was also no significant association between either homocysteine or Lp(a) and atherosclerotic disease, myocardial infarction or stroke in logistic regression analyses. The odds ratios varied from 1.00 to 1.26 for homocysteine and from 0.81 to 1.06 for Lp(a). The results of this prospective population-based study do not support the hypotheses that serum homocysteine or Lp(a) are risk factors for atherosclerotic disease. The lack of association between serum homocysteine and atherosclerotic disease may be due to the exceptionally low gene frequency predisposing to homocysteinemia in Finland.

Adult↗
PubMed1994

Haemostatic factors and prevalent coronary heart disease; the FINRISK Haemostasis Study.

We examined the association of haemostatic factors with prevalent coronary heart disease (CHD) in the population-based FINRISK Haemostasis Study. Of the 2365 participants aged 45-64 years, 88 men and 44 women with prevalent CHD were identified. The participants with CHD were older and had lower high density lipoprotein cholesterol, higher triglycerides and higher body mass index than the participants without CHD. Men with CHD had significantly higher plasma fibrinogen (P < 0.0001, adjusted for age and smoking) and lipoprotein (a) (Lp (a), P = 0.001, adjusted for age) concentrations than those without CHD. Among women, the findings were consistent with those of men, but only fibrinogen reached the conventional level of statistical significance (P = 0.04). No difference was observed in factor VII antigen, factor VII activity or in plasminogen. The association of Lp (a) with prevalent CHD was dependent on fibrinogen concentration (P = 0.02 for the interaction-term) and on non-HDL cholesterol concentration (P = 0.01 for the interaction-term). These cross-sectional results do not prove causality, but they demonstrate a considerable accumulation of prevalent CHD among persons with high Lp (a), fibrinogen and non-HDL cholesterol.

Antigens↗
PubMed1994

Cardiovascular risks and socioeconomic status: differences between men and women in Finland.

STUDY OBJECTIVE: The study aimed to assess the association of different indicators of socioeconomic status with levels of cardiovascular disease risk factors in men and women aged 25-64 years. DESIGN: This was a cross sectional survey, using a community based random sample. SETTING: The provinces of North Karelia and Kuopio in eastern Finland and the cities of Turku and Loimaa and surrounding communities in southwestern Finland in 1987. PARTICIPANTS: Altogether 2164 men and 2182 women aged 25-64 years took part. MEASUREMENTS AND MAIN RESULTS: Data were collected using self administered questionnaires and the measurement of height, body weight, and blood pressure and blood sampling for lipid determinations were done at the survey site. The risk of cardiovascular disease was determined by calculating a simple risk factor score based on the observed values of HDL and total cholesterol, leisure time, physical activity, blood pressure, medication for hypertension, body mass index, and smoking. Indicators of socioeconomic position used were years of education, family income, marital status, and the person's occupation. Lower levels of education, occupation, and income were all significantly associated with an unfavorable risk factor profile in men and women. Education and occupation showed the strongest associations with the risk factor score in both men and women. The results changed little when adjusting for income and marital status. Family income was more strongly associated with the risk factor score in women than men. When adjusting for occupation and education, income was no longer significantly associated with the risk factor score in men. Marital status was not significantly associated with the risk factor score in either sex. CONCLUSIONS: Using the strength of the association with the cardiovascular risk factor score as the criterion for a good socioeconomic indicator, the present study suggests that education and occupation may be equally good indicators in both men and women. Family income may have some additional importance, especially in women.

Adult↗
PubMed1994

Ultrasonographic manifestations of common carotid atherosclerosis in elderly eastern Finnish men. Prevalence and associations with cardiovascular diseases and risk factors.

We investigated the prevalence and associations with cardiovascular symptoms, signs, and risk factors of common carotid atherosclerosis using B-mode ultrasonography in a population sample of 182 eastern Finnish men aged 70 to 89 years. Men were examined in 1989 as a part of the 30-year follow-up examination of the eastern Finnish cohort of the Seven Countries Study. The mean maximal intima-media thickness (IMT) of the right and left common carotid arteries was 1.5 mm (range, 0.7 to 5.3 mm; standard deviation, 0.7 mm). Fifty-one percent of the subjects had nonmineralized atheroma and 91% had single or multiple mineralizations in any of the arterial segments imaged. Both mean maximal IMT and nonmineralized atheromas were associated significantly (P < .05) with the presence of cerebral atherosclerosis, carotid murmur, at least one nonpalpable peripheral arterial pulse, ischemic resting electrocardiographic abnormalities, and history of coronary heart disease but not with intermittent claudication at the 30-year follow-up. No significant associations were found between carotid mineralizations and clinical cardiovascular disease. Long-term elevations of serum cholesterol and long-term smoking, measured as the number of risk factor elevations in the six examinations, were associated with the presence of nonmineralized atheroma in the elderly (in 1989). Smoking and repeatedly detected hypertension, on the other hand, had an association with the presence of mineralizations in 1989.

Adult↗
PubMed1994

Haemostatic factors and lipoprotein (a) in three geographical areas in Finland: the Finrisk Haemostasis Study.

BACKGROUND: An increasing volume of evidence suggests that haemostatic factors play a role in the risk of coronary heart disease. It is not known, however, whether between-population differences in haemostatic factors correspond with the differences in mortality related to coronary heart disease. We examined this question in Finland, where, in North Karelia (in the eastern part of the country), the mortality from coronary heart disease is 1.5-1.7 times higher than that in southwestern areas. METHODS: A random sample of 3000 people aged 45-64 years was drawn from the population registers of North Karelia, of the area surrounding Turku and Loimaa in southwestern Finland and of the Helsinki area in southern Finland. Of the 3000 people approached, 79.6% took part in the study. differences in coronary heart disease mortality and morbidity. RESULTS: Factor-VII coagulant activity was significantly higher in North Karelia than in the other areas (P = 0.0008). The fibrinogen level was also higher in North Karelia, although the difference was significant only among non-smokers (P = 0.02). Levels of factor-VII antigen, plasminogen and lipoprotein (a) did not differ between the areas. Within North Karelia, the levels of both factor-VII coagulant activity, and factor-VII antigen were higher in rural areas than in urban areas. Levels of factor-VII coagulant activity, factor-VII antigen and plasminogen were higher in women than in men and increased with age in women but not in men. The fibrinogen level increased with age in both sexes. CONCLUSION: These baseline findings of the Finrisk Haemostasis Study demonstrate that the geographical differences in levels of factor-VII coagulant activity and fibrinogen in Finland are consistent with the population

Age Distribution↗
PubMed1994

Change in blood pressure and 5-year risk of coronary heart disease among elderly men: the Finnish cohorts of the Seven Countries Study.

OBJECTIVE: To study the association between blood pressure and change in blood pressure with future coronary risk among elderly men. DESIGN: Cohort study. SETTING: Finnish cohorts of the Seven Countries Study. PARTICIPANTS: Four hundred and seventy-six men aged 65-84 years and free of clinically manifested coronary heart disease at baseline, in 1984. MAIN OUTCOME MEASURES: Fatal myocardial infarction (n = 29), any myocardial infarction (n = 42), and incidence of any new signs and symptoms of coronary heart disease (n = 80) during a 5-year follow-up. RESULTS: In multivariate analysis, a significant inverse U-shaped relationship was observed between baseline diastolic blood pressure and future fatal myocardial infarction, any myocardial infarction and any coronary heart disease. In models predicting the risk of coronary heart disease during 1984-1989, there was a significant interaction between both systolic (SBP) and diastolic blood pressure (DBP) in 1969-1974 and change in blood pressure between 1969-1974 and 1984. In categorical analyses, men (n = 42) who experienced a decline in DBP of > or = 4 mmHg from initial levels of > or = 90 mmHg had a higher risk of any myocardial infarction than men (n = 112) with a change of < 4 mmHg (odds ratio 4.5). For a decline of > or = 10 mmHg or more in SBP from levels of > or = 160 mmHg the corresponding odds ratio was 2.9. Men who experienced a decline in DBP or SBP from normotensive levels or an increase in blood pressure had no excess risk compared with men with stabile (change in DBP < 4 mmHg and change in SBP < 10 mmHg) blood pressure values. CONCLUSION: The present results suggest that among elderly men a decline in DBP or SBP from previously hypertensive levels may be associated with increased coronary risk.

Age Factors↗
PubMed1993

Genetic susceptibility to non-insulin dependent diabetes mellitus and glucose intolerance are located in HLA region.

OBJECTIVES: To test the hypothesis that the genetic susceptibility to non-insulin dependent diabetes mellitus is the same as that to insulin dependent disease and to see whether glucose intolerance is associated with specific HLA haplotypes. DESIGN: Population based study of men in 1989 first tested for glucose tolerance in 1984. HLA haplotypes, including HLA-A, C, B, DR, and DQ, were defined serologically. HLA haplotype data from a population based Finnish study of childhood diabetes were used for predicting non-insulin dependent diabetes and impaired glucose tolerance. SETTING: Two communities in Finland. SUBJECTS: Representative cohort of Finnish men aged 70-89, comprising 98 men with non-insulin dependent diabetes mellitus and a randomly selected group of 74 men, who served as controls, who were tested for glucose tolerance twice within five years. MAIN OUTCOME MEASURES: Non-insulin dependent diabetes, impaired glucose tolerance, blood glucose concentration. RESULTS: Diabetes associated HLA haplotypes were present in 94% (85/90) of diabetic subjects, 79% (27/34) of subjects with impaired glucose tolerance, and only 13% (3/23) of non-diabetic subjects. In this group of elderly men sensitivity of the diabetes associated HLA haplotypes for non-insulin dependent diabetes and impaired glucose tolerance was 90%, specificity 87%, and predictive power 97%. Mean fasting blood glucose concentration was only just significantly higher in men with diabetes associated haplotypes than in men with no such haplotypes, but there was a substantial difference in blood glucose values two hours after glucose loading (10.4 and 6.4 mmol/l in men with diabetes associated HLA haplotypes and men with no such haplotypes, respectively (p < 0.0001)). CONCLUSIONS: These findings support the hypothesis that specific HLA haplotypes exhibit a common genetic determinant for insulin dependent and non-insulin dependent diabetes. Furthermore, HLA is a major genetic determinant of glucose intolerance in elderly Finnish men. The belief that the HLA predisposition to diabetes is specific for insulin dependent diabetes mellitus is largely incorrect.

Aged↗
PubMed1993

The effects of saturated fat and n-6 polyunsaturated fat on postprandial lipemia and hemostatic activity.

The effect of different fat loads on postprandial lipemia and hemostatic activity was examined in 10 middle-aged men using 3 different meals. One meal was rich in saturated fatty acids (cream), the other rich in n-6 polyunsaturated fatty acids (sunflower oil) and the third was fat-free containing only carbohydrates. Lipoprotein lipids and hemostatic parameters were measured during fasting and 2, 4, 6 and 8 h after the test meal. In fasting samples, several hemostatic parameters were significantly associated with lipoprotein lipids. Most notable were the strong associations of fibrinolysis parameters tissue plasminogen activator antigen and plasminogen activator inhibitor activity (PAI-1) with total and very low density lipoprotein (VLDL) triglycerides. During lipemia, the associations were approximately similar or slightly weaker than in the fasting state. Both fat loads resulted in similar postprandial lipid responses: VLDL and high density lipoprotein (HDL) triglycerides reached maximum at 4 h after the meal. VLDL cholesterol also increased 4 and 6 h after the fat loads. HDL3 cholesterol declined after the fatty meals but no change was observed in the HDL2 fraction. The fat-free meal gave no significant lipid changes during the time course studied. Factor VII activity (F VII:C) increased 6 and 8 h after the fatty meals, whereas a decrease was observed after the fat-free meal. The changes (+/- S.D.) at 8 h after cream, sunflower oil and fat-free meal were 5.2 +/- 3.3, 3.3 +/- 4.2 and -5.8 +/- 7.9 percentage points, respectively, and the effect of the meal on the changes was statistically significant (F (8,99) = 2.99, P = 0.0048). F VII antigen (F VII:Ag) tended to decline during the day but there was no difference between the meals. Factor VIII activity (F VIII:C) was highest after the polyunsaturated fat meal and lowest after the fat-free meal. PAI-1 declined during the day and the decline tended to be steepest after the fat-free morning meal. The effect of the meal on the changes in lipoprotein lipids and hemostatic factors varied significantly between individuals. In conclusion, postprandial lipemia after a single fatty meal was associated with procoagulatory change in F VII:C but there was no difference between saturated fat and n-6 polyunsaturated fat.

Adult↗
PubMed1993

Prevalence of coronary heart disease and associated risk factors among elderly Finnish men in the Seven Countries Study.

The prevalence of coronary heart disease (CHD), cardiovascular disease (CVD) and associated risk factors was studied in 413 men aged 70-89, the survivors of the Finnish cohorts of the Seven Countries Study. Men were divided into five categories according to manifestations of prevalent CVD: I, history or ECG evidence of previous myocardial infarction (MI; 48 men, 12%); II, typical angina pectoris (AP; 56 men, 14%); III, other ischaemic electrocardiographic (ECG) changes (82 men, 20%); IV, stroke, transient ischaemic attack, intermittent claudication or minor ECG changes (other CVD; 78 men, 19%); V, free of CVD (149 men, 36%). Both systolic and diastolic blood pressures were lowest in men with previous MI and in men free of CVD, and highest in men with other ischaemic ECG changes (P = 0.017). Low HDL-cholesterol (< 0.9 mmol/l) was more prevalent and the total/HDL-cholesterol ratio and triglyceride levels were higher in men with prevalent CHD (P < 0.05). Total and LDL-cholesterol, smoking, body mass index, fibrinogen, coagulation factor VIIc, apolipoprotein A-I, apolipoprotein B and lipoprotein(a) were not associated with prevalent CVD. The results show that manifestations of CHD and CVD are common among elderly Finnish men. Low HDL-cholesterol, total/HDL ratio, triglycerides and hypertension were associated with manifest CVD cross-sectionally.

Aged↗
PubMed1993

Prevalence and change of cardiovascular risk factors among men born 1900-19: the Finnish cohorts of the Seven Countries Study.

In a 30-year follow-up survey of the Finnish cohorts of the Seven Countries Study in 1989, 470 men aged 70-89 years were examined in two rural areas of Finland, in the East and in the West. Life-style-related coronary heart disease risk factors were at high levels in both groups, but the difference between areas found in the same cohort in middle age had mostly disappeared or partially reversed. Mean levels of systolic/diastolic blood pressure, body mass index, fasting blood glucose, blood glucose after 2-hour oral glucose tolerance test, and plasma fibrinogen were higher in the West than in the East (p < or = 0.05), but serum thiocyanate which reflects smoking habits was higher in the East (p < 0.001). No significant differences between areas were found in mean levels of serum total cholesterol, LDL-cholesterol, HDL-cholesterol, triglycerides, apolipoprotein A-I, apolipoprotein B, lipoprotein (a), plasma coagulation factor VIIc, and prevalence of current smoking.

Aged↗
PubMed1993

Does the predictive value of baseline coronary risk factors change over a 30-year follow-up?

The association of baseline serum total cholesterol, systolic blood pressure, smoking and body mass index with coronary heart disease (CHD) mortality was analyzed among 1,619 men aged 40-59 at baseline. Analyses were made separately for the first, second and third decade of follow-up. Serum cholesterol and smoking more than 9 cigarettes daily were strong predictors of risk of CHD death (n = 450) occurring early and late during the 30-year follow-up. After 20 years of follow-up, systolic blood pressure was no longer associated with CHD risk. In contrast, highest tertile of body mass index (over 24.7 kg/m2) was only then associated with increased CHD risk. The correlations between the baseline and the 30-year risk factor values were 0.42 for serum cholesterol (n = 444), 0.28 for systolic blood pressure (n = 444) and 0.57 for body mass index (n = 429). Our results showed large differences in the long-term predictive power of the classical coronary risk factors. The reasons for these differences are discussed.

Adult↗
PubMed1993

Changes in glucose tolerance among elderly Finnish men during a five-year follow-up: the Finnish cohorts of the Seven Countries Study.

UNLABELLED: It was the purpose of this study to report the natural history of glucose tolerance during a five-year follow-up among elderly Finnish men, and to evaluate the role of age and body-mass index in explaining the variation in glucose tolerance both cross-sectionally and longitudinally. In the survivors of the Finnish cohorts of the Seven-Countries Study, aged 65 to 84 years at baseline, a two-hour oral glucose-tolerance test was performed according to current WHO criteria. Subjects with fasting blood glucose > 10 mmol/l were directly classified as having diabetes at baseline.-- FINDINGS: At baseline, of the 637 men 216 had normal and 234 had impaired glucose tolerance, 187 were diabetic. At follow-up, 172 men had died; 38 (18%) of the subjects with normal glucose tolerance at baseline had either impaired glucose tolerance or diabetes; 17 (7%) of the men with initially impaired glucose tolerance had developed diabetes, and 79 (34%) were normalized. 25 (13%) and 22 (12%) of the initially diabetic subjects had reverted to impaired or normal glucose tolerance, respectively. The age was able to explain 1-2% of variation in blood glucose level in cross-sectional but not in longitudinal comparison. Body-mass index was an important predictor of abnormal glucose tolerance in previously normally responding men. By contrast, obesity did not contribute to the development of diabetes among men with impaired glucose tolerance.-- PRINCIPAL CONCLUSIONS: The incidence of impaired glucose tolerance was high in these elderly Finnish men as compared with studies in middle-aged Caucasoid subjects.(ABSTRACT TRUNCATED AT 250 WORDS)

Age Factors↗
PubMed1992

Short- and long-term association of serum cholesterol with mortality. The 25-year follow-up of the Finnish cohorts of the seven countries study.

The association of serum cholesterol with cause-specific and all-cause mortality was assessed in a cohort of 1,426 men aged 40-59 years who were free of clinically evident heart disease at baseline (1959). A total of 748 deaths (53 percent of the participants) occurred during the 25-year follow-up period. Men with high serum cholesterol levels at baseline had high mortality due to coronary heart disease during both the early and later parts of the follow-up period. In contrast, the association of serum cholesterol with mortality due to causes other than coronary heart disease changed during follow-up (interaction of cholesterol with follow-up period: p = 0.004). During the first 10 years of follow-up, despite their high coronary mortality, men with high cholesterol levels had lower all-cause mortality (age-adjusted relative risk = 0.71 for serum cholesterol above 5.79 mmol/liter vs. below 5.80 mmol/liter; p = 0.03) because of their low cancer mortality (relative risk = 0.55, p = 0.03) and residual mortality (relative risk = 0.49, p less than 0.01). During the last 15 years of follow-up, cholesterol at baseline was no longer associated with mortality due to causes other than coronary heart disease, and consequently, because of their high coronary mortality, men with high cholesterol levels also had higher all-cause mortality (relative risk = 1.22, p = 0.05). The results suggest that to fully analyze the association of serum cholesterol with all-cause mortality, the follow-up period should be sufficiently long--possibly more than 10 years--and the possibility of a change in the direction of the association studied should always be considered.

Adult↗