The authors examined the effect of several risk factors in relation to coronary heart disease (CHD) mortality in the Corfu cohort of the Seven Countries Study. The population studied in this analysis consisted of 529 rural men (age 40–59 years) enrolled in 1961. Multivariate analysis was performed with CHD death as the end point; age, blood pressure, heart rate, serum total cholesterol, smoking, physical activity, body mass index, skinfold thickness, vital capacity, and forced expiratory volume were the predictors. The 40-year CHD survival rate was 63% (108 deaths). Age (hazard ratio [HR]=1.093; p<0.001), smoking (HR=1.79; p<0.05), body mass index (HR=1.05; p<0.1), and serum total cholesterol (HR=1.004; p<0.2) were independently associated with 40-year CHD mortality. Conclusively, in men from the Corfu cohort, age, lifestyle habits (expressed as smoking and body mass index), and serum cholesterol levels were consistently associated with 40-year CHD mortality. In addition, the current status of men in the Corfu cohort is marked by long-term adoption of a Mediterranean type of diet, physical activity, and optimism.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
By the 1950s and early 1960s a remarkable series of epidemiologic studies had begun, some of which continue to the present. The goal of these studies was to identify factors that could explain either differences in rates of occurrence of coronary heart disease (CHD) between populations or differences in risks of coronary events among individual members of a population. The Seven Countries Study on cardiovascular diseases has been completed in most of the 16 cohorts, marking approximately four decades of prospective investigation. In the mid-1950s, Keys1 enrolled and examined more than 12,000 men. Since then, several analyses2–5 have been performed showing that the Mediterranean cohorts, and especially the Greeks (Corfu and Crete), had a marked decrease in cardiovascular death rates, among all the investigated cohorts. According to previous analyses2–5 (10–25 years of follow-up), baseline age, systolic blood pressure, and total cholesterol levels, as well as smoking habits were significant predictors for CHD mortality in the Greek cohorts. However, during recent decades, Greece has experienced marked but uneven socioeconomic development which has been associated with considerable alteration in lifestyle. Stable, age-old dietary habits and high habitual physical activity have gradually given way to “Western”-type diets and a more sedentary lifestyle.6–8 Additionally, smoking prevails in almost one half of the male population.9 Thus, it remains unclear why a population with few healthy habits still enjoys one of the longer life expectancies and the lowest rates of CHD mortality among the 16 cohorts of the Seven Countries Study.10
In order to evaluate predictors for CHD mortality, we examined the information obtained from the 40 years of the Corfu cohort of the Seven Countries Study. To the commonly studied baseline risk factors (i.e., age, systolic and diastolic blood pressure, smoking, serum total cholesterol, and body mass index [BMI]) we added physical activity status, vital capacity, forced expiratory volume, heart rate, and arm circumference (as an index of muscularity). Also, we evaluated entry values of triceps and subscapular skinfold, as indexes of body fat. Additionally, current information regarding the previous factors as well as psychologic status and several lifestyle habits (physical activity, nutritional habits, afternoon siesta, psychosocial stress, and depression) were recorded.
METHODS
Study Population
In the early 1960s, Keys and colleagues1,2 enrolled 12,763 middle-aged men from 16 cohorts in the Seven Countries Study (United States, Italy, Japan, Netherlands, Greece, former Yugoslavia, and Finland). Included from Greece were 1215 rural men from the Corfu and Crete islands. The population studied in this analysis was the Corfu cohort. From September to October 1961, an international team of experts (physicians, nurses, and technicians) examined 529 men from six villages (San Marcus, Scriperon, Sokraki, Doucades, Gardelades, Korakiana), who were almost all (95.3%) between 40–59 years old in the early fall of 1961. The group was mainly small-scale, non-mechanized farmers (55%) at entry.10
Follow-Up
Forty years after entry into the study, survivors of the Corfu cohort were examined by a team of specialists from the Cardiology Department of the University of Athens Medical School under the guidelines of the Seven Countries Study coordinators. Collection of data on mortality was complete for all men for the subsequent 40 years, through visits at the aforementioned villages. Causes of death were obtained from previous clinical records completed by the study's research group or from hospital records, necroscopy records, or information from family or hospital physicians, other specialists, family or relatives, and any other witnesses.1,2 Information from the police, in cases of violent death, was also obtained. The coding of death was based on standard criteria according to the Laboratory of Physiological Hygiene (LPH) codes.1 The LPH mortality codes including coronary heart disease are summarized in Table I.
Table I. Laboratory of Physiologic Hygiene Codes (LPH)
LPH Code 710
Coronary heart disease, myocardial infarction, not sudden
Coronary heart disease, severe (recognized) arrhythmia, sudden
LPH Code 721
Coronary heart disease, sudden without further specification
LPH Code 730
Coronary heart disease, with heart failure or arrhythmia, not sudden
LPH Code 731
Coronary heart disease, with sudden heart failure, usually pulmonary edema
LPH Code 740
Hypertensive heart disease: WHO-ICD, 402 and 404
WHO-ICD=World Health Organization-International Classification of Diseases
For practical purposes, it is advised by the coordinators of the study to use, for CHD, the block of 710 to 731. However it has been shown that 740 (hypertensive heart disease) is predicted by the same risk factors as CHD and, therefore, code 740 can be joined to the 710–731 block in order to identify the bulk of heart diseases and conditions likely bound to atherosclerosis and hypertension.1
Parameters Investigated
In the present analysis, the following entry information has been considered: age; physical activity status (classified as sedentary or moderately active to hard physical work); supine resting systolic and diastolic blood pressure of the right arm in mm Hg (mean of two consecutive measurements); resting heart rate (calculated from the resting electrocardiogram); daily cigarette smoking status (number of cigarettes smoked, as reported by the subjects); total serum cholesterol (mg/dL) measured in a casual blood sample by the Abell-Kendall method, modified by Anderson and Keys11; BMI (kg of weight divided by m2 of height); unclothed right arm circumference (mm); triceps and subscapular skinfold thickness (measured to the nearest 0.5 mm on the unclothed chest just below the tip of the right scapula); vital capacity in dL/height in meters and forced expiratory volume in 3/4 seconds (dL/height in meters).
In order to describe the present status of men alive at 40 years of follow-up, most of the previous measurements using the same techniques were also obtained. In addition, we assessed daily food consumption, based on a validated questionnaire that has been used in several analyses in the past,7 mental status, according to the Mini-Mental State questionnaire,12 and the presence of psychologic stress and depression, according to the Geriatric Depression Scale (GDS) questionnaire.13 For the purposes of the analysis the GDS was weighted by a group of specialists at the Hellenic Institute of Psychometric Research. Thus, a sum of the GDS score over 5 defines significant depression.13 Finally, current information regarding physical activity, smoking habits, and afternoon siesta (nap or rest) was recorded from all men. Further details regarding the protocol and procedures applied in the Seven Countries Study have been presented in detail elsewhere.1,5,8,14,15
Statistical Analysis
Continuous variables are presented as means±1 SD, while qualitative variables are presented as absolute and relative frequencies. Death rate was calculated using the observed person-time, in years. Multivariate analysis was performed using the Cox proportional hazards model with CHD death as the end point and the aforementioned risk factors as predictors. The proportion of surviving persons was recorded every 6 months. Subjects with missing information on risk factors were excluded from the analysis. The final model was developed through backward elimination procedures for the selection of variables, based on the Wald's statistic and using 5% of the probability for inclusion and 10% of the probability of removal of a variable from the survival model. Deviance residuals were calculated in order to evaluate the model's goodness-of-fit. The assumption of proportionality was graphically assessed. SPSS 10 software (SPSS Inc., Chicago, IL) was used for all statistical calculations.
RESULTS
Table II presents baseline information regarding the investigated factors described in the methodology section.
Table II. Entry Characteristics of Men in the Corfu Cohort (1961)
Number
529
Age (yr)
49.7±5.7
Prevalence of CHD at entry
3
Systolic blood pressure (mm Hg)
136.8±19.9
Diastolic blood pressure (mm Hg)
82.65±11.65
Heart rate
69±13
Hypertension (>140/90 mm Hg)
156 (29%)
Serum total cholesterol (mg/dL)
204.5±46.0
Hypercholesterolemia (>220 mg/dL)
168 (32%)
Smoking (>=1 cigarette/day)
393 (74.3%)
Cigarettes (n/day)
11.1±10.7
Physical activity (yes)
362 (68.4%)
Body mass index (kg/m2)
25.3±3.5
Arm circumference (cm)
25.2±2.1
Subscapular skinfold (mm)
10.7±5.7
Triceps skinfold (mm)
7.2±4.3
Vital capacity (dL/m)
22.3±3.6
Forced expiratory volume (dL/m)
15.7±3.2
CHD=coronary heart disease
At 40 years of follow-up, 120 (26%) of 461 deaths were attributed to CHD, i.e., eight events per 1000 person-years of observation. Figure 1 demonstrates causes of death, while Table III shows the cumulative survival rates.
Cumulative coronary heart disease (CHD) survival curve, for the Corfu cohort during a 40-year period (1961–2001)
Table IV shows the results of the multivariate survival model. The levels of the investigated factors, at entry, are presented as the exponentials (hazard ratios [Hrs]) of the estimated coefficients. Among the investigated risk factors, we found that age, smoking, and BMI were significant determinants for CHD mortality (Table IV). Furthermore, we concluded that 40 mg/dL differences in baseline serum total cholesterol levels were associated with a 17% excess CHD mortality (HR=1.17, p<0.18) after controlling for age, smoking habits, BMI, and exclusion of the other covariates previously presented from the final model.
Table IV. Estimates From a Backward Step-Wise Cox Model Predicting 40-Year CHD Mortality as a Function of Baseline Risk Factors Levels
Risk Factor
Hazard Ratio
95% CI
p Value
Age at entry (per 1 year)
1.10
1.06–1.14
<0.001
Smoking status (yes vs. no)
1.79
1.15–2.77
0.010
Body mass index (per 1 kg/m2)
1.05
0.99–1.10
0.090
Variables entered in the model: age, systolic and diastolic blood pressure, smoking, serum cholesterol, body mass index, physical activity status, vital capacity, forced expiratory volume, arm circumference, triceps and subscapular skinfold CI=confidence interval
Finally, Table V briefly demonstrates the current social, lifestyle, and psychologic status of the surviving men alive from the Corfu cohort. In addition to the results presented in Table V, we found that the majority of the men (64 out of 67) had excellent mental status (according to the Mini-Mental State scale) in that only one out of four reported no friends or feeling abandoned, whereas the rest usually met their family, friends, and neighbors almost daily.
Table V. Current Habits of the 67 Surviving Men at the 40-Year Follow-Up of the Corfu Cohort From the Seven Countries Study
Alcoholic beverages consumed (wineglass of 100 mL per week)
14±5
Weekly Nutritional Habits
Never-Rare (0–1 d/wk)
Regularly (2–4 d/wk)
“Every”day (5–7 d/wk)
Vegetables
3%
12%
85%
Meat
73%
22%
5%
Fruits
2%
15%
83%
Eggs
20%
57%
23%
Dairy products (milk, cheese, etc.)
18%
61%
21%
Bread
5%
8%
87%
Olive oil
2%
3%
95%
GDS=Geriatric Depression Scale questionnaire
DISCUSSION
In this paper we present mortality data on 40 years of follow-up of the Seven Countries Study for the first time. The survival analysis in men enrolled in the early 1960s in the Corfu cohort showed that age, smoking, BMI, and total cholesterol levels were significant predictors of CHD mortality (Table IV). Additionally, the proportion of CHD deaths, during four decades of investigation, varied from 16%–28.5% of all causes of death (Figure 1), supporting the previously reported moderate to low CHD mortality in the investigated cohort.10
According to previous analyses (10-year follow-up) from the Seven Countries Study,2 the relation between smoking and CHD event rates was strongest for populations with the highest rates (Northern Europe) and notably less strong for the populations with lower rates (Yugoslavia, Italy, and Greece). Regarding the present cohort, the majority of the subjects (74.3%) were smokers at entry. The present analysis showed that smoking habits (>1 cigarette/day) at entry, constitute a highly significant predictor for cardiac deaths, constituting a 59% increase in the CHD hazard ratio (Table IV). Previous results from the Seven Countries Study showed that serum cholesterol was strongly related to CHD mortality both at the population and the individual levels.16 The strength of the association between serum cholesterol and CHD mortality was similar in different cultures. However, the absolute risks differed substantially. Kromhout16 reported that at a serum cholesterol level of 200 mg/dL, the 25-year CHD mortality rate was five times higher in Northern Europe compared to Mediterranean Southern Europe. Baseline measurements in the Corfu cohort showed that the prevalence of hypercholesterolemia (serum total cholesterol levels<220 mg/dL) was 31.8% (168 men). In addition, our analysis showed that total cholesterol measured at baseline constitutes a significant marker for CHD mortality in the investigated cohort (Table IV). Moreover, we observed that a 40 mg/dL difference in baseline serum total cholesterol levels is associated with an insignificant 17% excess risk of 40-year CHD mortality. However, according to Kromhout,16 it can be concluded that the relations between diet, serum cholesterol, and CHD are more complex than originally thought, because not only dietary cholesterol but other lipids and antioxidants may play a role in the genesis of atherosclerosis.16
A US consensus panel report has concluded that persons whose relative weight is 20% or more above the desirable weight for sex and height are at increased risk of chronic diseases and require special treatment. However, many investigators suggest that obesity and smoking are strongly confounded; that is, smokers are often found in the lowest category of obesity.17 In the present study, we found that BMI is significantly associated with CHD mortality, independent of smoking habits. Moreover, the analysis showed that obesity at baseline doubles the hazard of CHD deaths compared to normal weight (HR=2.023; p<0.05). The prospects for increasing obesity with economic and social development raise serious concern for public health strategies.17
In addition, it should be noted that those men alive after 40 years of follow-up were more likely to be physically active (at least three times per week), optimistic, nonsmoking, to practice afternoon napping, and consume olive oil, fresh fish, fruits, and vegetables. Many investigators report that dietary habits have changed during the past 20 years in the Greek population, approaching a “Western” type of diet.7 Nevertheless, the men of the Corfu cohort showed persistent adherence to the diet of their grandparents (Table V). Previous analyses have shown that the adoption of a dietary pattern rich in fruits, vegetables, and olive oil explains, in part, the large differences in 25-year CHD mortality of the Mediterranean cohorts of Italy and Greece compared to cohorts from northern Europe and the United States and even those of neighboring inland Mediterranean countries such as Croatia and Serbia (former Yugoslavia).18,19 The protective effect of diet seems to be related to lower BMI, blood pressure, and lipid levels, as well as possible improvement in endothelial function, decreases in inflammation and oxidation, and reduced insulin resistance.18–20 Our findings support the benefits of the Mediterranean diet on CHD mortality.
LIMITATIONS
The exact cause of death in several cases in this study could not be ascertained.14 Thus, the calculated cause-specific HR, after adjustment for several predictors, may be over- or underestimated. Also, regarding the current status of three out of the 67 surviving men, we obtained the information from family or neighbors, since they themselves were unable to provide accurate information. Finally, the small number of men alive after 40 years increases the uncertainty of the mathematical calculations of the estimated survival.
CONCLUSIONS
In previous analyses, the markers for longevity due to CHD were age, smoking habits, systolic blood pressure, and total cholesterol levels.4,10,15,21 In the present work with the exception of blood pressure levels, age, smoking habits, total cholesterol, and additionally, BMI, were significant determinants for the 40 years' CHD mortality. Moreover, the surviving men in the Corfu cohort benefited from the long-term adoption of a nutritional pattern close to the Mediterranean diet, the presence of physical activity, optimism, and a positive psychologic profile.
Acknowledgements: We would like to express our gratitude toDr. Dia Pitsavos, Mrs. Klio Vlahou-DontasandDr. Anastasia Katiniotifor their substantial assistance in the completion of the 40 years follow-up. We are grateful particularly to the men of the villages of Corfu whose cooperation made this follow-up study feasible. Also, we would like to thank the mayors of the villages, the registrar of the city of Corfu, the director and the staff of the San Marcos Medical Center, as well as the microbiological laboratory of the Corfu Prefectorial Hospital for their substantial support at the follow-up. The Seven Countries Study was funded by grants HE 04697, HE 6090, and HE 00278 from the National Heart, Lung, and Blood Institute, as well as by a grant from the Hellenic Heart Foundation.
