We studied the ecologic relationships of food groups, macronutrients, eating patterns, and an a priori food pattern score (Mediterranean Adequacy Index: MAI) with long-term CHD mortality rates in the Seven Countries Study.
Sixteen cohorts (12,763 men aged 40–59 years) were enrolled in the 1960s in seven countries (US, Finland, The Netherlands, Italy, Greece, former Yugoslavia: Croatia/Serbia, Japan). Dietary surveys were carried out at baseline and only in a subsample of each cohort. The average food consumption of each cohort was chemically analyzed for individual fatty acids and carbohydrates.
Ecologic correlations of diet were computed across cohorts for 50-year CHD mortality rates; 97% of men had died in cohorts with 50-year follow-up. CHD death rates ranged 6.7-fold among cohorts. At baseline, hard fat was greatest in northern Europe, olive oil in Greece, meat in the US, sweet products in northern Europe and the US, and fish in Japan. The MAI was high in Mediterranean and Japanese cohorts. The 50-year CHD mortality rates of the cohorts were closely positively ecologically correlated (r = 0.68–0.92) with average consumption of hard fat, sweet products, animal foods, saturated fat, and sucrose, but not with naturally occurring sugars. Vegetable foods, starch, and the a priori pattern MAI were inversely correlated (r = −0.59 to −0.91) with CHD mortality rates.
Long-term CHD mortality rates had statistically significant ecologic correlations with several aspects of diet consumed in the 1960s, the traditional Mediterranean and Japanese patterns being rich in vegetable foods, and low in sweet products and animal foods.
The relationship between diet and coronary heart disease (CHD) has been studied systematically in the Seven Countries Study (SCS) since mid-20th century [1,2,3,4]. Cross-cultural analyses among the 16 SCS cohorts found that the average population intake of saturated fatty acids (SFA) was positively associated with population CHD mortality rates through 25 years of follow-up [2, 3, 5,6,7]. Since 2009 seven meta-analyses of randomized controlled trials with CHD as endpoint showed that replacement of 5% of energy from SFA by polyunsaturated fatty acids reduced CHD mortality by 5–10% [8,9,10,11,12,13,14]. However, replacement of SFA by carbohydrates did not affect cardiovascular disease in the Women's Health Initiative . Thus, substituting polyunsaturated fat for SFA is preferable to replacing SFA by carbohydrates.
In the 1990s, an important contrasting theory was that low carbohydrate diets would promote weight loss and be generally healthful . Recently, meta-analyses of trials showed that added sugars or sugar-sweetened beverages increased body weight [17, 18]. A meta-analysis of cohort studies observed that a high consumption of sugar-sweetened beverages or soft drinks is associated with a higher risk of type 2 diabetes . A systematic review showed that the consumption of sugar-sweetened beverages is weakly associated with a higher incidence of hypertension . Increased body weight, diabetes, and hypertension are strong CHD risk factors.
The SCS surveys established the characteristics of traditional Mediterranean diets in Southern Italy, Dalmatia and Greece . The traditional Mediterranean diet was defined by Fidanza as the diet “which is usual for people from countries whose coasts are washed by this sea, and is moderate in cereal products, fish, legumes, olive oil, fruit, and vegetables in combination with little meat and wine” . The main characteristics of this diet pattern were that it was plant-food based, with olive oil as the principal source of fat. There were substantial variations in the Mediterranean pattern among SCS cohorts . To better describe these patterns, Fidanza and collaborators developed an a priori food pattern score, the Mediterranean Adequacy Index (MAI) [24, 25]. The MAI was inversely related to CHD mortality rates through 25 years follow-up in the SCS .
Here we extend the follow-up to 50 years and add new information on the ecologic associations of macronutrients, foods, and eating patterns with long-term CHD mortality. We also study the association of the MAI with population average intake of dietary fats and carbohydrates, with a view toward dietary patterns rather than macronutrient intake.
Material and methods
Of the 16 cohorts in the SCS, 11 were rural: East Finland and West Finland (Finland), Crevalcore and Montegiorgio (Italy), Dalmatia and Slavonia (Croatia, former Yugoslavia), Velika Krsna (Serbia, former Yugoslavia), Crete and Corfu (Greece), and Tanushimaru and Ushibuka (Japan). The remaining five cohorts were the US and Rome railroad, men from the town of Zutphen, The Netherlands and two Serbian cohorts, namely workers in a large agricultural cooperative in Zrenjanin and Belgrade university professors. A total of 12,763 men aged 40–59 were initially examined in the late 1950s and early 1960s [2,3,4].
Baseline data were collected before the era of the Helsinki Declaration, with consent implied by participation in the examinations, while verbal or written consent was obtained to collect follow-up data.
In subsamples of 13 cohorts from 1959 to 1964, a single, multiday dietary survey per man was carried out at baseline, spread over each subsample across all seasons of the year, each man’s dietary survey was comprised of food records with precise weighing and chemical analysis for macronutrients . Only food records were collected in the US railroad in 1960–1962, the Rome railroad in 1969 and in Ushibuka, Japan in 1971 . The estimated average diet in Rome and Ushibuka was assumed to be similar to the average diets ~10 years earlier. The 498 men in diet survey subsamples across cohorts provided 3282 days of food consumption surveyed. These were random subsamples of all participants in each cohort, with replacements for the few who refused the diet survey. Additional details of SCS diet methods are presented in the Supplementary Dietary Methods Appendix online.
Foods and food groups
Individual food items and their weighed amounts eaten were recorded for each participant. All foods eaten by the participants in the subsamples at entry were coded in a standardized way as the edible part of raw products . For this report, the foods were classified into 15 food groups (listed in the next paragraph) and a group of foods not otherwise classified.
The MAI is an a priori food pattern score computed according to Fidanza et al. [24, 25], based on 15 food groups all expressed as gram per 1000 kcal at baseline . The numerator of the score included foods typical of the traditional Mediterranean diet (bread, cereals, legumes, potatoes, vegetables, fruit, fish, vegetable oils, mainly olive, and wine), while the denominator includes foods not typical of the Mediterranean diet (high-fat milk (solid fraction), cheese, meat, eggs, hard fat (butter, lard, and hard margarine), and sweet products. High scores for the MAI represented the “traditional” Mediterranean diet. In this analysis, all alcoholic beverages replaced wine due to the wide variety of drinking patterns (wine, beer, and hard liquor) in the different countries.
Chemically analyzed nutrients
In 1987–1988, foods identical to those consumed in the 1960s were purchased from local markets for each of the 16 cohorts and shipped for chemical analysis in the laboratory of the Division of Human Nutrition of Wageningen University, The Netherlands. The foods were cleaned and food composites were prepared according to the average consumption pattern of each cohort. The food samples were analyzed for total proteins, total and individual fatty acids and carbohydrates, and total dietary fiber. Additional details are presented in the online Supplement.
Mortality data collection and coding
Mortality data for each man, including person years to death or censoring and causes of death, was collected systematically. During 25 years of follow-up 13 men from the two Croatian cohorts and after 45 years six men from two Serbian cohorts were lost to follow-up. Of the other 12 cohorts 20 men from the US, two from Zutphen, one from Crete, and 25 from the two Japanese cohorts were also lost to follow-up. In the remaining seven cohorts no one was lost to follow-up.
Causes of death were coded following the 8th Revision of the WHO-ICD , adjudicated by a single reviewer (AM) following predefined criteria. In the presence of multiple causes of death or serious doubt about cause, cause was assigned hierarchically as violence, cancer, CHD, stroke, and other causes. CHD deaths included the ICD-8 codes 410, 411, and 412, plus cases of sudden death attributable to CHD, after exclusion of other reasonable causes.
Total energy was calculated from the intake of total proteins, fats, and carbohydrates. For six cohorts: one from Italy, two from Croatia, and three from Serbia, 50-year mortality data were estimated using regression equations derived from the ten cohorts with complete mortality data. The dependent variable was the 50-year findings and the independent variables were either 25- or 45-year findings (Supplementary Table 1). Observed and predicted 50-year CHD mortality rates per 1000 person-years were adjusted by linear regression for the mean cohort age.
Only ecologic analysis of diet data is possible in the currently extant data. Pearson linear correlation coefficients were computed among the average population food groups (g/1000 kcal), the MAI, macronutrients (expressed as percent of total energy intake) and 50-year CHD death rates. Correlation coefficients of >0.50 for 16 cohorts and >0.63 for ten cohorts were statistically significant, two-sided P-value <0.05. For analytical purposes the MAI was transformed by the natural log, which improved the fit.
Around 1960 there were large differences among cohorts in consumption by food groups (Table 1). The highest consumption of hard fat was found in northern Europe. In the Mediterranean areas, olive oil was the main vegetable oil, with the highest consumption in Greece. The greatest consumption of meat was in the US and of sweet products in the US and northern Europe. Japan had the highest consumption of fish and legumes . Mediterranean-style diets were observed in Italy, Greece, and Dalmatia; the Japanese cohorts also had high MAI. In ten cohorts 97% of men had died after 50 years. The highest 50-year CHD mortality rates were in the US, northern and central Europe and the lowest in the Mediterranean areas and Japan, a 6.7-fold range from 2.6 to 17.3/1000 person-years.
The average population consumption of cereals, vegetables, legumes, and alcohol (r = −0.52 to −0.62) was inversely associated with the long-term CHD mortality rates (Table 2), while higher rates were found in cohorts eating predominantly hard fat, high-fat milk and sweet products (r = 0.69 to 0.84). The combined food groups hard fat plus sweet products (r = 0.84) and animal foods (r = 0.68) were positively related to 50-year CHD morality rates. The combined vegetable foods (r = −0.59) and the natural log of the MAI food pattern score (r = −0.91) were also inversely related to CHD mortality (Fig. 1).
Large variations in average population intakes were observed for fatty acids (Supplementary Table 2). There was a 4-fold range across cohorts in the average total fat intake, 6-fold for SFA, 2-fold for total carbohydrate, 3-fold for total dietary fiber, and 4-fold for total sugars and starch. The 50-year mortality rates of cohorts were positively related to average intake of total fat and inversely to total carbohydrate (Supplementary Table 3). Average cohort intake of SFA was positively associated with 50-year CHD mortality rates (r = 0.92) (Fig. 2) and a similar result was observed for the sum of SFA and trans fatty acids (Supplementary Figure 1).
The average intake of sucrose was related to 50-year CHD mortality rates (r = 0.71, Fig. 3) but the remaining naturally occurring sugars were not associated (Supplementary Table 3). The population intake of sucrose was correlated with saturated fat (r = 0.63) (Supplementary Figure 2). The average intake of starch was inversely associated with 50-year CHD mortality rates (r = −0.71) (Supplementary Figure 3). The average intake of total dietary fiber was not associated with 50-year CHD mortality rates (Supplementary Table 3). The population average SFA intake was statistically nearly indistinguishable from the MAI (r = −0.96) (Fig. 4). The MAI was also inversely correlated with sucrose (r = −0.66) (Supplementary Figure 4) and positively with the intake of starch (r = 0.73) (Supplementary Figure 5).
These ecological findings among habitual diet and CHD death rates showed that dietary characteristics estimated among cohorts of the SCS at baseline in the 1960s continued to relate to CHD death rates after 50 years of follow-up when almost all men had died. The associations with coronary deaths were direct for the average intake of SFA and sucrose, hard fat, sweet products and animal foods, and inverse for starch and vegetable foods. The MAI closely represented all these aspects of diet and was related to low population long-term CHD mortality rates in both the Mediterranean and Japanese cohorts.
The 50-year data confirm and extend a finding of the SCS in the earlier 25-year data: average SFA intake was strongly related to CHD mortality rates [2, 3, 5,6,7]. However, the major SFA: lauric acid (C12:0), myristic acid (C14:0), palmitic acid (C16:0), and stearic acid (C18:0) and the trans fatty acid elaidic acid (C18:1T) were strongly correlated (r > 0.8) with total SFA and could not be reliably separated from each other in ecologic analysis of the 16 cohorts . Therefore, we limited the statistical analysis to the association of total SFA intake with long-term CHD mortality rates. The effects of the individual fatty acids can only be investigated in controlled dietary intervention studies.
Critics frequently questioned the primary ecologic finding of the SCS on the associations of the population SFA intake with serum cholesterol and CHD mortality rates . The arguments are that the countries participating in the SCS were selectively excluded on the desired outcome and that France was purposefully excluded. It is true that selection of the cohorts was not-random, being based on dietary variation and practicality. Colleagues who were interested in the study had to participate in one of the pilot studies and to follow the study protocol. Researchers from France did participate in a pilot, but decided not to participate due to lack of funds. Already the first publication of the SCS in 1970 showed that the actual difference in average serum cholesterol between Finland and Greece was 1.53 mmol/L . The Keys equation, based on controlled cross-over dietary experiments, predicted a difference of 1.07 mmol/L . A part of the 30% difference could also be explained by a serum cholesterol lowering effect of apples and legumes, and life-long differences in dietary patterns . So, we conclude that replacement of SFA by polyunsaturated fatty acids causally reduced serum cholesterol and CHD mortality rates as shown in trials with at least 2 years of follow-up . We also note that high SFA intake is embedded in a dietary pattern, as discussed below.
Critics also contend that sugar was not considered as possible contributor to CHD in the SCS . We observed a positive relationship of the population intake of sucrose with 50-year CHD mortality rates. Similar correlations were found by Keys after 5- and 10-years of follow-up [3, 30]. Based on partial correlation analysis Keys concluded that SFA, but not sucrose, was related to CHD . However, in case of correlated intakes, sucrose and SFA were both associated with CHD mortality rates and can not be disentangled. In the 1960s the traditional Western diets were high in SFA and sucrose, while the traditional Mediterranean and Japanese diets were low in SFA and sucrose (Figs. 2 and 3). In 2015, the WHO endorsed an added sugar (sucrose) recommendation of <50 g/day and suggested a further reduction of <25 g/day . This is in accord with the low CHD risk of the traditional Mediterranean and Japanese diets that contained <25 g of added sugar per day. The remaining naturally occurring sugars (glucose and fructose in fruit, lactose and galactose in milk, and maltose in bread and pasta) were not related to 50-year CHD mortality rates. This could be due to opposite health effects of nutrients and bioactives present in foods that also contain sugars.
High intakes of the macronutrients SFA and sucrose and the food groups hard fat and sweet products were characterized by high 50-year CHD mortality rates. The intake of starch was correlated (r = 0.74) with the consumption of the basic foods bread, cereals, legumes, and potatoes. In the 1960s, foods were prepared at home and had more fiber than the ultra-processed foods produced by the food industry nowadays. The intake of starch and vegetable foods was inversely related to CHD mortality rates. These ecological results are in accord with the findings of randomized controlled trials of nutrients or foods on LDL-cholesterol, a major risk factor of CHD. Meta-analyses of controlled dietary experiments have shown that replacement of added sugars by starch, butter by soft margarine and addition of legumes to the diet reduced LDL-cholesterol [32–35].
Variations in habitual consumption of food groups in the SCS cohorts were summarized in the MAI for this ecological analysis. A high MAI score is an indicator of the traditional Mediterranean, as well as traditional Japanese dietary patterns and is associated with low CHD mortality rates. Though these ecological findings are not directly pertinent to inference about diet-disease relationships for individuals, they are nevertheless concordant with reported meta-analyses of individual diet data in cohort studies, which showed that other representations of the Mediterranean Diet were associated with lower individual CHD risk [36, 37]. Furthermore, they are concordant with the findings of the randomized PREDIMED diet trial, which showed that a Mediterranean-style diet, with either an additional amount of 50 ml extra-virgin olive oil daily, or 35 g/day of mixed tree nuts, compared with a lower fat diet, reduced the risk of cardiovascular events by 30% during 4.8 years of follow-up .
The structure of the analysis of dietary data used in the SCS is confined to the ecologic, rather than individual correlations. Only average population intake data were used for correlations. A disadvantage was that only ten cohorts had complete 50-year CHD mortality data; for the remaining six cohorts we relied upon strong dose-response relationships (r > 0.98) for predicted 50-year CHD mortality rates. However, the diet correlations of the 16 cohorts were lower than those of the ten cohorts and our description of lower correlations is conservative. Caution must be taken in interpreting relevance to the individual from ecologic differences and correlations. Even when viewed at the ecologic level, 16 is a small number of statistical units and leads to imprecise estimates and reduced power. Such results are prone to ecological biases such as confounding and ecological fallacy. An advantage of this study, however, is the original sampling of each cohort with the intention to use these dietary data to best characterize the cohorts. Additional advantages of this study are the accuracy of chemical analysis of the food, fatty acid and carbohydrate intake at entry, as well as the complete near lifetime mortality follow-up. These population-level data ignore variation within and between persons; consequently, reducing generalizability to the individual.
In conclusion, correlations with CHD rates varied widely across specific fatty acids and carbohydrates, consistent with the concept that neither total fat nor carbohydrate is a rational basis for choosing what foods to eat. Further, this ecological analysis of the SCS diets of the 1960s and the 50-year experience shows that cohorts of middle-aged populations of men consuming traditional diets of mainly plant foods and of a high MAI index have low long-term CHD death rates. Cohorts with a dietary pattern high in SFA and sucrose found in foods such as hard fat and sweet products are characterized by excess long-term population CHD death rates. This leads to the conclusion that in the 1960s a low population intake of SFA and sucrose was an indicator of overall dietary habits resembling the traditional Mediterranean and Japanese diets and that these patterns were associated with low long-term CHD mortality rates.
Den Hartog C, Buzina R, Fidanza F, Keys A, Roine P. Dietary studies and epidemiology of heart diseases. The Hague, The Netherlands: Voeding; 1968. p. 1–157. http://edepot.wur.nl/380320. Accessed 2 Nov 2017.
Keys A. Coronary heart disease in seven countries. Circulation. 1970;41(suppl 1):1–211. (Ed)
Keys A, Aravanis C, Blackburn H, Buzina R, Djordjevic BS, Dontas AS, et al. Seven Countries Study. A multivariate analysis of death and coronary heart disease. Cambridge, MA: Harvard Univ Press; 1980. p. 1–381.
Kromhout D, Menotti A, Blackburn H. Prevention of coronary heart disease. Diet, lifestyle and risk factors in the Seven Countries Study. Norwell, MA: Dordrecht, NL: Kluwer Publ; 2002. p. 1–267.
Keys A, Menotti A, Karvonen MJ, Aravanis C, Blackburn H, Buzina R, et al. The diet and 15-year death rate in the Seven Countries Study. Am J Epidemiol. 1986;124:903–915.
Menotti A, Keys A, Aravanis C, Blackburn H, Dontas A, Fidanza F, et al. Seven Countries Study. First 20-year mortality data in 12 cohorts of the Seven Countries. Ann Med. 1989;21:175–179.
Kromhout D, Menotti A, Bloemberg B, Aravanis C, Blackburn H, Buzina R, et al. Dietary saturated and trans fatty acids and cholesterol and 25-year mortality from coronary heart disease: the Seven Countries Study. Prev Med. 1995;24:308–315.
Skeaff CM, Miller J. Dietary fat and coronary heart disease: summary of evidence from prospective cohort studies and randomised controlled trials. Ann Nutr Metab. 2009;55:173–201.
Mozaffarian D, Micha R, Wallace S. Effects of coronary heart disease increasing polyunsaturated fat in place of saturated fat: a systematic review and meta-analysis of randomized controlled trials. PLoS Med. 2010;7:e1000252.
Ramsden CE, Hibbeln JR, Majchrzak SF, Davis JM. n-6 fatty acid-specific and mixed polyunsaturated dietary interventions have different effects on CHD risk: a meta-analysis of randomised controlled trials. Br J Nutr. 2010;104:1586–1600.
Hooper L, Summerbell CD, Thompson R, Sills D, Roberts FG, Moore HJ, et al. Reduced or modified dietary fat for preventing cardiovascular disease. Sao Paulo Med J. 2016;134:182–3.
Ramsden CE, Zamora D, Leelarthaepin B, Majchrzak-Hong SF, Faurot KR, et al. Use of dietary linoleic acid for secondary prevention of coronary heart disease and death: evaluation of recovered data from the Sidney Diet Heart Study an updated meta-analysis. BMJ. 2013;346:e8707.
Chowdhury R, Warnakula S, Kunutsor S, Crowe F, Ward HA, Johnson L, et al. Association of dietary, circulating, and supplement fatty acids with coronary risk: a systematic review and meta-analysis. Ann Intern Med. 2014;160:398–406.
Hooper L, Martin N, Abdelhamid A, Davy Smith G. Reduction in saturated fat intake for cardiovascular disease. The Cochrane database of systematic reviews. Cochrane Database Syst Rev. 2015;6:CD011737.
Howard BV, Van Horn L, Hsia J, Manson JE, Stefanick ML, Wassertheil-Smoller S, et al. Low-fat dietary pattern and risk of cardiovascular disease: the Women´s Health Initiative Randomized Controlled Dietary Modification Trial. JAMA. 2006;8:655–666.
Atkins RC. Dr. Atkins new diet revolution. New York: Harper Collins; 1992.
Te Morenga L, Mallard S, Mann J. Dietary sugars and body weight: systematic review and meta-analyses of randomised controlled trials and cohort studies. BMJ. 2012;345:e7492.
Malik VS, Pan A, Willett WC, Hu HB. Sugar-sweetened beverages and weight gain in children and adults: a systematic review and meta-analysis. Am J Clin Nutr. 2013;98:1084–1102.
Imamura F, O’Connor L, Ye Z, Mursu J, Hayashino Y, Bhupathiraju SN, Forouhi NG. Consumption of sugar sweetened beverages, artificially sweetened beverages, and fruit juice and incidence of type 2 diabetes: systematic review, meta-analysis, and estimation of population attributable fraction. BMJ. 2015;351:h3576.
Malik AH, Akram Y, Shetty S, Malik SS, Njike VY. Impact of sugar-sweetened beverages on blood pressure. Am J Cardiol. 2014;113:1574–1580.
Keys A, Keys M. Eat well and stay well. The Mediterranean Way. Garden City, NY: Doubleday & Co, Inc; 1975. p. 1–488.
Fidanza F. The Mediterranean Italian diet: keys to contemporary thinking. Proc Nutr Soc. 1991;50:519–526.
Kromhout D, Keys A, Aravanis C, Buzina R, Fidanza F, Giampaoli S, et al. Food consumption patterns in the 1960s in Seven Countries. Am J Clin Nutr. 1989;49:889–894.
Alberti-Fidanza A, Fidanza F. Mediterranean adequacy Index of Italian diets. Publ Hlth Nutr. 2004;7:937–941.
Fidanza F, Alberti A, Lanti M, Menotti A. Mediterranean Adequacy Index: correlation with 25-year mortality from coronary heart disease in the Seven Countries Study. Nutr Metab Cardiovasc Dis. 2004;14:254–258.
WHO International Classification of Diseases. Eighth revision. Geneva: World Health Organization; 1965.
Pett KD, Kahn J, Willett WC, Katz DL. Ancel Keys and the Seven Countries Study. White paper, 2017, http://www.truehealthinitiative.org/wordpress/wp-content/uploads/2017/07/SCS-White-Paper.THI.8-1-17.pdf. Accessed 2 Nov 2017.
Keys A, Anderson JT, Grande F. Serum cholesterol response to changes in diet. IV. Particular saturated fatty acids in the diet. Metabolism. 1965;14:776–787.
Sacks FM, Lichtenstein AH, Wu JHY, Appel LJ, Greager MA, Kris-Etherton PM et al. Dietary fats and cardiovascular disease. A presidential advisory from the American Heart Association. Circulation. 2017. https://doi.org/10.1161/CIR.0000000000000510.
Keys A. Sucrose in the diet and coronary heart disease. Atherosclerosis. 1971;14:193–202.
Guideline: sugars intake for adults and children. Geneva: World Health Organization; 2015, www.who.int/nutrition/publications/guidelines/sugars_intake/en
Te Morenga LA, Howatson AJ, Jones RM, Mann J. Dietary sugars and cardio-metabolic risk: systematic review and meta-analysis of randomized controlled trials on blood pressure and lipids. Am J Clin Nutr. 2014;100:65–79.
Fattore E, Botta F, Agostini C, Bosetti C. Effects of free sugars on blood pressure and lipids: a systematic review and meta-analysis of nutritional isoenergetic intervention trials. Am J Clin Nutr. 2017;105:42–56.
Zock PL, Katan MB. Butter, margarine and lipoproteins. Atherosclerosis. 1997;131:7–16.
Ha V, Sievenpiper JL, de Souza RJ, Jayalath VH, Mirrahimi A, Agarwal A, et al. Effect of dietary pulse intake on established therapeutic lipid targets for cardiovascular risk reduction: a systematic review and meta-analysis of randomized controlled trials. CMAJ. 2014;186:E252–E262.
Sofi F, Macchi C, Abbate R, Gensini GF, Casini A. Mediterranean diet and health status: an updated meta-analysis and a proposal for a literature-based adherence score. Publ Health Nutr. 2013;17:2769–2782.
Martinez-Gonzalez MA, Bes-Rastrollo M. Dietary patterns, Mediterranean diet, and cardiovascular disease. Curr Opin Lipidol. 2014;25:20–26.
Estruch R, Ros E, Salas-Salvadó J, Covas MI, Corella D, Arós F, et al. Primary prevention of cardiovascular disease with Mediterranean diet. N Engl J Med. 2013;368:1279–1290.
The authors express their gratitude for the role of the late professor Srecko Nedeljkovic from Belgrade, who was for many years the enthusiastic and highly professional principal investigator of the Serbian cohorts in the Seven Countries Study.
Drs. Menotti and Kromhout had full access to all data of the Seven Countries Study and take responsibility for the integrity of the data and the accuracy of data analysis. Study concept and design: the first generation of investigators lead by the late professor Ancel Keys of the University of Minnesota, Minneapolis conceptualized and designed the Seven Countries Study. Acquisition, analysis, or interpretation of data: all authors Drafting of the manuscript: Kromhout, Menotti, and Jacobs. Critical evaluation of the manuscript for important intellectual content: all authors. Chemical analysis: the first round of chemical analysis from 1959 to 1964 was coordinated by Keys and the second round from 1987 to 1988 by Kromhout. Statistical analysis: Menotti, Puddu, and Jacobs. Obtained funding: Nutrition Council, The Hague; Royal Netherlands Academy of Arts and Sciences, Amsterdam; University of Leiden; National Institute for Public Health and the Environment, Bilthoven; Wageningen University; University of Groningen, all in The Netherlands.
Conflict of interest
The authors declare that they have no conflict of interest.
About this article
Cite this article
Kromhout, D., Menotti, A., Alberti-Fidanza, A. et al. Comparative ecologic relationships of saturated fat, sucrose, food groups, and a Mediterranean food pattern score to 50-year coronary heart disease mortality rates among 16 cohorts of the Seven Countries Study. Eur J Clin Nutr 72, 1103–1110 (2018). https://doi.org/10.1038/s41430-018-0183-1
Effects of Mediterranean diet on hospital length of stay, medical expenses, and mortality in elderly, hospitalized patients: A 2-year observational study
Coronary heart disease mortality trends during 50 years as explained by risk factor changes: The European cohorts of the Seven Countries Study
European Journal of Preventive Cardiology (2020)
How the Seven Countries Study contributed to the launch and development of cardiovascular epidemiology in Italy. A historical perspective
Nutrition, Metabolism and Cardiovascular Diseases (2020)
Association of serum cholesterol with coronary heart disease mortality during 50-year follow-up in ten cohorts of the seven countries study
Nutrition, Metabolism and Cardiovascular Diseases (2020)
Energy-adjusted Dietary Inflammatory Index scores predict long-term cardiovascular disease mortality and other causes of death in an ecological analysis of the Seven Countries Study
European Journal of Preventive Cardiology (2020)