Comparison of diets of diabetic and non-diabetic elderly men in Finland, The Netherlands and Italy

Transcription

1 (2000) 54, 181±186 ß 2000 Macmillan Publishers Ltd All rights reserved 0954±3007/00 $ Comparison of diets of diabetic and non-diabetic elderly men in Finland, The Netherlands and Italy SM Virtanen 1 *, EJM Feskens 2,LRaÈsaÈnen 3, F Fidanza 4, J Tuomilehto 5, S Giampaoli 6, A Nissinen 7 and D Kromhout 2 1 School of Public Health and Medicine, and Tampere Diabetes Research Center, University of Tampere, and Department of Pediatrics, Tampere University Hospital, Tampere, Finland; 2 Department of Chronic Diseases and Environmental Epidemiology, National Institute of Public Health and the Environment, Bilthoven, The Netherlands; 3 Department of Applied Chemistry and Microbiology, Division of Nutrition, University of Helsinki, Helsinki, Finland; 4 Nutrition Section, Department of Internal Medicine and Endocrinological and Metabolic Sciences, University of Perugia, Italy; 5 Department of Epidemiology and Health Promotion, National Public Health Institute, Helsinki, Finland; 6 Laboratory of Epidemiology and Biostatistics, Istituto Superio di Sanita, Rome, Italy; and 7 Department of Community Health and General Practice, University of Kuopio, Kuopio, Finland Objective: To evaluate whether dietary recommendations for subjects with diabetes are met among Finnish, Dutch and Italian elderly men with diabetes, and whether the diets of diabetic and non-diabetic men differ in these three countries. Design: A dietary survey using cross-check dietary history method. A cross-sectional comparison. Setting: Thirty-year follow-up of survivors from the Finnish, Dutch and Italian cohorts of the Seven Countries Study. Subjects: 227 elderly men from Finland, 537 from The Netherlands, and 417 from Italy, of whom 8 ± 9% had diabetes. Main results: The diets of non-diabetic men from the three countries differed markedly from each other. In all three countries diabetic men consumed less added sugar than non-diabetic men. In Italy, in addition, diabetic men consumed more fruits and berries and vegetables. The Dutch diabetic men ate relatively more cereal products, fruits and berries, milk and milk products, cheese, and meat and meat products and drank less alcoholic beverages than non-diabetic men. The diet of both diabetic and non-diabetic Finnish and Dutch men was characterized by high fat content (41% and 40% of energy, respectively). The fat content of the diet was even higher for diabetic than non-diabetic men in Finland and The Netherlands, but not in Italy. The bre content of the diet was the highest among Dutch men and diabetic men received more dietary bre than non-diabetic men in The Netherlands and Italy, but not in Finland. The diet of diabetic and non-diabetic Finnish men differed little from each other and was characterized by high nutrient density of several vitamins and minerals. The proportion of protein of energy intake was higher among diabetic than non-diabetic Dutch and Italian men. Conclusions: The diet of the diabetic men from Finland, the Netherlands, and Italy resembled more the diet of non-diabetic men from the respective countries than the diet of diabetic men from the other countries. In the diet of Italian diabetic men, the proportions of fat, saturated fatty acids and carbohydrates were nearest the recommended levels. Sponsorship: The National Institute on Aging, Bethesda, USA, the Dutch Prevention Foundation, the Hague, The Netherlands, the Academy of Finland, and the Sandoz Gerontological Foundation. Descriptors: food consumption; nutrient intake; diabetic men; Finland; The Netherlands; Italy (2000) 54, 181±186 Introduction Dietary recommendations for subjects with diabetes are nowadays uniform in Western countries and do not substantially differ from what is recommended for the general population (Food and Nutrition Board, National Research Council, 1989; American Diabetes Association, 1991; Diabetes and Nutrition Study Group of the EASD, 1995). Poor dietary compliance regarding intakes of saturated fat, dietary cholesterol and dietary bre was observed among subjects with insulin-dependent diabetes mellitus (IDDM) aged 15 ± 60 y from several European countries *Correspondence: SM Virtanen, School of Public Health, University of Tampere, PO Box 607, FIN Tampere, Finland. suvi.virtanen@uta. Received 13 November 1998; revised 6 September 1999; accepted 9 September 1999 (Toeller et al, 1996). Very little is known about how dietary recommendations are being adhered among elderly subjects with diabetes in different European countries and how the dietary habits of the non-diabetic population affect dietary compliance. The most important aim of dietary recommendations among elderly subjects with diabetes is secondary and tertiary prevention of diabetic complications, especially cardiovascular diseases. The present study aims at evaluating whether dietary recommendations for diabetes subjects are being adhered among elderly men in three geographically and culturally different countries and whether the diet of diabetic and nondiabetic men differs from each other in these populationbased subject series. The advantage of the study is that diet was studied in connection of a survey directed towards elderly men in the general population, not speci cally towards men with diabetes.

2 182 Subjects and methods Study populations The Seven Countries Study was an observational study originally designed for the investigation of cardiovascular diseases and risk factors. The study started in the period 1958 ± 1964 in 16 cohorts of middle-aged men born between 1900 and 1919 in seven different countries (Keys et al, 1967). The survivors of the cohorts in Finland, The Netherlands, and Italy were re-examined after 30 y. In Finland, the original cohorts in 1959 consisted of men from two geographically de ned areas, Ilomantsi in East Finland (n ˆ 823) and PoÈytyaÈ and MellilaÈ in West Finland (n ˆ 888). In 1989, at the time of the 30 y follow-up the subjects were 70 ± 89 y old. Altogether 524 men of the original cohort of 1711 men were still alive (Nissinen et al, 1993). Of these, 470 (90%) were examined, and a random sample of 227 of these men participated in the dietary survey (RaÈsaÈnen et al, 1992). The Dutch cohort comes from the town of Zutphen, in the eastern part of The Netherlands. Of this cohort, 878 men were examined for the rst time in In 1985, 555 of the men in this cohort were still alive and in addition, a random sample was selected (n ˆ 711) of all other men aged 65 ± 85 y who were living in Zutphen. Of them 939 (74%) took part in the study in In 1990, 560 (78%) of those 718 still alive from the Dutch cohort, took part in the study. Complete information on the diet was obtained from 537 men. The Italian cohorts come from two rural villages, Crevalcore from northern Italy and Montegiorgio from central Italy (Keys et al, 1967). The original cohorts in 1960 consisted of 1712 men. In 1991, 614 of the men were still alive, 427 were examined, and 417 of them (68%) took part in the dietary study. Complete data on diet from the 30 y follow-up examination were available from 227 Finnish, 537 Dutch, and 417 Italian men and were used in the present analyses. Examinations All men were examined according to the international protocol used in the previous surveys of the Seven Countries Study (Keys et al, 1967). Medical examinations were carried out by trained physicians. Height and weight were measured in light clothing without shoes. Body mass index (BMI) was calculated by dividing weight by height squared (kg=m 2 ). Standardized questions on the history of diabetes mellitus were asked. The diabetic men were also asked about treatment of diabetes, ie the use of oral anti-diabetic treatment and=or insulin and whether they followed any dietary instructions for subjects with diabetes. In Finland and in Italy the standardized questionnaire of the Seven Countries Study core protocol was used to obtain information on smoking habits (Keys et al, 1967). In The Netherlands information on smoking habits was collected using a locally developed questionnaire. The presence of diabetes was diagnosed at the examination by a positive history. Men who were not previously known to have diabetes but had asymptomatic hyperglycaemia were considered as nondiabetic in the present analysis. Food and beverage consumption data were collected by the cross-check dietary history method (Burke, 1947; Block, 1982; Bloemberg et al, 1989). The interviews were carried out by experienced dietitians and nutritionists in all cohorts. In Finland, the interviews took place in autumn 1989, in The Netherlands in spring 1990, and in Italy in autumn Information was obtained about the usual food consumption pattern during the previous month preceding the interview. First, the usual food consumption pattern of a person was assessed during week and weekend days. This part concerned questions about the foods used at breakfast, lunch, dinner and between meals. Thereafter a checklist with an extensive number of foods was used to record the frequencies and amounts of foods consumed. The information about the food pattern was checked with the information from the checklist. Portion sizes were estimated by a picture booklet with 126 color photographs of various portion sizes of foods in Finland, by a portable scale in The Netherlands, and by arti cial models of different foods in Italy. Although the dietary history method was adjusted to the local situation, the methodology was comparable across the cohorts (Bloemberg et al, 1989; Farchi et al, 1995; RaÈsaÈnen et al, 1992). Nutrient intake was assessed using computerized versions of the local food tables for Finland, The Netherlands and Italy (AhlstroÈm et al, 1972; Netherlands Of cial Dutch Food Table (NEVO), 1989; Fidanza & Versiglioni, 1989). The Finnish, Dutch and Italian nutrient databases were complete for the nutrients studied and the Finnish and Dutch ones were complete also for foods. For the very few foods not included in the Italian databases, values of similar foods were used. The computing programmes contained data on recipes of all composite dishes. In Finland and in The Netherlands dietary bre was measured by the method of Englyst (1981). In Italy both the method of Englyst and that of Southgate (1981) had been used. The Southgate method is known to overestimate dietary bre, because of incomplete starch solubilization (Schneeman & Tietyen, 1994). This means that there may be some overestimation in the dietary bre intake of Italian men. Carbohydrates were digestible carbohydrates and they were calculated by subtracting water, ash, protein and fat, the rest being carbohydrates. Lactose was included in mono- and disaccharides. Apart from nutrients all the food groups and items which were comparable between the three countries were used in the analyses: cereals, potatoes, vegetables, legumes, fruits and berries, milk and milk products, cheese, eggs, dietary fats and oils, meat and meat products, sh and sh products, alcoholic beverages, and added sugar. Added sugar included all sucrose consumed as such, for example, in coffee, and that added to dishes and baked products. Statistical analyses Statistical analyses were carried out using the SAS programs version The food consumption and nutrient intake were expressed per 1000 kcal, to adjust for energy intake. Mean values of selected characteristics were compared between diabetic and non-diabetic men in each country and between countries among diabetic and nondiabetic men. Analysis of covariance was used to adjust for age as a potential confounder. For skewed variables logarithmic transformations were used. All P-values were based on two-sided tests of statistical signi cance. Results Characteristics of study populations The proportion of diabetic men was similar in the Finnish, Dutch and Italian cohorts (9%, 8% and 9%, respectively). All the 21 Finnish diabetic men reported adhering to

3 diabetic diet, 10 of them had also oral anti-diabetic treatment and one had insulin. In the Dutch cohort, 34 of the 43 diabetic men reported adherence to the diet therapy, of them 19 also had oral anti-diabetic treatment and eight also had insulin, whereas nine men had only oral anti-diabetic treatment without any dietary prescription. Of the 36 diabetic men, 31 reported having received dietary instructions, of them 18 also used oral anti-diabetic treatment and one had insulin. Four men used only oral anti-diabetic treatment and one man insulin alone without any dietary prescription. The study groups were similar in age, only Italian nondiabetic men were somewhat older than non-diabetic men from other countries (P < 0.001, Table 1). The diabetic and non-diabetic men from the three countries were not different regarding cardiovascular risk factors (Table 1). Dutch men had lower BMI (P < 0.01) and lower systolic blood pressure (P < 0.001) than men from Finland and Italy. Italian men had a higher serum HDL cholesterol concentration than Finnish or Dutch men, respectively (P < 0.001). Food consumption The food consumption patterns of Finnish, Dutch, and Italian men differed considerably from each other. In the diet of Finnish non-diabetic men, the percentage of energy from milk and milk products, eggs, sh and sh products, and sugar was highest and that from cereal products, vegetables, legumes, fruits and berries, cheese, meat and meat products, and alcoholic beverages lowest (P < 0.01, Table 2). Dutch non-diabetic men consumed relatively most potatoes and least sh and sh products (P < 0.01), whereas the relative consumption of cereal products, legumes, fruit and berries, and alcoholic beverages was highest and that of potatoes, milk and milk products, eggs, and sugar lowest for the Italian men (P < 0.01). In Finland the only difference in food consumption patterns between diabetic and non-diabetic men was seen in the consumption of added sugar which was about onehalf in diabetic compared with non-diabetic men (Table 2). The Dutch and Italian diabetic men consumed very little added sugar, only about 1 g=day, much less than nondiabetic men in the same areas. The consumption of vegetables and fruits and berries was higher for Italian diabetic than non-diabetic men. In The Netherlands, the consumption of cereal products, fruits and berries, milk and milk products, cheese, and meat and meat products was greater for diabetic men and that of alcoholic beverages greater for non-diabetic men. Differences in food consumption patterns between diabetic men from the three countries were similar to differences observed between non-diabetic men. Among non-diabetic men, energy intake was highest for the Finnish and lowest for the Dutch men (P < 0.001, Table 183 Table 1 Characteristics of diabetic (D) and non-diabetic (ND) men from Finland, The Netherlands and Italy Finland The Netherlands Italy D ND D ND D ND n Age (y) (mean s.d.) Body mass index (kg=m 2 ) (mean s.d.) Systolic blood pressure (Hgmm) (mean s.d.) Diastolic blood pressure (Hgmm) (mean s.d.) Serum total cholesterol (mmol=l) (mean s.d.) Serum HDL cholesterol (mmol=l) (mean s.d.) Proportion of smokers (%) Prevalence of hypertension (%) Prevalence of myocardial infarction (%) Table 2 and Italy Mean ( s.d.) food consumption (g=1000 kcal) by diabetic (D) and non-diabetic (ND) men from Finland, The Netherlands Finland The Netherlands Italy Differences between countries a D ND D ND D ND D ND n Cereal products * b a, b, c A, B, C Potatoes a, b, c A, B, C Vegetables ** a, b A, B Legumes A, B, C Fruit and berries * *** b A, B, C Milk and milk products ** a, b, c A, B, C Cheese ** a, b A, B Eggs b, c A, B, C Dietary fats and oils b B, C Meat and meat products *** a, b A, B Fish and sh products c a, c A, B, C Alcoholic beverages * b, c A, B, C Added sugar *** *** *** a, b A, B, C a Differences between countries indicated by small letters in diabetic men and by capital letters in non-diabetic men: a, A, Finish men differ from Dutch men; b, B, Finnish men differ from Italian men; c, C, Dutch men differ from Italian men. b Differences *P < 0.05, **P < 0.01, ***P < between diabetic and non-diabetic men in each country. c Log-transformed in analyses, the values are retransformed.

4 184 3). The Finnish diet was characterized by the highest content of saturated fatty acids, mono- and disaccharide, dietary cholesterol, thiamin, ribo avin, calcium, phosphorus, and potassium, and the lowest content of monounsaturated fatty acids and alcohol (P < 0.001). The intake of protein per kg of body weight was highest for the Finnish non-diabetic men (P < 0.001). The diet of Dutch non-diabetic men contained the greatest amount of dietary bre, polyunsaturated fatty acids, and vitamin C and the least carbohydrates, especially polysaccharides (P < 0.001). In the diet of Italian non-diabetic men the relative amounts of monounsaturated fatty acids, polysaccharides, and alcohol were highest and those of protein, fat, especially saturated and polyunsaturated fatty acids, monoand disaccharide, dietary cholesterol, dietary bre, thiamin, ribo avin, vitamin C, phosphorus, potassium and iron lowest (P < 0.01). In Finland, the energy intake of diabetic and nondiabetic men did not differ, whereas Dutch and Italian diabetic men had lower intakes of energy than non-diabetic men, respectively (Table 3). On the whole, the differences between diabetic and non-diabetic men in nutrient intakes were greatest in The Netherlands and smallest in Finland (Table 3). The proportion of protein was higher by diabetic than non-diabetic men in The Netherlands and Italy. In the diet of both Finnish and Dutch men, the proportion of fat of energy intake was higher for diabetic than non-diabetic men and consequently the proportion of carbohydrates lower in the diet of diabetic men. The proportion of saturated fatty acids was similar among diabetic and nondiabetic men in all the three countries. The lower proportion of carbohydrates in the diet of Finnish and Dutch diabetic men compared with non-diabetic men was due to the higher intake of mono- and disaccharide by nondiabetic men. Dutch diabetic men had higher intake of polysaccharides than non-diabetic men. The intake of alcohol was lower by Dutch diabetic mean and that of dietary cholesterol lower by non-diabetic men. The nutrient densities of bre and several vitamins and minerals were higher for Dutch and Italian diabetic men as compared with non-diabetic men (Table 3). Among diabetic men, the differences between the countries in the intake of energy and nutrients were similar to those observed among nondiabetic men. In the relative intakes of vitamins and minerals the differences between diabetic cohorts were somewhat smaller than between non-diabetic cohorts. Discussion The differences in food consumption patterns of Finnish, Dutch and Italian cohorts of the Seven Countries Study have become fewer and smaller since the 1960s (Kromhout et al, 1989; Huijbregts et al, 1995). However, in 1989 ± 1991 marked differences both in absolute and energyadjusted food consumption were still between the three cohorts of men. The diet of Finnish men could be regarded as North European and that of Italian men Mediterranean, whereas the diet of Dutch men is positioned between these two cohorts (Huijbregts et al, 1995). The differences in the composition of diet between the diabetic patients re ected the differences observed between non-diabetics. In all the three cohorts, diabetic men consumed less added sugar than non-diabetic men. In Finland no other differences were seen in food consumption between diabetic and non-diabetic men. In Italy, in addition, diabetic men consumed more fruits and berries and vegetables. The Dutch diabetic men had the greatest differences in their diet compared with non-diabetic men Ð they ate relatively more cereal products, fruits and berries, milk and milk products, cheese, and meat and meat products and fewer alcoholic beverages. In the Dutch study data on some food subgroups were also available. Dutch diabetic men ate more bread with high bre content, low-fat meat products, low-fat or fat-free Table 3 Mean ( s.d.) intakes of energy and nutrients by diabetic (D) and non-diabetic (ND) men from Finland, The Netherlands and Italy Finland The Netherlands Italy Differences between countries a D ND D ND D ND D ND n Energy (kcal) * b *** a, b A, B, C Protein (percentage of energy) *** *** a, c A, B, C Protein (g=kg of body weight) ** b A, B, C Fat (percentage of energy) ** * b, c A, B, C saturated a, b, c A, B, C monounsaturated *** * b, c A, B, C polyunsaturated ** b, c A, B, C Carbohydrate (percentage ** *** b, c A, C of energy) mono- and disaccharide ** *** a, b, c A, B, C polysaccharides ** b, c A, B, C Alcohol (percentage of energy) * b, c A, B, C Dietary cholesterol (mg=1000 kcal) ** a, b, c A, B, C Dietary bre (g=1000 kcal) *** *** a, c A, B, C Thiamin (mg=1000 kcal) *** *** a, b, c A, B, C Ribo avin (mg=1000 kcal) *** ** a, b, c A, B, C Vitamin C (mg=1000 kcal) ** ** a, c A, B, C Calcium (mg=1000 kcal) *** * b, c A, B, C Phosphorus (mg=1000 kcal) *** * b, c A, B, C Potassium (mg=1000 kcal) *** *** b, c A, B, C Iron (mg=1000 kcal) *** c B, C a Differences between countries indicated by small letters in diabetic men and by capital letters in non-diabetic men: a, A, Finish men differ from Dutch men; b, B, Finnish men differ from Italian men; c, C, Dutch men differ from Italian men. b Differences *P < 0.05, **P < 0.01, ***P < between diabetic and non-diabetic men in each country.

5 milk and milk products, and both low-fat and high-fat cheese than non-diabetic men (data not shown). At the time of the study, around 1990, dietary recommendations directed towards subjects with diabetes were similar in Finland, The Netherlands, and Italy (Diabetes and Nutrition Study Group of the EASD, 1988). For obese subjects with non-insulin dependent diabetes mellitus (NIDDM) energy restriction is recommended. Total fat should provide about 30% of total energy intake, saturated fatty acids less than 10%, and carbohydrates 50 ± 60%. The recommended amount of dietary bre is about 20 g per 1000 kcal. The intake of dietary cholesterol should not exceed 300 mg per day and that of added sugar 50 g per day (30 g of added sucrose). In the earlier dietary recommendations towards subjects with diabetes the amount of sucrose was more strictly controlled (Huttunen et al, 1982). In the most recent dietary recommendations for subjects with diabetes, bene cial effects of monounsaturated fatty acids on glucose and lipid metabolism have been recognized (Diabetes and Nutrition Study Group of the EASD, 1995). No quantitative recommendations for total fat and carbohydrates are given anymore, instead, it is recommended that most dietary energy intake should come from combination of carbohydrates and monounsaturated fatty acids with cis-con guration. The intake of saturated and polyunsaturated fatty acids each should not exceed 10% of energy intake. The greater energy intake of Finnish men compared with Dutch and Italian men can be partly explained by higher amount of physical activity by Finnish men. A part of it may re ect methodological problems in assessing food consumption (Huijbregts et al, 1995). For the sake of comparability we used energy-adjusted food and nutrient intakes in the present study. The differences in energy intake between the countries were greater among diabetic than non-diabetic men. The energy intake of Finnish diabetic men were about 700 and 600 kcal higher than that of Italian and Dutch counterparts, whereas among non-diabetic men the differences were about 600 and 300 kcal, respectively. The lower energy intake observed among diabetic compared with non-diabetic men in The Netherlands and Italy may re ect their smaller amount of physical activity and=or underreporting of food consumption. Although according to BMI the diabetic men in the present study were not more obese than non-diabetic men, more dietary advice to control body weight may have been directed towards diabetic than nondiabetic men which may have resulted in underreporting of food consumption among diabetic men. It is also known that body weight can decrease in diabetic patients with increasing severity of the disease. The mean proportion of total fat and saturated fatty acids in the diet was well below the recommended levels for the Italian diabetic men only, whereas Finnish and Dutch diabetic men remarkably exceeded the recommended levels. Similar differences have also been reported among younger subjects with IDDM from Italy, Finland and The Netherlands (Toeller et al, 1996). In the present study the total fat intake was even higher for diabetic than nondiabetic men in Finland and The Netherlands probably due to avoidance of carbohydrates. Finnish diabetic men received also much more dietary cholesterol than recommended. The dietary bre intake of all the diabetic cohorts was below the recommended level, although Dutch and Italian diabetic men had somewhat increased their intake of dietary bre, especially from fruit and berries compared with non-diabetic men. The nutrient density of the diet was generally higher for the Finnish men than for the Dutch and Italian men, and among Dutch and Italian men higher for diabetic than non-diabetic men. The mean daily intake of thiamin by Italian diabetic and non-diabetic men, of ribo avin by Italian non-diabetic men, and of calcium by Italian diabetic and non-diabetic men were lower by 10% or more than Recommended Dietary Allowances (Food and Nutrition Board, National Research Council, 1989). Concerning the intakes of total fat and carbohydrates the differences between the countries were greater among diabetic men than among non-diabetic men. On the other hand, differences in the intake of mono- and disaccharide and in the intakes of many vitamins and minerals between the countries were smaller among diabetic as compared with non-diabetic men. Diabetic men from all the three countries had unnecessarily high intakes of protein, and in The Netherlands and Italy even higher intake than nondiabetic men (recommendation 0.75 g per kg of body weight per day, Food and Nutrition Board, National Research Council, 1989). Our results in Finnish and Dutch diabetic men resemble those obtained in US women with NIDDM and in German men and women with NIDDM in the high proportion of fat, especially saturated fatty acids, and low proportion of carbohydrates in the diet (Shimakawa et al, 1993; Toeller, 1993). In a recent study, a higher proportion of protein (20% of energy intake) and a lower proportion of carbohydrate (39%) were observed in Spanish men than among the present Finnish, Dutch and Italian men (Diabetes and Nutrition Study Group of the Spanish Diabetes Association, 1997). The fat intake (36% of energy intake) was somewhat higher and that of alcohol lower (1%) in that study than among Italian men in the present study. To conclude, the diet of the diabetic men from Finland, Italy, and the Netherlands resembled more the diet of nondiabetic men from respective countries than the diet of diabetic men from the other countries. In the diet of Italian diabetic men, the proportion of fat, saturated fatty acids, and carbohydrates was nearest the recommended levels. Acknowledgements ÐWe are indebted to the many people involved in this longitudinal study, including all participants and the eldwork teams in East and West Finland, Italy, and in Zutphen, The Netherlands. We thank Ms Y Volebregt for the contribution to the data analysis. During 1993 ± 1994 Suvi M Virtanen was a postdoctoral researcher at the Department of Epidemiology and Public Health, Wageningen Agricultural University, The Netherlands, and at the Department of Chronic Diseases and Environmental Epidemiology, National Institute of Public Health and the Environment, Bilthoven, The Netherlands. References AhlstroÈm A, RaÈsaÈnen L & Kuvaja K (1972): A method of data processing for food-consumption surveys. Ann. Acad. Sci. Fenn. AIV 194, 1± 8. American Diabetes Association (1991): Nutritional recommendations and principles for individuals with diabetes mellitus. Diabetes Care 14, 20 ± 27. Block G (1982): A review of validations of dietary assessment methods. Am. J. Clin. Nutr. 115, 492 ± 505. Bloemberg BPM, Kromhout D, Obermann-de Boer GL et al (1989): The reproducibility of dietary intake data assessed with the cross-check dietary history method. Am. J. Epidemiol. 130, 1047 ± Burke BS (1947): The dietary history as a tool in research. J. Am. Diet. Assoc. 23, 1041 ±

6 186 Diabetes and Nutrition Study Group of the EASD (1995): Recommendations for the nutritional management of patients with diabetes mellitus. Diabetes Nutr. Metab. 8, 1±4. Diabetes and Nutrition Study Group of the EASD (1988): Nutritional recommendations for individuals with diabetes mellitus. Diabetes Nutr. Metab. 1, 145 ± 149. Diabetes and Nutrition Study Group of the Spanish Diabetes Association (1997): Diabetes nutrition and complications trial (DNCT): food intake and targets of diabetes treatment in a sample of Spanish people with diabetes. Diabetes Care 20, 1078 ± Englyst H (1981): Determination of carbohydrate and its composition in plant materials. In: The Analysis of Dietary Fibre in Food, ed. WPT James & O Theander, pp 71 ± 93. New York: Marcel Dekker. Farchi G, Fidanza F, Grossi P, Lancia A, Mariotti S & Menotti A (1995): Relationship between eating patterns meeting recommendations and subsequent mortality in 20 years. Eur. J. Clin. Nutr. 49, 408 ± 419. Fidanza F & Versiglioni N (1989): Tabelle di composizione degli alimenti. Napoli: Idelson. Food and Nutrition Board, National Research Council (1989): Recommended Dietary Allowances, 10th edn. Washington, DC: National Academy Press. Huijbregts PPCW, Feskens EJM, RaÈsaÈnen L, Alberti-Fidanza A, Mutanen M, Fidanza F & Kromhout D (1995): Dietary intake in ve ageing cohorts of men in Finland, Italy and the Netherlands. Eur. J. Clin. Nutr. 49, 852 ± 860. Huttunen JK, Aro A, Pelkonen R, Puomio M, Siltanen I & AÊ kerblom HK (1982): Dietary therapy in diabetes mellitus. Description of a recommendation prepared by the Finnish Diabetes Association's Committee on Nutrition Therapy. Acta Med. Scand. 211, 469 ± 475. Keys A, Aravanis C, Blackburn H et al (1967): Epidemiological studies related to coronary heart disease; characteristics of men aged 40 ± 59 in seven countries. Acta Med. Scand. (Suppl), 1 ± 392. Kromhout D, Keys A, Aravanis C et al (1989): Food consumption patterns in the 1960s in seven countries. Am. J. Clin. Nutr. 49, 889 ± 894. Netherlands Of cial Dutch Food Table (NEVO) (1989): Nederlands voedingsstoffenbestand 1989=90. Zeist, Stichting NEVO. Nissinen A, Tervahauta M, Pekkanen J et al (1993): Prevalence and change of cardiovascular risk factors among men born 1900 ± 19. The Finnish cohorts of the seven countries study. Age Ageing 22, 365 ± 376. RaÈsaÈnen L, Mutanen M, Pekkanen J et al (1992): Dietary intakes of 70 to 89 years old men in eastern and western Finland. J. Intern. Med. 232, 305 ± 312. Schneeman BO & Tietyen J (1994): Dietary bre. In: Modern Nutrition in Health and Disease, ed. ME Shils, JA Olson & M Shike, pp 89 ± 100. Philadelphia, PA: Williams and Wilkins. Shimakawa T, Herrera-Acena MG, Colditz GA, Manson JE, Stamper MJ & Willett WC (1993): Comparison of diets of diabetic and nondiabetic women. Diabetes Care 10, 1356 ± Southgate DAT (1981): Use of the Southgate method for unavailable carbohydrates in the measurement of dietary bre. In: The Analysis of Dietary Fibre in Food, ed. WPT James & O Theander. New York: Marcel Dekker. Toeller M (1993): Diet and diabetes. Diabetes Metab. Rev. 9, 93 ± 108. Toeller M, Klischan A, Heitkamp G et al (1996): Nutritional intake of 2868 IDDM patients from 30 centres in Europe. Diabetologia 39, 929 ± 939.