In uence of educational level and marital status on dietary intake, obesity and other cardiovascular risk factors in a Hong Kong Chinese population

Transcription

1 European Journal of Clinical Nutrition (1999) 53, 461±467 ß 1999 Stockton Press. All rights reserved 0954±3007/99 $ In uence of educational level and marital status on dietary intake, obesity and other cardiovascular risk factors in a Hong Kong Chinese population J Woo* 1, SSF Leung 2,SCHo 3, A Sham 3, TH Lam 4 and ED Janus 5 1 Department of Medicine and Therapeutics, The Chinese University of Hong Kong; 2 Department of Paediatrics, The Chinese University of Hong Kong; 3 Department of Community and Family Medicine, The Chinese University of Hong Kong; 4 Department of Community Medicine, University of Hong Kong Objectives: To examine the in uence of education and marital status on dietary intake, body mass index, waist hip ratio, blood pressure, fasting and 2 h glucose, and lipid pro le in adult Hong Kong Chinese. Design: Randomized age and sex strati ed survery Subjects: One thousand and ten subjects aged 25±74 y (500 men, 510 women) recruited for the 1995±96 Hong Kong Dietary and Cardiovascular Risk Prevalence Survey. Measurements: Dietary intake was estimated using a food frequency method. Information on education level and marital status was included in the questionnaire. Anthropometry and biochemical parameters were measured using standard methods. Results: After adjustment for age, higher levels of education are associated with higher percentage protein intake in men, higher percentage fat intake in women, higher nutrient density of bre and calcium in both men and women, and higher nutrient density of protein, fat, niacin, vitamin D, and polyunsaturated fatty acid in women. Consumption of fruits was also higher in women, and that of dairy products higher in men. Body mass index and waist-hip ratio were lower with increasing levels of education in women, while lower mean systolic BP was observed in men. Single women had lower nutrient densities of vitamin D and iron, and lower consumption of vegetables and sh, compared with married women. Body mass index was lower in both single men and women. Single men had a better cardiovascular risk factor pro le, in that diastolic BP, triglycerides and cholesterol=hdl ratio were lower, in addition to a lower body mass index. Conclusion: Higher education level is associated with a healthier diet and lower prevalence of overweight. Sponsorship: Hong Kong Health Services Research Grant. Descriptors: education; marital status; dietary intake; obesity; cardiovascular risk factors; Chinese *Correspondence: Dr J Woo, Department of Medicine & Therapeutics, Prince of Wales Hospital, Shatin, N.T., Hong Kong. Contributions by authors: (J Woo)ÐWriting, planning analysis, design of study, grant application, running of the project. (SSF Leung)ÐDesign of questionnaire, training of eld workers, day to day supervision, data analysis. (SC Ho)ÐDesign of questionnaire, data analysis. (A Sham)Ð Data analysis. Received 23 June 1998; revised 13 January 1999; accepted 25 January 1999 Introduction Dietary surveys of food intake, prevalence of obesity and other cardiovascular risk factors among many populations have identi ed the in uence of socioeconomic factors (Dobson et al, 1997; Hulshof et al, 1991; Kant et al, 1991; Rose & Marmot, 1981; Shea et al, 1993; Ko et al, 1997; Popkin et al, 1993a, 1993b; Popkin et al, 1995; To & Donnan, 1988). Among Western populations, lower social class appears to be associated with a more atherogenic dietary pattern and an increase in prevalence of obesity and other cardiovascular risk factors (Dobson et al, 1997; Hulshof et al, 1991; Kant et al, 1991; Rose & Marmot, 1981; Shea et al, 1993). Among Chinese populations, the trend appears to be reversed, in that higher income groups have a higher prevalence of obesity and other cardiovascular risk factors, together with a more atherogenic dietary pattern (Popkin et al, 1993a, 1993b; Popkin et al, 1995; To & Donnan, 1988). A recent survey of a selected cohort undergoing oral glucose tolerance test suggested that the trend in some Chinese communities may have now changed to the Western pattern (Ko et al, 1997). In Hong Kong, cardiovascular disease is the second highest cause of mortality after neoplasm, and unlike many Western countries where substantial declines of age standardised ischaemic heart disease mortality in the past two decades have been observed, the mortality in Hong Kong showed a plateau or slowly decreasing trend only (Yu et al, 1995). In view of the association between dietary habits and cardiovascular disease, it is important to document the dietary intake of the Hong Kong Chinese population. Since results of previous surveys suggest that dietary habits and cardiovascular risk factors may be in uenced by social factors, data from the 1995 Hong Kong Dietary and Cardiovascular Risk Factor Prevalence study (Leung et al, 1997; Janus, 1997) were examined, to identify the more vulnerable section of the community in order that preventive action may be targeted towards that group. In this paper, the in uence of education and marital status on nutrient intake, consumption pattern of food groups, body

2 462 mass index, waist hip ratio, blood pressure and other laboratory parameters are examined. Subjects and method The Hong Kong Dietary Survey was conducted from October 1995±May 1996 as part of a territory wide cardiovascular risk factor study in ethnic Chinese. Subjects were contacted by a random telephone survey, and invited to a hospital based centre for physical examination and blood tests. Dietary assessment was carried out consecutively on those who attend until more than 100 subjects in each 10 y age and sex groups from 34 y to 55 y were recruited. The response rate for telephone interview was approximately 80%, and of those approximately 40% attended for blood tests and anthropometric measurements. The attendees closely matched the overall Hong Kong population, since there was no difference in age distribution or socioeconomic characteristics between subjects attending for laboratory tests, those who participated in the telephone interview, and the population as a whole as described in the 1996 Hong Kong by-census. The age distribution was also similar between the three geographical regions of Hong Kong. A detailed description of the sample for the territorywide cardiovascular risk factor study has been published elsewhere (Janus, 1997). Over the period of the dietary survey, the dietary assessment was carried out consecutively on those who attended until more than 100 subjects in each 10 y age and sex group from 34 y to 55 y were recruited. No signi cant differences in physical or laboratory parameters were noted between subjects from the three geographical regions (Hong Kong Island, Kowloon, and New Territories) and between those who received the dietary assessment and those who did not. The overall cardiovascular risk factor study included 2900 attendees aged 25±74 y of whom 1010 underwent dietary assessment. The mean age of the study sample was 45.6 y (s.d. 11.7), range 24±74 y. Dietary assessment was carried out using a food frequency questionnaire, the validity of which had been examined elsewhere (Woo et al, 1997). This consisted of items in the following seven categories; Bread=pasta=rice (16 items); vegetables (63 items); fruits (26 items); meat (39 items)= sh (31 items)=eggs (5 items); beverages (37 items); dimsum=snacks (39 items); soups (10 items); and oil=salt=sauces. Wherever possible, subjects were told prior to the visit that a survey on a week's diet would be carried out and were advised to make a brief record at home to help the interview. On the day of the interview, each subject was asked to complete the questionnaire: the food item, the size of each portion, the frequency of consumption on a daily and weekly basis. Portion size was explained to subjects using a catalogue of pictures of individual food portions developed by one of the authors (SSFL). Data were cross checked by examining the dietary pattern (for example, if meals were skipped) to see if it corresponded to the number of times staple foods such as rice or noodles were consumed over a one week period. In case of discrepancies, the questionnaire was rechecked with the subject. The amount of cooking oil was estimated according to the method of preparing different foods: 0.2 tablespoon for steaming sh or stir frying half a portion of vegetables, and one tablespoon for stir frying one portion of vegetables or one portion of meat. The type of oil used was also documented to allow estimation of the quantity of fat used in cooking. Quantitation of nutrients was carried out using food tables for Hong Kong compiled from McCance and Widdowson (Holland et al, 1992) and two food tables used in China published by Zhongshan University (1991) (Tsang & Yung, 1991) and the Institute of Health of the Chinese Medical Science Institute (Institute of Health, 1992). Information regarding educational level and marital status was also obtained. For analysis, those who were divorced, separated or widowed were classi ed as being single. Physical measurements Height was measured to the nearest 0.5 cm and weight to the nearest 0.1 kg. Participants were measured in socks, stockings or bare feet and light street clothing. Waist and hip circumferences were measured twice to the nearest 0.5 cm and the mean was used for subsequent analysis. Waist circumference was measured half way between the xiphisternum and the umbilicus; hip circumference was measured at the level of the greater trochanters. Two consecutive measurements of blood pressure were taken on the right arm 2±3 min apart with the subject being seated for at least 10 min. Systolic blood pressure was rst checked by palpation prior to the auscultatory measurements. Systolic and diastolic phase V pressures were recorded to the nearest 2 mmhg. Normal mercury sphygmomanometers (Baumanometer, W.A. Baum Co., Inc. Copiague, New York 11726, USA) were used with standard sized cuffs (12±14 cm wide). If the two systolic or diastolic blood pressures recorded were four or more mm apart, a third measurement was made. The average of the two blood pressures (both entered into the data base) was used for the subsequent analysis. Laboratory measurements All participants were asked to fast for 12 h before their visit and to eat normally in the period prior to their attendance. Venesection was performed after blood pressure readings with the subject seated. Assays for glucose, cholesterol, HDL-cholesterol, and triglycerides 75 g of anhydrous glucose was dissolved in 200 ml of distilled water were performed using standard techniques, described in detail elsewhere (Janus, 1997). The laboratory participated in the Murex External Quality Control Programme for glucose and in the Royal College of Pathologists of Australasia=Australasian Association of Clinical Biochemists (RCPA=AACB) Quality Control Program for Lipids and Lipoproteins. Statistical analysis Since dietary intake and cardiovascular risk factor prevalence change with age, (Leung et al, 1997) the MANOVA test adjusted for age was used to examine differences in mean values between different educational levels, and marital status, within the same sex. The Chi-square test was used to examine the association between age, education and marital status and being overweight. Results The age and sex distribution of the 1995 dietary survey sample is shown in Table 1. Data was incomplete for one subject for the analysis by education level, and for three subjects for the analysis by marital status. The mean nutrient intake per day by education level is shown in Table 2a, and by marital status, in Table 2b. After adjusting for age, with increasing educational level, the percentage

3 protein intake increases in men, and the percentage fat intake increases in women. The nutrient densities of protein, fat, niacin, vitamin D and PUFA also increase in women, while that of bre and calcium increase in both men and women. The difference is mainly noted between the primary and higher educational level groups. After adjusting for age, women who are single have lower nutrient density of vitamin D and iron. Women with higher educational levels have higher fruit consumption per week, while men have higher consumption of dairy products (Table 3a). Women who are single have lower consumption of vegetables and sh per week (Table 3b). Only food groups showing signi cant differences are shown. Women with higher education level have lower mean body mass index (BMI) and waist hip ratio (WHR), while men with higher educational level have lower systolic BP and higher diastolic BP (Table 4a). No difference in laboratory parameters is observed. Both men and women who were single had lower mean BMI (Table 4b). Single men also had lower diastolic BP, serum triglycerides, and HDL-cholesterol. The overall prevalence of obesity (BMI 30 kg=m 2 ) was 5% for men and 6.7% for women, while that of overweight (BMI 27 kg=m 2 ) was 20.8% for both men and women. Signi cant factors associated with overweight Table 1 Age and sex distribution of subjects in the 1995 dietary survey Age (y) M F Total < ± ± All ages in women were increasing age, lower education level, and living with spouse (Table 5). Discussion As in other developed countries, the burden of disease in Hong Kong has shifted from infective illnesses to chronic diseases, such that the top three causes of mortality are neoplasm, ischaemic heart disease and stroke. There is an association between dietary factor and chronic diseases (Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group, 1994; Gurr, 1992; Jha et al, 1995; Ziegler, 1989), so that it is pertinent to examine the dietary habits of the population. The social class gradient in mortality and morbidity (Rose & Marmot, 1981; Antonovsky, 1968; James et al, 1997) may be mediated partly by differences in dietary habits, and it would be of interest to examine the effect of social factors in the Hong Kong Chinese population, as little data is available. In the 1995±96 Hong Kong Cardiovascular Risk Factor Prevalence Study and Dietary Survey of a representative sample of the Hong Kong Chinese adult population, information regarding educational level and marital status was also collected, enabling the in uence of these two factors on food consumption, anthropometry and other cardiovascular risk factors to be examined. Although details of income were not available, educational level may be used as an approximate surrogate. Marital status represents another social dimension, as dietary habits could be in uenced by living arrangement. Overall, higher levels of education appears to be associated with a `healthier' dietary pattern in relation to chronic diseases. Therefore bre and calcium intakes were higher among both men and women. The bene cial effects of bre (Burkitt et al, 1974) include reduced risk of gallstones, diverticulitis, neoplasm, and possibly hyperten- 463 Table 2a Age-adjusted mean nutrient intake per day by education level Mean value (s.d.) Tertiary Secondary Primary Total M F n ˆ 79 n ˆ 55 n ˆ 311 n ˆ 244 n ˆ 110 n ˆ 210 n ˆ 500 n ˆ 509 Energy (kcal) 2331.(480) 1909.(641) 2494.(712) 1828.(536) 2277.(626) 1773.(584) 2421.(667) 1814.(568) % Protein 18.7 (3.2) 19.0 (2.8) 18.5 (3.0) 19.0 (3.4) 17.5 (2.8) 18.6 (3.5) 18.3 (3.1) * 18.8 (3.4) % Fat 28.9 (6.7) 30.2 (5.0) 29.2 (5.5) 30.2 (5.3) 27.3 (6.1) 27.5 (5.2) 28.7 (5.9) 29.1 (5.4) * % Carbohydrate 53.2 (9.2) 52.1 (6.9) 52.9 (7.5) 52.2 (7.6) 55.8 (7.9) 55.1 (7.5) 53.6 (7.9) 53.4 (7.6) Nutrient=1000 kcal Protein (g) 46.6 (8.0) 47.6 (7.0) 46.3 (7.6) 47.5 (8.5) 43.7 (7.0) 46.5 (8.7) 45.8 (7.6) * 47.1 (8.4) Fat (g) 32.1 (7.5) 33.6 (5.5) 32.4 (6.1) 33.6 (5.9) 30.4 (6.8) 30.5 (5.8) 31.9 (6.5) 32.3 (6.0) * Carbohydrate (g) (22.9) (17.2) (18.8) (19.1) (19.7) (18.7) (19.9) (19.1) Cholesterol (mg) (56.6) (61.5) (51.7) (50.7) (60.6) (51.4) (54.7) (52.6) Fibre (g) 3.6 (1.8) 5.3 (2.5) 3.4 (1.5) 4.8 (2.0) 3.3 (1.3) 4.7 (1.8) 3.4 (1.5) *** 4.8 (2.0) ** Vitamin A (ug) (337.8) (516.8) (373.0) (460.1) (423.7) (583.1) (379.9) (519.3) Thiamine (mg) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) Ribo avin (mg) 0.5 (0.1) 0.6 (0.1) 0.5 (0.1) 0.6 (0.2) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) 0.6 (0.1) Niacin (mg) 8.7 (1.9) 8.9 (2.0) 8.3 (1.9) 8.8 (2.1) 7.8 (1.9) 8.6 (2.3) 8.3 (1.9) * 8.7 (2.1) Vitamin C (mg) 62.4 (35.4) (62.8) 59.5 (35.4) 95.5 (53.9) 66.6 (37.6) 99.1 (54.9) 61.6 (35.9) 97.8 (55.3) Vitamin D (ug) 9.1 (10.2) 10.1 (11.1) 7.8 (6.5) 7.7 (7.1) 7.4 (8.8) 5.8 (5.7) 7.9 (7.7) 7.2 (7.3) *** Calcium (mg) 257.(63) 325.(99) 256.(93) 321.(117) 240.(83) 312.(111) 252.(87) *** 318.(112) * Iron (mg) 7.1 (1.7) 8.0 (1.9) 7.3 (1.7) 8.2 (2.5) 7.4 (1.7) 8.3 (2.0) 7.2 (1.7) 8.2 (2.3) Zinc (mg) 6.2 (4.2) 5.8 (1.5) 5.8 (2.1) 6.4 (8.8) 5.9 (2.4) 5.8 (2.0) 5.9 (2.6) 6.1 (6.3) SFA (g) 8.9 (2.5) 9.2 (2.3) 9.1 (2.3) 9.1 (2.2) 8.3 (2.5) 8.1 (2.1) 8.9 (2.3) 8.7 (2.2) MUFA (g) 11.4 (3.0) 11.6 (2.7) 11.6 (2.7) 11.5 (2.4) 10.7 (3.1) 10.3 (2.3) 11.4 (2.8) 11.0 (2.5) PUFA (g) 7.3 (1.8) 7.8 (1.6) 7.0 (1.6) 7.9 (1.7) 6.7 (1.5) 7.1 (1.7) 7.0 (1.6) 7.5 (1.7) * SFA ˆ saturated fatty acid. MUFA ˆ monounsaturated fatty acid. PUFA ˆ polyunsaturated fatty acid. MANOVA Test adjusted for age within the same sex: * P<0.05; ** P<0.01; *** P<0.001.

4 464 Table 2b Mean nutrient intake per day by marital status Mean value (s.d.) Single Married Total n ˆ 72 n ˆ 88 n ˆ 427 n ˆ 420 n ˆ 499 n ˆ 508 Energy (kcal) 2637.(743) 1762.(525) 2388.(644) 1825.(578) 2424.(664) 1814.(569) % Protein 18.6 (3.0) 18.8 (3.0) 18.3 (3.1) 18.9 (3.5) 18.3 (3.1) 18.9 (3.4) % Fat 30.1 (6.1) 29.8 (5.7) 28.5 (5.8) 29.0 (5.3) 28.7 (5.9) 29.1 (5.4) % Carbohydrate 51.9 (7.8) 52.6 (7.8) 53.8 (8.0) 53.6 (7.6) 53.6 (8.0) 53.4 (7.6) Nutrient/1000 kcal Protein (g) 46.4 (7.4) 47.0 (7.4) 45.7 (7.7) 47.2 (8.7) 45.8 (7.6) 47.2 (8.5) Fat (g) 33.4 (6.7) 33.1 (6.3) 31.7 (6.5) 32.2 (5.9) 31.9 (6.5) 32.3 (6.0) Carbohydrate (g) (19.4) (19.4) (19.9) (19.0) (19.9) (19.1) Cholesterol (mg) (49.2) (52.0) (55.6) (52.9) (54.7) (52.8) Fibre (g) 2.9 (1.2) 4.5 (1.7) 3.5 (1.5) 4.9 (2.0) 3.4 (1.5) 4.8 (2.0) Vitamin A (ug) (245.6) (375.8) (392.7) (544.0) (379.4) (519.7) Thiamine (mg) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) 0.5 (0.1) Ribo avin (mg) 0.5 (0.1) 0.6 (0.1) 0.5 (0.1) 0.6 (0.2) 0.5 (0.1) 0.6 (0.1) Niacin (mg) 8.3 (1.8) 8.6 (1.8) 8.3 (1.9) 8.7 (2.2) 8.3 (1.9) 8.7 (2.1) Vitamin C (mg) 47.7 (28.0) 90.0 (53.5) 63.8 (36.7) 99.2 (55.6) 61.5 (36.0) 97.6 (55.3) Vitamin D (ug) 7.4 (5.3) 6.5 (5.6) 8.0 (8.1) 7.4 (7.6) 7.9 (7.7) * 7.2 (7.3) Calcium (mg) 232.(63) 299.(96) 256.(90) 321.(115) 252.(87) 317.(112) Iron (mg) 6.8 (1.4) 7.6 (2.0) 7.3 (1.7) 8.3 (2.3) 7.2 (1.7) 8.2 (2.3) * Zinc (mg) 5.6 (1.2) 5.6 (1.6) 6.0 (2.8) 6.2 (6.8) 5.9 (2.6) 6.1 (6.3) SFA (g) 9.5 (2.3) 9.3 (2.3) 8.8 (2.3) 8.5 (2.2) 8.9 (2.3) 8.7 (2.2) MUFA (g) 12.1 (2.7) 11.6 (2.7) 11.2 (2.8) 10.9 (2.4) 11.3 (2.8) 11.0 (2.5) PUFA (g) 7.0 (1.6) 7.5 (1.6) 7.0 (1.6) 7.6 (1.8) 7.0 (1.6) 7.5 (1.7) MANOVA Test adjusted for age within the same sex: * P < Table 3a Mean intake of food groups per week by education level Mean (s.d.) g=week Tertiary Secondary Primary Total M F n ˆ 79 n ˆ 55 n ˆ 311 n ˆ 244 n ˆ 110 n ˆ 210 n ˆ 500 n ˆ 509 Fruits 1530 (1140) 1899 (1174) 1410 (952) 1572 (897) 1323 (907) 1453 (920) 1410 (975) 1558 (947) ** Dairy products 390 (435) 356 (412) 351 (477) 366 (457) 218 (310) 338 (493) 332 (446) * 353 (465) MANOVA Test adjusted for age within the same sex: * P < 0.05; ** P < Table 3b Mean food intake per week by marital status Mean.(s.d.) g/week Single Married Total n ˆ 72 n ˆ 88 n ˆ 427 n ˆ 420 n ˆ 499 n ˆ 508 Vegetables 1877.(1022) 1743.(1103) 2154.(1318) 2189.(1242) 2114.(1282) 2111.(1230) * Fish 668.(504) 505.(316) 700.(595) 638.(440) 695.(579) 615.(424) * MANOVA Test adjusted for age within the same sex: * P < sion (He et al, 1995), while calcium intake is important in the prevention of osteoporosis related fractures (Lau et al, 1988). The bene cial effect is more marked in women, as higher levels of education are associated with a higher density of a wider range of nutrients compared with men. The consumption of fruits and dairy products also increases with educational level. It is uncertain whether the increase in consumption of protein and fat with higher education level is bene cial or deleterious in terms of cardiovascular disease risk. Increasing fat intake as a percentage of total calories is considered a risk factor for coronary heart disease (Murphy et al, 1992). However, it has been observed that as the nutritional adequacy of the diet rises, the percentage of energy from fat also tends to rise slightly, implying selection of higher-fat sources of nutrients (The International Lipid Information Bureau, 1995). It has been observed that diets high in protein and bre content are associated with a fall in blood pressure (He et al, 1995). The effect of the amount of protein consumed on health would depend on whether dietary intake is de cient in protein or otherwise. Higher educational level appears to be associated with a lower risk of being overweight, but this only applies to women. Similar to the Whitehall study of civil servants

5 Table 4a Cardiovascular risk factors by educational level Mean (s.d.) 465 Tertiary Secondary Primary Total M F n ˆ 79 n ˆ 55 n ˆ 311 n ˆ 244 n ˆ 110 n ˆ 210 n ˆ 500 n ˆ 509 Body mass index (kg=m 2 ) 24.3 (2.8) 22.0 (2.9) 24.3 (3.3) 23.1 (3.4) 24.4 (3.7) 25.6 (4.1) 24.3 (3.3) 24.0 (3.9) *** Waist hip ratio (cm=cm) 0.87 (0.06) 0.76 (0.07) 0.89 (0.23) 0.78 (0.07) 0.91 (0.08) 0.84 (0.08) 0.89 (0.19) 0.81 (0.08) *** Systolic BP (mmhg) 120 (17) 110 (13) 120 (17) 112 (18) 123 (17) 126 (19) 121 (17) * 118 (19) Diastolic BP (mmhg) 78 (10) 71 (9) 76 (11) 70 (11) 76 (10) 77 (11) 77 (10) *** 73 (11) MANOVA Test adjusted for age within the same sex: * P < 0.05; ** P < 0.01; *** P < Table 4b Cardiovascular risk factors by marital status Mean (s.d.) Single Married Total n ˆ 72 n ˆ 88 n ˆ 427 n ˆ 420 n ˆ 499 n ˆ 508 Body mass index (kg=m 2 ) 23.4 (3.3) 22.6 (4.1) 24.5 (3.3) 24.3 (3.8) 24.3 (3.3) * 24.0 (3.9) * Waist hip ratio (cm=cm) 0.85 (0.06) 0.77 (0.07) 0.89 (0.2) 0.81 (0.08) 0.89 (0.19) 0.81 (0.08) Systolic BP (mmhg) 113 (13) 112 (18) 122 (17) 119 (20) 121 (17) 118 (19) Diastolic BP (mmhg) 71 (10) 71 (11) 77 (10) 74 (11) 77 (10) * 73 (11) Triglyceride (mmol) 1.08 (0.60) 0.82 (0.37) 1.43 (1.12) 1.10 (0.89) 1.38 (1.07) * 1.05 (0.83) Chol=HDL (mmol=mmol) 4.2 (1.2) 3.6 (1.0) 4.8 (1.8) 3.9 (1.4) 4.7 (1.7) * 3.8 (1.3) MANOVA Test adjusted for age within the same sex: *P < Table 5 Factors associated with overweight (BMI 27 kg/m 2 ) BMI < 27 (kg=m 2 ) BMI > ˆ 27 (kg/m 2 ) n ˆ 396 n ˆ 404 n ˆ 104 n ˆ 106 n ˆ 500 n ˆ 510 Age: % % % % O.R. (95% C.I.) < *** ± (0.5±1.7) 1.8 (0.8±4.1) 45± (0.6±2.0) 4.7 (2.2±10.0) (0.7±2.5) 4.9 (2.3±10.3) After adjusting for age: Education: Tertiary *** Secondary (0.7±2.4) 2.4 (0.7±8.0) Primary (0.5±2.3) 7.1 (2.1±24.2) Marital status: Married ** Single (0.3±1.5) 0.4 (0.2±1.0) Chi-Square test, same sex. ** P 0.01; *** P (Rose & Marmot, 1981), a lower mean systolic blood pressure in men is associated with higher educational level. No other in uence on cardiovascular risk factors was observed. These ndings contrast with the results of a cardiovascular risk factor prevalence survey in Hong Kong Chinese carried out in 1983±85 (To & Donnan, 1988), showing that higher socioeconomic groups had an increased prevalence of risk factors, in that mean blood pressure, serum cholesterol and body mass index were higher. Men in the higher socioeconomic groups also had a higher prevalence of ischaemic heart disease. However, a later study carried out in a cohort of Chinese subjects undergoing routine oral glucose tolerance test in a laboratory showed that the prevalence of obesity, diabetes mellitus and other cardiovascular risk factors was higher among those with lower educational level (Ko et al, 1997), a nding similar to that of this present survey. Surveys in China also showed that higher socioeconomic groups have increased fat intake, increased consumption of rice, wheat, fresh fruit, dairy products, red meat, poultry, eggs, sh and seafood. However, the prevalence of obesity also increases (Popkin et al, 1993b). An examination of changes in body weight from 1989±1991 showed that higher income groups have an increasing tendency to become overweight (Popkin et al, 1995). This observation is similar to the situation in Hong Kong in 1983 (To & Donnan, 1988). A more recent Chinese National Nutrition Survey in 1992 showed that in general, higher income groups have a less healthy intake pattern with respect to cardiovascular disease, particularly in urban compared with rural areas. However, although the consumption of fat and animal product is higher in urban areas, the intakes of fruits, vitamin C and calcium were also higher (Institute of Nutrition and Food Hygiene, 1995).

6 466 In general, dietary surveys of Western populations show that a healthier diet is associated with higher socioeconomic or educational status. The Australian Surveys in 1983 and 1994 showed that women and those of higher socioeconomic status or education level have healthier diets (Dobson et al, 1997). A survey examining time trends in body mass index among young white and black adult men between 1960±1980 showed that initially there was a positive association between body mass index and education, which subsequently changed to a negative one (Flegal et al, 1988). Another survey in the USA (NHANES II) showed that increased level of education is associated with increased consumption of dairy products, meat, grain, fruit and vegetables (Kant et al, 1991). In the Netherlands, higher socioeconomic status is associated with a dietary intake closer to health recommendations (Hulshof et al, 1991). In contrast, in elderly Zimbabweans, body mass index increases with education level, as does the frequency of eating meat and fried food (Allain et al, 1997). It is likely that the in uence of education on dietary intake, obesity and other cardiovascular risk factors changes with time, depending on the degree of economic development of a country. Initially, increasing education (and probably income) may be associated with a less healthy diet in relation to cardiovascular health, obesity, and other risk factors. With time, the association becomes reversed. This appears to be the situation in Hong Kong, over a 10±20 y period. The same transition may be observed among different countries, re ecting the rate of economic development: therefore this transition was observed in Western populations rst, then in Hong Kong, followed by urban and rural mainland China, and perhaps Africa. In parallel, other lifestyle factors may also be affected by socioeconomic changes, such as the level of physical activity. The latter may account for differences in dietary habits and cardiovascular risk factor levels between rural and urban populations, or between economically strong countries and those which are poorer. The in uence of marital status on dietary intake is different between women and men. No association was observed for men. However, women not living with a spouse tend to have lower nutrient density of iron and vitamin D, and lower consumption of vegetables and sh. It is possible that in such circumstances, women may pay less attention to meal preparation. Other possible explanations are that these women tend to have less income, or have a lower educational level. In terms of cardiovascular risk factors, not living with spouse appears to result in a better pro leðlower body mass index in both men and women, and lower diastolic blood pressure, serum triglerides, and cholesterol=hdl ratio in men. The explanation for this observation is unclear, but may not be related to dietary habits, since there was no difference in energy and other macronutrient intake between married and single subjects after adjusting for age. Differences in other factors such as level of physical activity may account for this difference. Conclusions This study shows that it is important to examine the effect of individual socioeconomic factors (education, marital status) on dietary intake and cardiovascular risk factors. For Hong Kong Chinese, higher levels of education are associated with a healthier diet and a lower prevalence of overweight, showing a reversal of the trend 10±20 y ago. From the public health view point, dietary education needs to be targeted particularly at the primary schools, and also women with lower educational levels, in addition to the general health education efforts for the whole population. References Allain TJ, Wilson AO, Gomo ZAR, Adamchak DJ & Matenga JA (1997): Diet and nutritional status in elderly Zimbabweans. Age Ageing 26, 463±470. Alpha-Tocopherol, Beta carotene Cancer Prevention Study Group (1994): The effect of vitamin E and beta-carotene on the incidence of lung cancer and other cancer in male smokers. N. Engl. J. Med. 330, 1029±1035. Antonovsky A (1968): Social class and the major cardiovascular diseases. J. Chron Dis. 21, 65±106. Burkitt DP, Walker AR & Painter NS (1974): Dietary bre and disease. JAMA 229, 1068±1074. Dobson A, Porteous J, Mc Elduff P & Alexander H (1997): Whose diet has changed? Aust. NZ. J. Public Health 21, 147±154. Flegal KM, Harlan WR & Landis JR (1988): Secular trends in body mass index and skinfold thickness with socioeconomic factors in young adult men. Am. J. Clin. Nutr. 48, 544±551. Gurr MI (1992): Dietary lipids and coronary heart disease: old evidence, new perspective. Prog. Lipid Res. 31, 195±243. He J, Klag MJ, Ghelton PK, Chen JY, Qian MC & He GQ (1995): Dietary macronutrients and blood pressure in Southwestern China. J. Hypertension 13, 1267±1274. Holland B, Welch AA, UnwinID, BussDH, Paul AA& Southgate DAI (1992): McCanceandWiddowson's: TheCompositionofFoods. 5thEdn.TheRoyal Society of Chemistry and Ministry of Agriculture, Fisheries and Food. Hulshof KF, Lowik MR, Kok FJ, Wedel M, Brants HA, Hermus RJ & tenhoor F (1991): Diet and other life-style factors in high and low socio-economic groups (Dutch Nutrition Surveillance System). Eur. J. Clin. Nutr. 45, 441±50. Institute of Health (1992): Food Composition Tables. Chinese Medical Science Institute, China: Peoples Health Press. (Chinese). Institute of Nutrition and Food Hygiene (1995): The Dietary and Nutritional Status of Chinese Population (1992 National Nutrition Survey), Vol. 1. eds. K Ge, L Cheng, X Guo, Q Wang, J Hu, F Jia, J Liu & J Li. Chinese Academy of Preventive Medicine: People's Medical Publishing House. James WPT, Nelson M, Ralph A & Leather S (1997): The contribution of nutrition to inequalities in health. Br. Med. J. 314, 1545±1549. Janus ED (1997): The Hong Kong Cardiovascular Risk Factor Prevalence Study, 1995±1996. Hong Kong: Prime Media. Jha P, Flather M, Lonn E, Farkouh M & Yusuf S (1995): The antioxidant vitamins and cardiovascular disease. A critical review of epidemiologic and clinical trial data. Ann. Int. Med. 123, 860±872. Kant AK, Block G, Schatzkin A, Ziegler RG & Nestle M (1991): Dietary diversity in the US population, NHANES II 1976±1980. J. Am. Diet Assoc. 91, 1526±1531. Ko GTC, Chan JCN, Yeung VTF, Chow CC, Li JKY, So WY, Tsang LLW & Cockram CS (1997): Socio-economic status, prevalence of diabetes and other cardiovascular risk factors in Chinese. Diabetologia 40, A209. Lau E, Donnan S, Barker DJP & Cooper C (1988): Physical activity and calcium intake in fracture of the proximal femur in Hong Kong. Br. Med. J. 297, 1441±1443. Leung SSF, Ho SC, Woo J, Lam TH & Janus ED (1997): Hong Kong Adult Dietary Survey, Hong Kong: Prime Media. MurphySP,RoseD,HudesM& ViteriFE(1992):Demographicandeconomic factors associatedwith dietary qualityforadultsinthe1987±88nationwide Food Consumption Survey. J. Am. Diet Assoc. 92, 1352±1357. Popkin BM, Ge K, Zhai F, Guo X, Ma H & Zohoori N (1993a): The nutrition transition in China: a cross-sectional analysis. Eur. J. Clin. Nutr. 47, 333±346. Popkin BM, Paeratakul S, Zhai F & Gek (1993b): Dietary and environmental correlates of obesity in a population study in China. Obes. Res. 3, 135S±143S. Popkin BM, Paeratakul S, Ge K & Zhai F (1995): Body weight patterns among the Chinese: Results from the 1989 and 1991 China Health and Nutrition Surveys. Am. J. Public Health 85, 690±694. Rose G & Marmot MG (1981): Social class and coronary heart disease. Br. Heart J. 45, 13±9. Shea S, Melnik TA, Stein AD, Zansky SM, Maylahn C & Basch CE (1993): Age, sex, educational attainment, and race=ethnicity in relation to consumption of speci c foods contributing to the atherogenic potential of diet. Prev. Med. 22, 203±218.

7 The International Lipid Information Bureau (1995): The IL1B Lipid handbook: Clinical Guide. The International Lipid Information Bureau, USA. To CY & Donnan SPB (1988): Health risks, tness and quality of life in adults in Hong Kong. The Chinese University of Hong Kong. Tsang YS & Yung SW (1991): Diet and Health Guide and Composition of Food in Southern China.China: Zhongshan University Press. (Chinese). Woo J, Leung SSF, Ho SC, Lam TH & Janus ED (1997): A food frequency questionnaire for use in the Chinese population in Hong Kong: description and examination of validity. Nutr. Res. 17, 1633±1641. Yu TS, Wong SL, Lloyd OL & Wong TW (1995): Ischaemic heart disease: trends in mortality in Hong Kong, 1970±89. J. Epidemiol. Community Health 49, 16±21. Ziegler RG (1989): A review of epidemiological evidence that carotenoids reduce the risk of cancer. J. Nutr. 119, 116±