Nutrient Intake of Adult Women in Jilin Province, China, with Special Reference to Urban ± Rural Differences in Nutrition in the Chinese Continent

Nutrient intake of adult women in Jilin province, China, with special reference to urban ± rural differences in nutrition in the Chinese continent

J-B Qu1, Z-W Zhang2, S Shimbo2, Z-M Liu3, X-C Cai4, L-Q Wang4, T Watanabe5, H Nakatsuka6, N Matsuda-Inoguchi6, K Higashikawa7 and M Ikeda7*

1Department of Hygiene, Faculty of Public Health, Shandong Medical University, Jinan 250012, Shandong, People's Republic of China; 2Department of Food and Nutrition, Kyoto Women's University, Kyoto 605 ± 8501, Japan; 3Division of Environmental Health Sciences, Tohoku University Postgraduate School of Medicine, Sendai 980 ± 8575, Japan; 4Jilin Provincial Sanitary and Anti-epidemic Station, Changchun 130051, Jilin Province, People's Republic of China; 5Miyagi University of Education, Sendai 980 ± 0845, Japan; 6Miyagi University, Taiwa-cho 981 ± 3298, Japan; and 7Kyoto Industrial Health Association, Kyoto 604 ± 8472, Japan

Objectives: The objectives of this study were to investigate possible urban ± rural differences in food intakes in Jilin province and in continental China as a whole, and to examine possible implications for nutritional status of urban and rural populations. Design: Cross-sectional study. Setting: Communities. Subjects: In total, 499 adult women in six urban sites and four rural sites, 10 sites in total, including two sites in Jilin province. Methods: A pair of surveys were conducted in September, 1999, in the provincial capital of Changchun and a farming village in Dehui county, both in Jilin province, in northeast China. Each of 50 adult women per survey site provided a 24 h duplicate food sample and a blood sample, and had an interview on health history including anthropometry and blood pressure measurement. Nutrient intakes were estimated from the food duplicates, using national food composition tables. Results from the two sites were supplemented with data from eight sites where surveys had been conducted following the same protocol, and the pooled material were subjected to analyses for possible urban ± rural differences. Results: The Jilin participants consumed daily, on average, about 1600 kcal energy, 44 g protein, and 60 g lipid with a lipid energy ratio (ie the ratio of lipid over total nutrients in terms of energy) of 33%. When nutrient intakes were compared between the urban (ie Changchun) and rural (Dehui) groups, urban women consumed more energy, protein (especially animal protein) and lipid than rural women. Similar examination of data from six urban and four rural sites, including the present two, showed that adult women in urban areas eat more animal protein and animal fat than their counterparts in villages, and suggested that the observation on urban ± rural difference in Jilin province can be extrapolated to a nationwide scale. Conclusions: Urban ± rural differences in nutrient intakes still persist in 1999 not only in Jilin but in other provinces, typically in the terms of intakes of animal-based foods. Sponsorship: The Ministry of Education, Science, Sports and Culture, the Government of Japan. Descriptors: adult women; food duplicate study; Jilin province, China; nutrient intake; urban ± rural difference European Journal of Clinical Nutrition (2000) 54, 741±748

Introduction surveys in various sites, differences in food habits and major nutrient sources have been observed between people It was made clear in the 1992 National Nutrition Survey in urban areas and their counterparts in rural areas in many (Ge et al, 1995,1996) that rural people in continental China east Asian countries, such as Japan (eg Ikeda et al, 1988; ate less animal-based foods (eg pork and poultry) and more Nakatsuka et al, 1988) and Korea (Moon et al, 1997), and cereals than their counterparts in urban areas. Through ®eld most probably also China (Tian et al, 1995). The differences in food consumption are often associated with better *Correspondence: M Ikeda, Kyoto Industrial Health Association, 67 nutrient intake in urban communities than in their rural Nishinokyo-Kitatsuboicho, Nakagyo-ku, Kyoto 604 ± 8472, Japan. counterparts (Ge et al, 1996). Guarantor: M Ikeda. The objectives of the present analysis are to clarify if the Contributors: J-B Qu designed the study and took responsibility for local urban ± rural differences in nutrient intake still persist in the arrangements. Z-W Zhang and S Shimbo were responsible for clinical evaluation. Z-M Liu, X-C Cai and L-Q Wang made arrangements for late 1990s and, if they do, determine the effects on the collection of food and blood samples including health examination. T nutritional status of the people. This study group has been Watanabe obtained a research grant for the study and made nutritional conducting ®eld surveys to elucidate the nutritional char- analysis in cooperation with H Nakatsuka and N Matsuda-Inoguchi. K acteristics of working women in the Chinese continent Higashikawa made statistical analysis including table preparation, and M (Zhang et al, 1997; Qu et al, 1997; Shimbo et al, 1997; Ikeda drafted the paper. Most of the contributors worked together in the ®eld survey and all contributors reviewed the paper. Watanabe et al, 1998; Nakatsuka et al, 1999). As an Received 2 March 2000; revised 17 May 2000; accepted 14 June 2000 extension of these studies, surveys were conducted in a Nutrient intakes in China J-B Qu et al 742 provincial capital in combination with a farming village in duplicate (Acheson et al, 1980) of her food together with a the same area of northeast China. Results are presented in menu record, to allow blood sampling (without fast) and to this report, in connection with the observations in the be interviewed on her health history (including anthropo- foregoing surveys at eight sites, to examine possible metry, blood pressure measurement and information on urban ± rural differences in nutrition in the continent in smoking and drinking habits); each participant was care- the late 1990s. fully instructed to cook ordinary everyday meals and no special dishes. Only women were selected because they had better opportunities to offer food duplicate samples. Materials and methods Survey population, study design and collection of food Nutritional evaluation and other examinations duplicate samples Procedures for nutritional evaluation have also been Two surveys were conducted in a paired manner in the described previously (Shimbo et al, 1994, 1996; Ikeda middle of September 1999, one in Changchun, the provin- et al, 1996; Moon et al, 1997; Zhang et al, 1997). Each cial capital, and the other in a farming village in Dehui food item in the duplicate set was separated manually, and county (to be called Dehui) in Jilin province in northeast its weight was recorded. The items were coded by experi- China; the village is some 100 km north-east-north of enced Chinese nutritionists using the Food Composition Changchun (for locations of the survey sites, see Figure Tables for China (Part 1, National values; Institute of 1). In each site, 50 adult healthy working women volun- Nutrition and Food Hygiene, Chinese Academy of Preven- teered to participate; those with systemic diseases at the tive Medicine 1991a). Estimation of nutrients in each meal time of survey or in the past were excluded from the and snack was made by use of the code numbers and the database. Changchun participants (the urban group) were weight records (Zhang et al, 1997) in accordance with the women staff of a public health institution, whereas those food Composition tables. from the village (the rural group) were farmers engaged in Nationally recommended dietary allowances for Chinese production of wheat, rice and maize. (RDA; Institute of Nutrition and Food Hygiene, Chinese The details of the study protocol have been described Academy of Preventive Medicine, 1991b) were employed; previously (Shimbo et al, 1996, 1997; Ikeda et al, 1996; the RDA values for adult women in three age groups are Moon et al, 1997; Zhang et al, 1997; Nakatsuka et al, cited in Table 1. In applying the RDA, light-grade physical 1999). In short, each participant was asked to offer a 24 h exercise was assumed even for farmers, because the sur-

Figure 1 Locations of the paired survey sites in Jilin province, northeast China.

European Journal of Clinical Nutrition Nutrient intakes in China J-B Qu et al 743 Table 1 Recommended dietary allowances for Chinese women for the weight of clothing by subtracting 1 kg from the measured weight, so that BMI (body weight Age (y) (kg) 7 1)=(height (m))2. A criterion of 25 (kg=m2) was Energy and nutrient (unit=day) 18 ± 44 45 ± 59 60 ± 69 taken as identifying overweight cases (Institute of Nutrition and Food Hygiene, Chinese Academy of Preventive Energy (kcal) 2100 1900 1700 Medicine, 1991b). Protein (g) 65 65 60 Lipid energy ratio (%) 20 ± 25 20 ± 25 20 ± 25 Minerals Statistical analysis Calcium (mg) 800 800 800 STATVIEW Software Version 5 for PC, was employed. A Iron (mg) 18 12 12 normal distribution was assumed for intake of most nutri- Zinc (mg) 15 15 15 Vitamins ents, so that arithmetic means (AM) and arithmetic stan- Vitamin A (REa mg) 800 800 800 dard deviations (ASD) were taken to represent distribution. Vitamin B1 (mg) 1.1 1.2 1.2 Student's unpaired t-test and ANOVA were employed for Vitamin B2 (mg) 1.1 1.2 1.2 detection of signi®cant differences in means, and the chi- Nicotinic acid (mg) 11 12 12 square test in prevalence. Vitamin C (mg) 60 60 60 Vitamin D (mg) 5 5 10 Vitamin E (mg) 10 12 12 Results Cited from Institute of Nutrition and Food Hygiene, Chinese Academy of Preventive Medicine (1991b); light physical exercise is assumed. Demography of the study populations, and health aIn retinol equivalency. parameters Ages of the participants in the two sites were 41 y on average with a wide variation (Table 2). Body weight was veys were made in an off-season for agriculture. The distributed in a wide range from 35 to 70 kg. Accordingly, nutrient intake was considered as suf®cient when the ASD for BMI was also large ( > 3). Overweight cases were estimate was within a range of 80 ± 120% (thus with an distributed evenly (P > 0.10) in the two study populations. allowance of+ 20% RDA) of the corresponding RDA, The prevalence of anemia was quite low: 0% when taking day-to-day variation in food intake into considera- evaluated in terms of red blood cell (RBC) counts (ie tion (Shimbo et al, 1994; Zhang et al, 1997). A ratio of erythrocytopenia with < 3.306106 RBC=mm3) and 5% in 40 ± 50% (or 45% as an average) of animal protein over terms of hypohemoglobinemia (ie hemoglobin (Hb) total protein was taken as adequate (Committee on Health concentration below 12.0 g=100 ml) with no inter-site and Nutrition, Japan, 1999). difference (P > 0.10). Triglyceride (TG) was elevated Hematology and serum biochemistry tests were carried (ie > 151 mg=100 ml) in 36% of the women. The observed out with automated analyzers, whereas urinalysis for pro- elevation might be due to temporary effects of a meal at tein etc was by clinical tapes. To calculate BMI (body mass least in some cases because no fasting was requested, but index), body weight was measured clothed and corrected BMI was also substantially elevated in two women in the

Table 2 Age distribution, anthropometry, hematology and other ®ndings among participants in the two survey sites

Participants in

Parameter (unit) All Urban site Rural site P a

No. of participants 100 50 50 Age (y) 41.0+ 9.5 39.5+ 8.6 42.4+ 10.2 NS Minimum 7 maximum (y) 21 ± 61 25 ± 58 21 ± 61 Height (H) (cm) 157.2+ 6.0 159.5+ 6.0 155.0+ 5.2 ** Weight (W) (kg) 56.0+ 8.7 59.0+ 8.4 53.0+ 8.0 ** BMI (W=H2) 22.6+ 3.2 23.2+ 3.1 22.1+ 3.2 { Overweight (%)b 20 (20%) 13 (26%) 7 (14%) NS Erythrocyte counts (x104=mm3) 444+ 32 448+ 31 440+ 33 NS Erythrocytopenia casesc (%) 0 (0%) 0 (0%) 0 (0%) NS Hemoglobin (g=100 ml) 13.4+ 0.9 13.3+ 0.9 13.4+ 0.8 NS Hypohemoglobinemia casesd (%) 5 (5%) 3 (6%) 2 (4%) NS Albumin (g=100 ml) 4.0+ 0.2 4.1+ 0.1 3.8+ 0.2 ** Hypalbuminemia casese (%) 1 (1%) 0 (0%) 1 (2%) NS Hypertriglyceridemia casesf (%) 36 (36%) 23 (46%) 13 (26%) * Liver dysfunction casesg (%) 1(1%) 1 (2%) 0 (0%) NS Hypertensive casesh (%) 12 (12%) 8 (16%) 4 (8%) NS

Values are AM+ ASD or number of cases (%), unless otherwise speci®ed. aMarks indicate that the means (by t-test) or the prevalence (by chi-square test) are signi®cantly (**,* and { for P < 0.01, 0.05 and 0.10, respectively; NS for P > 0.10) different between the two survey sites. bThe ratio of subjects who had BMI of 25 (Institute of Nutrition and Food Hygiene, Chinese Academy of Preventive Medicine, 1991b). c < 3.306106 RBC=mm3. d < 12.0 g Hb=100 ml. e < 3.5 g=100 ml. f > 151 mg triglyceride=100 ml. gAST > 40 IU=1, and ALT > 35 IU=1. hEither systolic pressure of > 140 mmHg or diastolic pressure of > 90 mmHg.

European Journal of Clinical Nutrition Nutrient intakes in China J-B Qu et al 744 urban group (BMI 29.2 and 27.2, respectively) and one in of the participants. Intakes of Ca and vitamins A, B1 and B2 the rural group (BMI 32.7). A possible liver dysfunction were below 80% RDA in almost all cases. case was detected in the urban group; g-GTP, AST and In contrast, intakes of Fe and vitamins C and E were ALT were slightly elevated, but the elevation in TG and suf®cient in most of the participants. When the intakes total cholesterol suggested fatty liver as the etiology. were compared between the urban and rural groups, insuf- With regard to hypertension prevalence, ®ve and three ®ciencies of energy, protein, vitamins A, C and E and cases in the urban group and two and one cases in the rural excess lipid were all more prevalent (P < 0.01 ± 0.05, group were slightly or moderately hypertensive, respec- depending on the nutrient compared) in the rural group tively. None of them, however, had > 170 mmHg SP or than in the urban group. Cases of excess intake of energy > 110 mmHg DP. or insuf®cient intakes of Ca, Fe and vitamin B2 were distributed evenly between the two groups.

Intake of energy, macro-nutrients, minerals and vitamins Leading meal for intake of food and energy Intake of energy, macro-nutrients, minerals and vitamins The amounts (both by physical weight and in terms of are summarized in Table 3. The total AM values for the energy) of foods and energy taken on each occasion of total participants show that the participants daily took about meals and snacks were compared between the two survey 1600 kcal energy, 44 g protein, 60 g lipid (with a lipid sites. Physical weight of the three meals was quite even energy ratio (the ratio of lipid over total nutrients in (28 ± 29% per meal) in the rural group, whereas breakfast terms of energy) of 33%), about 300 mg Ca, 22 mg Fe, was lighter (17%), and lunch (31%) and dinner (27%) were Na equivalent to 7.3 g NaCl, 0.55 mg vitamin B , and 1 heavier in the urban group, with lunch (and not dinner) 0.69 mg vitamin B . Study from the viewpoint of urban ± 2 being the heaviest meal (P < 0.01) of the day. rural difference showed that the urban people ate signi®- Evaluation in terms of energy recon®rmed even intake cantly (P < 0.01) more energy, protein and lipid, although among meals (31 ± 33%) in the urban group and heavy the intake of carbohydrate was comparable (P > 0.10). lunch (39%) in the rural group. Interestingly, snacks were Such was also the case for Ca and the ®ve vitamins studied heavy by weight (ie on average 531 and 202 g in the urban (P < 0.01). The intakes of Fe and Na were comparable. and rural group, respectively), but accounted for only 5% and 4% of energy in the former and the latter respectively, Prevalence of inadequate nutrient intake among the implying that snacks were predominantly of low energy, participants such as Chinese tea. Adequacy of energy or nutrient intakes of the participants was evaluated taking the national RDA (Institute of Nutri- Plant-based foods as major sources of energy, protein, lipid tion and Food Hygiene, Chinese Academy of Preventive and iron Medicine, 1991b) with 20% allowance as the criterion All food items were classi®ed by origin into plant- and (Table 4). With regard to energy intake, three women animal-based foods, and cereals and pulses were further took energy in excess, and about half had insuf®cient identi®ed among the plant-based foods. Animal-based intake. Protein intake was insuf®cient in the majority foods were separated into meats (ie meats of animals and (87%) of women, but lipid intake was in excess in 63% birds, mostly pork, and less beef and poultry), and ®sh and

Table 3 Energy and nutrient intake of participants in the two survey sites

Participants in

Energy and nutrient (unit=day) All a Urban site Rural site P a

No. of participants 100 50 50 Weight (g) 1788+ 599 2100+ 565 1475+ 454 ** Energy (kcal) 1588+ 419 1745+ 324 1432+ 447 ** Protein (g) 43.8+ 17.6 53.8+ 15.5 33.9+ 13.7 ** Lipid (g) 60.2+ 24.5 73.2+ 19.4 47.2+ 22.1 ** Lipid energy ratio (%) 33.3+ 8.5 37.7+ 6.7 29.0+ 7.8 ** Carbohydrate (g) 218.0+ 55.9 218.2+ 44.8 217.9+ 65.7 NS Crude ®ber (g) 11.2+ 4.1 11.7+ 3.6 10.7+ 4.4 NS Minerals Calcium (mg) 305+ 159 374+ 163 236+ 121 ** Iron (mg) 21.5+ 6.3 21.3+ 4.7 21.6+ 7.5 NS Phosphorus (mg) 779+ 266 893+ 232 665+ 250 ** Potassium (mg) 1683+ 521 1803+ 502 1564+ 518 * Sodium (mg) 2887+ 1437 2860+ 1023 2913+ 1767 NS NaClb (g) 7.3+ 3.6 7.3+ 2.6 7.4+ 4.5 NS Vitamins Vitamin A (RE mg) 292+ 178 356+ 187 229+ 144 ** Vitamin B1 (mg) 0.55+ 0.30 0.75+ 0.27 0.35+ 0.18 ** Vitamin B2 (mg) 0.69+ 0.26 0.79+ 0.23 0.60+ 0.25 ** Vitamin C (mg) 118.5+ 49.0 138.8+ 49.0 98.2+ 40.2 ** Vitamin E (mg) 44.1+ 20.6 51.1+ 12.7 37.1+ 24.5 **

Values are AM+ ASD. aMarks indicate that the means are signi®cantly (** and * for P < 0.01 and 0.05, respectively; NS for P > 0.10) different between the two survey sites as examined by t-test. bCalculated with an assumption that all sodium was taken as sodium chloride.

European Journal of Clinical Nutrition Nutrient intakes in China J-B Qu et al 745 Table 4 Excess or insuf®cient intake of energy and nutrients of participants in the two survey sites

No. of cases (%) in Total no. of cases Energy and nutrient Excess or Insuf®ciency (%) Urban site Rural site P a

No. of participants 100 (100%) 50 (100%) 50 (100%) Energy Excessb 3 (3%) 2 (4%) 1 (2%) NS Energy Insuf®cientc 51 (51%) 18 (36%) 33 (66%) ** Protein Insuf®cient 87 (87%) 38 (76%) 49 (98%) ** Lipid Excessd 63 (63%) 41 (82%) 22 (44%) ** Calcium Insuf®cient 100 (100%) 50 (100%) 50 (100%) NS Iron Insuf®cient 25 (25%) 13 (26%) 12 (24%) NS Vitamin A Insuf®cient 99 (99%) 49 (98%) 50 (100%) NS Vitamin B1 Insuf®cient 96 (96%) 46 (92%) 50 (100%) * Vitamin B2 Insuf®cient 94 (94%) 46 (92%) 48 (96%) NS Vitamin C Insuf®cient 10 (10%) 1 (2%) 9 (18%) ** Vitamin E Insuf®cient 5 (5%) 0 (0%) 5 (10%) *

aMarks indicate that the prevalences are signi®cantly (** and * for P < 0.01 and 0.05, respectively; NS for P > 0.10) different between the two survey sites as examined by chi-square test. bMore than 120% RDA. cLess than 80% RDA. dThe lipid energy ratio is > 30%, 1.2 times the upper limit of RDA (20 ± 25%; Institute of Nutrition and Food Hygiene, Chinese Academy of Preventive Medicine, 1991b).

shell®sh as well as eggs, milk, and others. Intakes were Table 5 Protein and fat sources of participants in the two survey sites counted in terms of energy, protein, lipid and iron, and Participants in the results were examined for detection of urban ± rural Energy and difference (Table 5). nutrient source Alla Urban site Rural site P a In all participants combined, intakes of energy, protein, No. of participants 100 50 50 lipid and iron were mostly from plant-based foods (Table Energy (kcal=day) 5); the account of plant-based foods was as high as 84% Plants 1334 (84.0%) 1431 (82.0%) 1237 (86.4%) ** and 87% for energy and iron, respectively. With regard to [Cerealsb] [730 (46.0%)] [679 (38.9%)] [781 (54.5%)] * energy and protein, almost a half of the total came from c [Pulses ] [74 (4.7%)] [73 (4.2%)] [75 (5.2%)] NS cereals. Meat 104 (6.5%) 73 (4.2%) 134 (9.4%) * Fish and shell®sh 8 (0.5%) 13 (0.7%) 3 (0.2%) * The contribution of animal-based foods (ie a combina- Eggs, milk, etc 57 (3.6%) 100 (5.7%) 14 (1.0%) ** tion of meats, ®sh, eggs and others) as protein sources was Totald 1588 (100%) 1745 (100%) 1432 (100%) ** about 23%, and that of pulses was 14%. Even lipid was Protein (g=day) primarily of plant-origin (74%). Similarly, iron was also Plants 30.8 (70.3%) 33.0 (61.3%) 28.5 (84.0%) * [Cerealsb] [17.1 (39.0%)] [16.2 (30.1%)] [18.0 (53.1%)] NS predominantly from plants (87%). Urban ± rural compari- [Pulsesc] [6.1 (13.9%)] [5.8 (10.8%)] [6.3 (18.6%)] NS son on protein intake showed that urban people took much Meat 4.6 (10.5%) 8.6 (16.0%) 0.7 (2.1%) ** more (P < 0.01) meat, eggs, milk etc than their rural Fish and shell®sh 1.3 (3.0%) 2.2 (4.1%) 0.4 (1.2%) * counterparts. Interestingly, rural people took less meat Eggs, milk, etc 4.1 (9.4%) 7.0 (13.0%) 1.1 (3.2%) ** protein (P < 0.01) but more meat-based fat (P < 0.01) Totald 43.8 (100%) 53.8 (100%) 33.9 (100%) ** Lipid (g=day) than their urban counterparts, which implies that larger Plants 44.6 (74.1%) 58.3 (79.8%) 31.0 (65.7%) ** intake of meat-based energy by the rural people was most [Cerealsb] [3.6 (6.0%)] [4.6 (6.3%)] [2.5 (5.3%)] * probably due to the consumption of animal fat. The situa- [Pulsesc] [1.3 (2.2%)] [1.9 (2.6%)] [0.7 (1.4%)] * tion was different for Ca intake (Table 6). Whereas the Meat 9.2 (15.3%) 3.9 (5.3%) 14.6 (30.9%) ** Fish and shell®sh 0.2 (0.3%) 0.4 (0.5%) 0.1 (0.2%) NS basic source of Ca in the diet was plant-based foods, which Eggs, milk, etc 3.7 (6.1%) 6.5 (8.9%) 1.0 (2.1%) ** was common to the two groups, the urban people took a Totald 60.2 (100%) 73.2 (100%) 47.2 (100%) ** substantial amount of Ca (about 100 mg=day) from animal- Iron (mg=day) based foods. Study of menu records disclosed that the Plant-based 18.6 (86.5%) 17.8 (83.6%) 19.4 (89.8%) NS major source was cow's milk and its products (data not Animal-based 1.5 (7.0%) 2.6 (12.2%) 0.3 (1.4%) ** Totald 21.5 (100%) 21.3 (100%) 21.6 (100%) shown). Calcium (mg=day) Plant-based 235.2 (77.1%) 259.4 (69.4%) 210.9 (89.3%) * Animal-based 52.8 (17.3%) 100.7 (26.9%) 4.9 (2.1%) ** Analysis of urban ± rural differences on a larger scale Totald 305.0 (100%) 373.8 (100%) 236.2 (100%) NS database Values are AM (percentages). Values in square brackets are as a fraction of In order to examine whether the observation on urban ± plant-based foods. rural differences in Jilin province can be generalized to aMarks are for the statistical signi®cance of the difference between the two other provinces, surveys were done for ®ve urban and rural survey sites by t-test (** and * for P < 0.01 and < 0.05, respectively; NS survey sites (for locations, see Figure 2), ie Jinan and a for P > 0.05). bSum of rice, wheat, maize and foxtail millet. village in Zhangqiu area (Shandong province; Qu et al, cPulses include peanuts. 1997), Xian and Bamiao county, and Xian and Changan dIncluding those of unclassi®able sources. county (Shaanxi province; Nakatsuka et al, 1999) as three

European Journal of Clinical Nutrition Nutrient intakes in China J-B Qu et al 746 Table 6 Urban ± rural comparison on energy and macro-nutrient intakes urban ± rural pairs (Xian thus being counted twice), and also for the three large cities of Beijing, Shanghai and Energy (kcal=day) and nutrient intake (g=day) Nanning (Zhang et al, 1997); the surveys were made Animal Animal following the same protocol as the present one. Demo- Parameter Energy Protein protein Lipid fat Carbohydrate graphic characteristics were similar among the 10 sites (including the two in Jilin; data not shown). An urban ± Urban sites (six sites) Site means rural comparison of the 10 sites in terms of daily intakes of (A) Changchun 1745 53.8 17.8 73.2 12.0 218.2 energy and macro-nutrients (including animal protein and (B) Beijing 1746 63.7 26.4 70.5 32.3 214.3 fat) is summarized in Table 6. (C) Jinan 1968 66.8 26.6 76.1 19.4 254.2 Whereas the variation was large (P < 0.01 by ANOVA) (D) Xian 1824 53.5 12.1 56.3 16.1 276.2 within the six urban sites as well as within the four rural (E) Nanning 1947 61.2 20.4 74.5 23.7 257.8 (F) Shanghai 1635 46.0 18.5 81.0 25.0 181.0 sites, it was also clear that there were substantial urban ± Subtotals rural differences in nutrient intakes. The women in the AM 1811 57.5 20.3 71.9 21.4 223.6 urban sites took signi®cantly (P < 0.01) more animal pro- + ASDa 118 7.1 5.1 7.7 6.6 32.1 tein, lipid and animal fat than the women in the rural sites, ANOVAb ** ** ** ** ** ** although the intakes of energy and protein differed between Rural sites (four sites) them (P < 0.10), and carbohydrate intake was larger in the Site means rural sites than the urban sites (P < 0.05; Table 6). Urban ± (G) Dehui 1432 33.9 2.2 47.2 15.8 217.9 (H) Zhangqiu 2071 61.1 4.5 54.0 8.1 335.1 rural differences were less remarkable in the intakes of (I) Changan 2002 58.2 1.2 26.4 2.1 383.6 vitamins and minerals (data not shown) than in macro- (J) Baoji 1797 50.5 1.7 27.7 2.3 336.9 nutrient intakes (Table 6). Subtotals AM 1826 50.9 2.4 38.8 7.1 318.4 + ASDa 249 10.6 1.3 12.0 5.6 61.2 Discussion ANOVAb ** ** ** ** ** ** Grand totals (10 sites) The present twin surveys in Jilin province made it clear that AM 1817 54.9 13.1 58.7 15.7 267.5 there were substantial differences in energy as well as + ASDa 182 9.2 9.6 18.9 9.4 62.0 nutrient intakes among women in the urban and rural ANOVAc ** ** ** ** ** ** d sites, and that those in the city of Changchun had intakes t-test NS NS ** ** ** * of protein and fat than their counterparts in a farming Changchun and Dehui (both in Jilin province; this study), Jinan and village in Dehui county. Whereas people in both survey Zhangqiu (Shandong province; Qu et al, 1997), Xian and Bamiao, and sites depended primarily on cereals (almost exclusively Xian and Changan (Shaanxi province; Nakatsuka et al, 1999) are pairs of rice) for major sources of energy and protein, the depen- urban and rural survey sites; Beijing, Shanghai and Nanning are cited from Zhang et al (1997). Values are AM, unless otherwise speci®ed. ASDs, up dency was more evident in the latter than in the former to 30% of AM, are not shown to avoid congestion of the table. Marks (Table 5). The urban group had more animal protein than indicate that the values are signi®cantly (** and * for P < 0.01 and 0.05, the rural group, even though the latter had more animal fat respectively; NS for P > 0.10) different between the two survey sites, as than the former (Table 5). assayed by t-test or chi-square test. In a review of dietary surveys in north-eastern China, aAM+ ASD of site means. bP for the difference among six (urban) or four (rural) sites. conducted mostly around 1930, Miura (1972) summarized cP for the difference among a total of 10 sites. that farmers there typically consumed 3500 kcal=day of dP for the difference between six urban and four rural sites. energy during active agricultural seasons with 100 g=day

Figure 2 Locations of the six urban and four rural survey sites for urban-rural comparison. (A) ± (J) Locations of the sites, the names of which are given in Table 6.

European Journal of Clinical Nutrition Nutrient intakes in China J-B Qu et al 747 protein (90 ± 99% from plant foods) and 45 g=day lipid. and milk-products in particular, which is rather exceptional Sorghum (Chinese variety, or kaoliang), maize and millet in China where consumption of dairy products is low in (foxtail millet) were staple cereals, accounting for 85% of general (Ge et al, 1995). the total energy intake, whereas those in cities consumed Analysis of the data from other provinces (Shimbo et al, rice and wheat (¯our) with a better supply of animal-based 1997) showed that substantial calcium intake can be protein. Compared with the ®ndings in Table 5, it is achieved through consumption of non-rice cereals such as apparent that supply of animal-based foods has been wheat and millet, because wheat and millet, for example, improved even in the rural group, and that there have contain almost 10 times as much calcium as rice on a dry been remarkable changes in staple cereals towards rice. It weight basis (Institute of Nutrition and Food Hygiene, should be noted, however, that the political situation then Chinese Academy of Preventive Medicine, 1991a). Con- was quite different and rice consumption was strictly sidering such inter-provincial differences in staple cereals regulated at the time of the surveys which Miura (1972) in combination with low dairy product consumption in the reviewed. The difference in energy intake may be attribu- central part of the continent (ie 36 g=day in cities and table at least in part to heavy physical loads in the past and 4g=day in villages; Ge et al, 1995), calcium supplementa- low bioavailability of energy in some cereals such as tion might be necessary to solve the calcium insuf®ciency sorghum. issue in the regions where people take rice as the principal With one exception of high animal fat consumption in source of daily energy. Dehui, the conclusions drawn from the survey in Jilin Jiang and Wang (1996) estimated from the 1992 (Tables 3 and 5) were also applicable to 10 urban and National Nutrition Survey results that lipid energy ratios rural survey sites (Table 6) in the sense that adult women for urban and rural women were 29% and 19%, respec- (and most probably people in general) in urban areas eat tively. The present ®nding of 71.9 g lipid (647 kcal) in more animal protein and animal fat than those in villages. 1811 kcal energy (or 36% as the lipid energy ratio) in urban The comparison of the results with the observations by Ge women (Table 6) suggests that the problem of high lipid et al, (1995,1996) on the 1992 survey suggests that urban ± intake remains or has been even aggravated in recent years rural differences in intakes of animal protein and animal fat in urban areas. Overweight is the issue emerging rather (therefore animal-based foods in general) still persist in the recently. Popkin (1994) noted that the problems of under late 1990s. and over-nutrition presently coexist in the continent; high Several points of interest in public nutrition are con- income in urban areas is associated with higher consump- sidered to be worthy of discussion in relation to urban ± tion of lipid and obesity, whereas under-nutrition is still an rural differences in nutrient intakes: anemia, calcium issue in less developed rural areas. The present observation insuf®ciency, overweight=obesity and hypertension. that 26 and 19% of women in Changchun and Dehui had Anemia, especially iron-de®ciency anemia, had been con- BMI of > 25 (Table 2) is closely in line with this opinion sidered as a long-standing public health problem, even in on the urban ± rural difference in overweight. Reduced fat large cities, in the past (Zhao, 1992; Li et al, 1993). In the intake is thus recommended to wealthy people in cities present survey in Jilin, however, those with < 12 g (Guo et al, 1999). Hb=100 ml blood were only 5% of the participants, with Because of relatively cold climate in the survey sites no signi®cant urban ± rural difference (Table 2), which (with average temperature of 7 16.7C in January and suggests that anemia is no longer a major public health 22.9C in July in Changchun; National Astronomical problem, at least in the sites surveyed. The absence of Observatory, 1999), hypertension could be a public health severely anemic cases may be related to the fact that iron concern. The prevalence observed in the present survey was intake was suf®cient in most (75%) of the people (Table 4), 12%, and there were no critically ill cases (Table 2). Such a even though a majority of the iron was from plants and not rather mild clinical picture might be associated with low from animal-based foods (Table 6). It is also known that NaCl consumption with an average intake of 7.3 g=day calculation of iron intake by use of food composition tables (Table 3), a value well below the RDA of 10 g=day in (Institute of Nutrition and Food Hygiene, Chinese Acad- China (Chen and Xu, 1996), although Tian et al (1996) emy of Preventive Medicine, 1991a) may lead to substan- reported much higher salt intake of 15 (in urban areas) tial over-estimation (Zhang et al, 1999). General to 16 g=day (in rural areas) in Tianjin in north China. In improvement in nutrition (Ge et al, 1995) may also have this connection, it may be worth adding that the table- affected the statistic favorably. It is of interest to note that based estimation of Na intake agrees fairly well with the the role of pulses as a source of protein was 21% (thus only measurement by ICP-MS (Shimbo et al, 1999). one-third of that of cereals), even among farmers, and their role as a lipid source was even less (Table 5). In contrast to the case of iron intake, calcium intake was quite insuf®cient both in Changchun and Dehui (Table 2). The RDA of 800 mg Ca=day for adult Chinese women References (Table 1; Institute of Nutrition and Food Hygiene, Chinese Acheson KJ, Campbell IT, Edholm OG, Miller DS & Stock MJ (1980): Academy of Preventive Medicine, 1991b) might be too The measurement of food and energy intake in man Ð an evaluation of demanding when compared with the RDA of 600 mg some techniques. Am. J. Clin. Nutr. 33, 1147 ± 1154. Chen JD & Xu H (1996): Historical development of Chinese dietary Ca=day set for adult Japanese women (Committee on patterns and nutrition from the ancient to the modern society. Wld Rev. Health and Nutrition, Japan, 1999). The current calcium Nutr. Diet 79, 133 ± 153. intake both in Changchun and Dehui is, however, far below Committee on Health and Nutrition, Japan (1999): Recommended Dietary any of the RDA values. With regard to the major calcium Allowances for the Japanese, 6th edn. Tokyo: Dai-ichi Shuppan (in sources, it was obtained primarily from plant-based foods Japanese). Ge K-Y, Zhai F-Y, Yan H-C, Cheng L, Wang Q & Jia F-M (1995): The (Table 6). It is worth noting that Changchun women in dietary and nutritional status of Chinese populations in 1990s. Acta addition took calcium from animal-based foods, from milk Nutrimenta Sinica 17, 123 ± 134.

European Journal of Clinical Nutrition Nutrient intakes in China J-B Qu et al 748 Ge K-Y, Zhai F-Y & Yan H-C (1996): The Dietary and Nutritional Status Qu J-B, Zhang Z-W, Xu F-G, Song L-H, Wang J-J, Shimbo S, Watanabe of Chinese Populations Ð 1992 National Nutrition Survey, Vol 1. T, Nakatsuka H, Higashikawa K & Ikeda M (1997): Urban ± rural Beijing: People's Medical Publishing House (in Chinese with English comparison of nutrient intake by adult women in Shandong Province, translation). China. Tohoku J. Exp. Med. 183, 21 ± 36. Guo X-G, Popkin BM, Mroz TA & Zhai F-Y (1999): Food price policy can Shimbo S, Kawamura S, Yamamoto K, Kimura K, Imai Y, Yasumoto M, favorably alter macronutrient intake in China. J. Nutr. 129, 994 ± 1001. Watanabe T, Iwami O & Ikeda M (1994): Reduced carbohydrate intake Ikeda M, Watanabe T, Kasahara M & Nakatsuka H (1988): Nutrient intake in past 10 years in two rural areas in Japan. Ecol. Food Nutr. 33, 123 ± of women in rural and urban areas in Japan. Asia-Paci®c J. Publ. Health 130. 2, 28 ± 32. Shimbo S, Hayase A, Murakami M, Hatai I, Higashikawa K, Moon C-S, Ikeda M, Zhang Z-W, Moon C-S, Imai Y, Watanabe T, Shimbo S, Ma W- Zhang Z-W, Watanabe T, Iguchi H & Ikeda M (1996): Use of food C, Lee C-C & Guo Y-LL (1996): Background exposure of general composition database to estimate daily dietary intake of nutrient or trace population to cadmium and lead in Tainan city, Taiwan. Arch. Environ. elements in Japan with reference to its limitation. Food Add. Contam. Contam. Toxicol. 30, 121 ± 126. 13, 775 ± 786. Institute of Nutrition and Food Hygiene, Chinese Academy of Preventive Shimbo S, Zhang Z-W, Qu J-B, Xu G-F, Song L-H, Wang J-J, Watanabe Medicine (1991a): Food Composition Tables (Nation-wide values, Part T, Nakatsuka H, Higashikawa K & Ikeda M (1997): Urban ± rural 1). Beijing: People's Hygiene Press (in Chinese). difference in cereal consumption in Shandong Province, China. Institute of Nutrition and Food Hygiene, Chinese Academy of Preventive Tohoku J. Exp. Med. 183, 211 ± 220. Medicine (1991b): Food Composition Tables; Appendix II. Beijing: Shimbo S, Zhang Z-W, Miyake K, Watanabe T, Nakatsuka H, Matsuda- People's Hygiene Press (in Chinese). Inoguchi N, Moon C-S, Higashikawa K & Ikeda M (1999): Estimates of Jiang J-P & Wang Y-M (1996): Status, prospect and policy of food mineral intakes using food composition tables vs measures by induc- development in China. Biomed. Environ. Sci. 9, 149 ± 163. tively-coupled plasma mass spectrometry: part 2; sodium, potassium, Li R-W, Chen X-C, Yan H-C, Deurenberg P, Garby L & Hautvast JGAJ magnesium, copper and zinc. Eur. J. Clin. Nutr. 53, 233 ± 238. (1993): Prevalence and type of anemia in female cotton mill workers in TianH-G,NanY,HuG,DongQ-N,YangX-L,PietinenP&NissinenA(1995): Beijing, China. Br. J. Nutr. 70, 787 ± 796. Dietary survey in a Chinese population. Eur. J. Clin. Nutr. 49, 25 ± 32. Miura, U (1972): Diet and nutrition of the Chinese in Manchuria. Bull. Tian H-G, Hu G, Dong Q-N, Yang X-L, Pietinen P & Nissinen A (1996): Koshien Univ. 2, 13 ± 42 (in Japanese with Englsh abstract). Dietary sodium and potassium, socioeconomic status and blood pressure Moon C-M, Zhang Z-W, Imai Y, Shimbo S, Watanabe T, Moon D-H, Lee in a Chinese population. Appetite 26, 235 ± 246. B-K, Lee S-H & Ikeda M (1997): Nutritional status of women in urban Watanabe T, Zhang Z-W, Qu J-B, Xu G-F, Song L-H, Wang J-J, Shimbo and rural areas in Korea as assayed by total food duplicate method. S, Nakatsuka H, Higashikawa K & Ikeda M (1998): Urban ± rural Tohoku J. Exp. Med. 181, 245 ± 265. comparison on cadmium exposure among general populations in Shan- Nakatsuka H, Kasahara M, Watanabe T, Hisamichi S, Shimizu H, Fujisaku dong Province, China. Sci. Total Environ. 217,1±8. S, Ichinowatari Y, Ida Y, Suda M, Kato K & Ikeda M (1988): Urban ± Zhang Z-W, Moon C-S, Qu J-B, Shimbo S, Watanabe T, Nakatsuka H, Yin rural differences in food habits in north-eastern Japan. Ecol. Food Nutr. S-A, Hu M-L, Chen Z-Q & Ikeda M (1997): Nutritional evaluation of 21, 77 ± 87. women in urban areas in continental China. Tohoku J. Exp. Med. 182, Nakatsuka H, Zhang Z-W, Qu J-B, Gao W-P, Deng Y-J, Shimbo S, 41 ± 59. Watanabe T, Matsuda-Inoguchi N, Higashikawa K & Ikeda M (1999): Zhang Z-W, Shimbo S, Miyake K, Watanabe T, Nakatsuka H, Matsuda- Urban ± rural comparison of nutrient intake by adult women in Shaanxi Inoguchi N, Moon C-S, Higashikawa K & Ikeda M (1999): Estimates of Province, China. Biomed. Environ. Sci. 12: 270 ± 284. mineral intakes using food composition tables vs measures by induc- National Astronomical Observatory (1999): The Almanac. Tokyo: tively-coupled plasma mass spectrometry: part 1; calcium, phosphorus Maruzen (in Japanese). and iron. Eur. J. Clin. Nutr. 53, 226 ± 232. Popkin BM (1994): The nutrition transition in low-income countries: an Zhao X-H (1992): Nutritional situation of Beijing residents. Southeast emerging crisis. Nutr. Rev. 52, 285 ± 298. Asian J. Trop. Med. Publ. Health 23(Suppl 3), 65 ± 68.

European Journal of Clinical Nutrition