Asia Pacific J Clin Nutr (1997) 6(3): 175-179

Body mass index of young Chinese adults

Keyou Ge

Institute of Nutrition and Food Hygiene, Chinese Academy of Preventive Medicine, Beijing, China

Body mass index of young Chinese adults aged 20 to 45 years was assessed in relation to household income, energy and macronutrient consumption. The data were collected in the 1992 China National Nutrition Survey, with a sample of 8477 urban and 20911 rural subjects from provinces, autonomous regions and municipalities.

The mean values of BMI were 21.5 and 21.9 for urban males and females, and 21.1 and 21.5 for rural males and females, respectively. Inhabitants of the three Chinese municipalities (Beijing, Tianjin and Shanghai) presented a higher BMI value than the national averages.

Undernourished people (BMI<18.5) accounted for 9.0% in the urban and 8.0% in the rural populations, and the overweight (BMI³ 25) sectors were 14.9% and 8.4% respectively.

In rural populations, the proportion undernourished is inversely related the average energy intake, and dietary fat intake is positively related to the proportion overweight. The change in recent years of BMI of Chinese adults is considered in relation to food consumption.

Key words: BMI (body mass index), China (Beijing, Tianjin and Shanghai) National Nutrition Survey of 1992, rural, urban, energy intake, fat intake, undernourished, overweight

Introduction

Anthropometric data of adults have been routinely collected in many surveys, but using these data to identify the undernourished groups has achieved limited results because of a lack of satisfactory indicators. Considerable efforts have been devoted to explore the nutritional significance of various weight/ height indices; body mass index (BMI) has become the focus of attention in recent years.

The report of a working party of the International Dietary Energy Consultative Group defined chronic energy deficiency (CED) using BMI for adults. It proposed 18.5 as the cut-off point for CED identification¹. Ferro-Luzzi et al examined three populations in Africa and Asia and determined that the use of BMI alone produced coherent data as BMI distributions changed, with only 5% of the population wrongly classified as being malnourished².

The optimum range of BMI was previously regarded as 18.7 to 23.8 for women and 20.1 to 25.0 for men³. Later, it was simplified to 20 to 25 as the acceptable range for both men and women in relation to body fatness⁴. A WHO study group, adopted cut off points of 18.5 for CED and 25 1000 for obesity⁵. The BMI of Chinese adults has been considered in connection with CED only recently. A nationwide sample of 13,300 subjects aged 20-45 years surveyed in 1982⁶ determined that 12.3% of them were below 18.5 kg/m². A sample of 5162 subjects selected from 8 provinces in 1989 showed a decline of the undernourished proportion (8.47%), and even less in the high income group⁷.

A recent national nutrition survey was carried out in China in 1992. It collected a wide range of information at individual, household and community levels. The present paper is based on the 1992 survey data, and the discussion focussed on the distribution of BMI in young adults in connection with their dietary and economic conditions so that the role of BMI as an indicator for assessing nutritional status could be evaluated.

Method

The 1992 China national nutrition survey consisted of a dietary survey, anthropometry measurements, haemoglobin assay and some socio-economic information. The dietary survey covered 25,285 households (8608 urban and 16677 rural) and 99,954 household members of all ages. Body weight and height were taken in 82,538 individuals which accounted for about 82.5% of the sampled subjects.

Sampling

A multistage random cluster process was used to draw the sample. All provinces, autonomous regions and municipalities (Beijing, Tianjin and Shanghai) under the central government were covered in the sample frame. In each province/region/municipality 4-8 units (towns and counties) were selected by the criteria of population size and economic situation. In each selected town or county, two districts or townships were chosen, and in each district or township, two communities or villages were sampled as the primary survey sites. In each survey site, a cluster of 30 households were drawn as the survey subjects. In each of the 3 municipalities, 5 central districts and 3 suburbs were chosen at the first stage, the sub-stages process were the same as in the provinces. Therefore, the sampling design was framed as:

• 30 provinces, autonomous regions, municipalities
• 30 x 8 = 240 towns, counties
• 240 x 2 = 480 districts, townships
• 480 x 2 = 960 communities, villages
• 960 x 30 = 28,800 households

This sample design was reviewed by an expert committee, and recognised as reasonably representative of the whole population.

Dietary data collection

Two sets of questionnaires were designed, for the collection of household food consumption data and for individual food intake data. The household dietary data were collected by recording the inventory change from the beginning to the end of the 3 survey days, in combination with weighing food stuffs used as prepared for meals every day. Household members who participated in each meal were recorded. Household average per capita food consumption was calculated, and household average per reference person food consumption was also calculated taking into consideration age, sex and labour intensity of the household members. The individual intake data were obtained by 3 consecutive 24 hour recalls. Some food stuffs which were used for preparing dishes, such as cooking oil and salt, were assigned to the meal participants in proportion to their energy requirements. During the survey period, qualified interviewers visited the household every day, and the questionnaires (filled in by the interviewer 1000 s) were reviewed by team leaders on the spot. The nutrient intakes were calculated based on the Chinese food composition tables⁸.

Anthropometry measurements

Body weight, height and upper arm circumference were measured by trained personnel with standardised scales and measures. A quality control process was in effect while the data collection was taking place in the field.

Socio-economic and community survey

The household income was reported by the head of the household. The economic categorisation of the households was based only on reported income, family assets were not taken into account. Community data were provided by community leaders and other knowledgeable individuals.

Result

The dietary and nutritional status were considered in relation to the household income tertile of a given population.

Nutrient intake

The national average intakes of energy, protein and fat are shown in Table 1. The overall consumption of energy was 2328 kcal per reference person per day. Urban inhabitants consumed about 100 kcal more than did rural people, on average. This difference mainly resulted from the greater consumption of the higher income groups. The low income inhabitants in both urban and rural areas consumed similar amounts of energy.

Table 1. National average of nutrients intake by income levels (per reference person per day).

1992 China Nutrition Survey

The protein consumption was 68g per day, on average, and increased in line with income in both urban and rural populations, but more obviously among urban groups. The high income group consumed 15g/day more protein than the poor, and urban people consumed 1lg/day more than the rural.

Fat intake was most obviously associated with house-hold income. High income groups consumed twice the amount of the low income groups. The fat intakes of urban inhabitants in each of the three income categories were significantly higher than those of the rural.

Body mass index

In general, the BMI of adults is lower in the younger population and higher in the middle aged populations, but decreases again in the older groups. This is true for both males and females, and for both urban and rural inhabitants in this survey. The discussion here, however, will focus on the segment aged 20 to 45 years.

Mean value of BMI

The mean BMI values of the nationwide sample (except the three municipalities directly under the central government) is shown in Table 2. The males had lower BMIs than females. The differences between males and females are statistically significant (P<0.05) for all the rural matchable pairs of groups, and for the urban overall and the low income groups. The mean BMI values of the high income category are significantly higher in comparison with the low income category in all listed populations except the urban females. Also, the overall means of the urban males and females are higher than the rural males and females respectively.

Table 2. Body mass index of Chinese adults in 1992, nationwide except the 3 cities of Beijing, Tianjin and Shanghai (subjects are 20-45 years).

Figures with the same letter are not statistically different. The BMI means of the inhabitants of the three municipalities, Beijing, Tianjin and Shanghai, are shown in Table 3. The overall trends among the populations and income groups are similar to the nationwide results but somewhat less obvious, particularly in urban populations. The total urban males and the low income urban males were heavier than their rural corresponding groups (p<0.05), but no significant difference was found between other comparable urban and rural pairs. It is interesting that the high income groups had a lower BMI value than the middle income groups and the low income groups in all the given areas and sex groups except the rural males. The urban high income male group presented a significantly lower BMI than the middle income group.

The means of BMI of the three municipalities were apparently higher in comparison with the nationwide results in the matchable groups. The difference is about 1kg/m² on average, and also in most of the comparable pairs.

Table 3. Body mass index of Chinese adults in 1992, in the three cities -- Beijing, Tianjin and Shanghai (subjects are 20-45 years).

Figures with the same letter are not statistically different.

Distribution of BMI

The distribution of BMI amongst Chinese young adults is summarised in Table 4. The cut off points <18.5, 18.5 to 25 and >25 are adopted to classify chronic energy deficiency (CED), normal and overweight respectively. About 73 to 82% of the urban people and 78 to 88% of the rural people were in the normal range. More overweight people were seen in all female groups than in the matchable male groups. The middle income urban male group presented a smaller proportion of CED but a larger proportion of overweight than the low and high income groups. In urban females, the high income group showed a higher percentage of CED, but there were more overweight females in the middle income group. The rural populations showed a better centripetal distribution. Both the CED and overweight portions were smaller than that of the urban populations. Within the rural section, the low income male group and the middle income female group presented a higher percentage of CED than their comparable groups. The proportion of overweight increased with income in both male and female groups.

Table 4. Distribution of BMI of Chinese adults in 1992, nationwide except in the 3 Cities (subjects are 20-45 years).

Figures with the same letter are not statistically different.

Correlation of BMI with nutrients intakes

The means and distribution of BMI by province, autonomous region or municipality are listed in Table 5. The means range from 20.9 to 23.7 for urban, and from 20.1 to 23.9 for rural populations. The underweight subjects (BMI<18.5) accounted for 3.3% to 16.9% of the urban and 2.9% to 22.1% of the rural populations. The overweight proportions were 3.8% to 32.8% of the urban people and 2.2% to 36.8% of the rural people.

Table 5. BMI of young adults of various areas, China 1992 (subjects are 20-45 years).

Table 6 shows the average energy and protein intakes as percentages of the Chinese RDA⁹, and the energy from fat and from animal food as a percentage of the dietary total energy. The correlations between dietary and BMI variables were analysed. The coefficient and probability values are listed in Table 7. Based on this analysis, the mean values of BMI were not correlated with the dietary factors in discussion. The energy intake is inversely correlated with BMI in the undernourished population, and the percentage of energy from animal food is positively related to BMI in the underweight population, and inversely related to BMI in the acceptably nourished rural sector. Fat intake is negatively related to BMI in the acceptably nourished population in both urban and rural populations, and shows a positive correlation with the overweight in rural populations.

Table 6. Dietary intake of nutrients of young adults, China 1992 (subjects of 20-45 years).