Food composition tables are essential to estimating nutrient intake. For the Thai diet, food composition tables have been established in Thailand and also by the Food and Agriculture Organization of the United Nations. However, nutritional evaluations to date have been limited because the number of food items and nutritional components in the tables is restricted. The aim of the present study was to calculate the nutrient intake of young Thais using different food composition tables from Thailand, Japan, and the US and to clarify the influence of these differences on calculated nutrient intake.
The study was conducted by the Chiang Mai Rajabhat University in Chiang Mai, Thailand in August of 2013, 2014 and 2015. A total of 297 female students were asked to complete a Self-Administered Food Frequency Questionnaire (SFFQ) asking about the habitual consumption of 80 kinds of listed foods and eight frequency categories within the past one year. Intakes of energy and nutrients were quantified using Japanese Food Composition Tables (J-FCT), USDA Food Composition Tables (U-FCT), and Thai or ASEAN Food Composition Tables (T-FCT). Mean estimated intakes of energy and nutrients were compared and Spearman’s correlation coefficients between the intakes of each nutrient using the different FCTs were calculated.
Means of almost all nutrients, excluding energy and protein, significantly differed with use of the different FCTs. Fat, vitamin E, and dietary fiber intake were lower when calculated using T-FCT than the other FCTs. Moreover, calcium and iron intake were lower and total energy and carbohydrate intake were higher when calculated using J-FCT than the other FCTs. Never the less, correlation coefficients between nutrients using the different FCTs were significant and more than 0.8 for all nutrients except vitamin E.
Ranking of participants by estimated nutrient intake did not change no matter which FCT was used. This finding suggests that nutrient composition values from other food composition tables can be substituted for missing values in the primary table.
Food composition table, Estimation of nutrient intake, Thailand, Japan, USDA
The reported prevalence of overweight and obesity in Thailand has dramatically increased over the last 30 years, among not only middle aged and older but also young Thais [1,2]. The leading causes are rapid changes in food intake and lifestyle pattern [3]. Thailand’s public health authorities have used dietary guidelines to inform and assist consumers in making healthy nutrition choices and in following healthy lifestyles [4]. Although the Thai Ministry of Public Health conducts a national nutritional survey [5], the intake of some generic food items, such as non-alcoholic beverages, prepared or semi-prepared foods, and snacks is not included [6,7], and the two Food Composition Tables (FCTs) used each list fewer than 800 kinds of food and less than 15 kinds of nutrient.
The Food and Agriculture Organization of the United Nations (FAO) has been working towards improving food composition data quality and availability, including the development and updating of FCTs [8]. However, because nutrient ingredients listed in FCTs are affected by climate and soil, it is inevitable that food ingredient values are different. Similarly, because measurement methods differ among countries, the measured values of a particular nutrient in a particular food are unlikely to have the same degree of accuracy when evaluated in different countries.
However, research in to the relationship between dietary intake and the morbidity or mortality of disease in epidemiological studies using food frequency questionnaires and dietary surveys is frequently done not to estimate absolute values, but rather to rank the correlations of nutrient intake in subjects [9,10]. Therefore, when nutrient intakes are estimated in epidemiological study, FCTs developed by the FAO may not be necessary - rather, if target foods are not listed in the FCT of the study country, it may be possible to use FCTs of other developed countries, such as Japan or the US, by replacing the foods with similar foods. If it is indeed possible to calculate nutrient intake using the composition table of a developed country, it would then be possible to epidemiologically study the association between life style-related chronic diseases and nutrient intake in Thailand.
The aim of this study was to calculate the nutrient intake of young Thais using food composition tables developed in Thailand, Japan, and the US, and to clarify differences in nutrient intakes calculated from them. If correlations of estimated nutrient intakes from these FCTs were high, it would be possible to calculate nutrient intakes for nutrients not listed in the Thailand or ASEAN FCT (T-FCT) [6,7] using the Japan [11] or USDA [12] FCT.
The study was conducted by Chiang Mai Rajabhat University in Thailand in the three respective Augusts of 2013, 2014 and 2015. The survey was conducted using a Self-Administered Food Frequency Questionnaire (Thai_SFFQ) in 334 female students. After excluding 37 participants who failed to answer all questions, 297 participants were included in the analysis. Written informed consent was obtained from all participants before distribution of the Thai_SFFQ and the study protocol was approved by the Institutional Review Board of Otsuma Women’s University, Tokyo (26-001).
The Thai_SFFQ asks about the habitual consumption of listed foods within the past one year. Details of the development of this SFFQ have been reported elsewhere [13]. In brief, 80 food items were included, in the 8 frequency categories of none, less than 1 day per month, 2-3 days per month, 1 day per week, 2-3 days per week, 4-6 days per week, almost daily, and more than twice per day. Portion sizes were described for every food item based on standard or typically presented portion sizes, namely less than half the standard portion size, 20% to 30% less than the standard portion size, same as the standard portion size, 20% to 30% more than the standard portion size, and more than 1.5 times the standard portion size. This Thai_SFFQ was developed in reference to a Japanese SFFQ (JP_SFFQ) which was designed to estimate nutrient intake in young Japanese, and has been validated [14]. The Thai_SFFQ was modified by removing or adding some food items according to data from a dietary record of 56 young Thai [13]. The food items listed in the Thai_SFFQ are described in the Appendix.
Appendix: Food items listed in FFQ. Appendix
Nutrient components listed in either the Thai or ASEAN Food Composition Tables (T-FCT) were selected [6,7]. Weighted average composition tables were developed using T-FCT [7], the Japan standard Food Composition Table (J-FCT) [11], and the USDA Food Composition Table (U-FCT) [12]. Nutrient components listed in the developed composition tables were energy and 14 kinds of nutrients, namely protein, fat, carbohydrate, calcium, phosphorus, iron, vitamin A, niacin, vitamin B1, vitamin B2, vitamin E, vitamin C, total dietary fiber, and salt.
The different FCTs express energy and nutrient intake as the mean, standard deviation, minimum value, and maximum value. Analysis of variance with Tukey - Kramer post-hoc comparison of means was performed to test for differences between the FCTs. Spearman correlation coefficients were calculated between each nutrient intake from the different FCTs. Statistical analyses were performed using SAS statistical software version 9.4 (SAS Institute Inc, Cary, NC, USA).
The mean age of participants was 20.3 ± 1.3 years (range 18-24 years) and the average BMI was 22.6 ± 5.2 kg/m2. The prevalence of overweight classified by the WHO criteria for Asians [15] (BMI > 25) was 23.9%. This prevalence of overweight participants was higher than the 20.6% for age 15-29 years in The Fourth National Health Examination Survey (NHES4) [16] (Table 1).
Table 1: Characteristics of Participants (n = 297). View Table 1
Nutrient intakes calculated using the different FCTs were significantly different for all nutrients except protein. In particular, mean intakes of total fat, vitamin E and dietary fiber intakes were lower, and mean intakes of iron and retinol were higher when calculated by T-FCT. Sodium intake was more than 20 g when calculated by any FCT. Meanwhile, all nutrients were significantly correlated between each FCT and Spearman correlation coefficients were more than 0.8, except for vitamin E (Table 2).
Table 2: Energy and nutrient intake of young Thais (n = 297) estimated from food composition tables developed in Japan, Thai and the USDA. View Table 2
In this study, we calculated the mean intakes of young Thais using a FFQ and different food composition tables from Thailand, Japan and the US. Results showed that the intake of the 13 kinds of nutrients listed in the T-FCT significantly differed by FCT. Nevertheless, intake levels of all nutrients were highly correlated among the different FCTs, and the J-FCT or U-FCT may accordingly be used to calculate nutrient intakes and to classify subjects on the basis of their intake ranking, which is generally assigned by quartile rank in dietary studies in Thailand.
When dietary surveys are conducted in developed countries, it is natural that nutritional calculations should use their own FCT. This is because the FCT is developed using values analyzed from the country’s own foods. In international comparisons of nutrient intake between developed and developing countries, however, when the degree of completion of the FCT of one country is low, it seems preferable to use the FCT of a developed country. Because, it is unnecessary to compare in consideration of the difference of the ingredient in FCT.
A recent study estimated nutrient intake in Thais [17-19]. However, dietary assessment in that study was limited to a single 24-hour dietary recall or non-weighed dietary record, and nutrient intakes and subjects were restricted to certain nutrients and preschool children or sedentary workers in Bangkok. The development and validation of an SFFQ in Thailand has not been reported, and we are therefore unable to compare our estimated nutrient intakes in young Thais with other those of studies. Moreover, nutrient intake was estimated using the SFFQ, which aims to rank subjects by nutrient intake rather than to estimate absolute values in subjects. It is therefore difficult to compare our estimated nutrient intakes with the dietary reference intake in Thailand [4].
Because fatty meat was not listed in the T-FCT, meat intake was calculated using the value for lean meat. Estimated fat intake is accordingly lower using the T-FCT than when using the other FCTs. On the other hand, the T-FCT includes vitamin E only in cooking oil and coffee milk, and estimated intake levels using T-FCT were lower than those using other FCTs. Because the amount of dietary fiber in rice and noodles, which are staple foods in Thailand, is low in the T-FCT, the dietary fiber intake estimated using the T-FCT appears low. Because the amount of carbohydrate in staple rice noodles is high in JP_FCT, the estimated carbohydrate and energy intake using the J-FCT appears to be high. Meanwhile, because the amount of calcium and iron in vegetables is low in the J-FCT, the estimated calcium and iron intake using the J-FCT appears to be low. Salt intake was high no matter which FCT was used, likely because salt is an important ingredient in Nam Pla, a fish sauce which is the most commonly used seasoning in Thailand.
The FFQ we used in this study was developed from the findings of two-day dietary records of young Thais (n = 56). Its validity has been examined by comparing it with other FFQs developed using the Fourth National Health Examination Survey (NHES4) in Thailand. It was modified to incorporate a setting for the portion size of each food, and included photographs of the food examined [13]. However, it appears likely that this FFQ did not allow adequate quantification of the intake of any nutrients. Because this study was conducted in young women, verification in men and in people of other ages is necessary.
Our data indicate that FCTs developed by different countries may be used to calculate nutrient intake and to classify subjects on the basis of their relative ranking, as assigned by quartile rank in Thailand. This finding suggests that nutrient composition values from other food composition tables can be substituted for missing values in the primary table.
We thank the staff and participants for their contribution to this work. This study was supported by the Institute of Human Culture Studies, Otsuma Women’s University (Grant Number K2806). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.