National income, which enables people to purchase animal proteins, is regarded as a key determinant of adult heights   . Japan and South Korea made rapid economic progress since the end of WW II, with S. Korea some two decades behind Japan due to the Korean War (1950-53) which devastated the country. Per capita GDP (in 2010 U.S. Dollars) in Japan increased from $8608 in 1960, to $25,489 in 1980, and eventually to $42,170 in 2000. That in S. Korea also increased steadily, but remained considerably behind Japan, at $944, $3700, and $15,104 over the corresponding period  . Accordingly, per capita daily caloric intake from animal products increased from 251 kcal in 1961 to 537 kcal in 1980 and 604 kcal in 2000 in Japan, whereas that in S. Korea also increased from 56 kcal, to 212 kcal, and eventually to 447 kcal over the same period  . As is shown in greater details in Table 2 and Table 3, Japan’s per capita intake of animal products was nearly twice as large as that in S. Korea in 1990, for example.
Based on national school health examinations conducted with similar methods both in Japan and S. Korea   , school boys at age 17 - 18 years in Japan were 166 cm in mean height in the early 1960s, 2 - 3 cm taller than their Korean peers, 170 cm in the 1980s, 1 cm taller than in Korea, but ceased to grow taller in height around 1990. Their Korean pees kept growing taller in height, overtaking Japanese boys by 3 cm in the mid-2000s, as shown in greater detail in Table 1.
Intakes of animal proteins alone likely cannot explain the differences in child height development between the two countries over the past half century from 1960 to 2010. A genetic difference between Japanese and Korean population, however, is not easy to establish. The observed differences in child height development between Japan and S. Korea do not appear to be greater than those observed between different regions in the same country   . Also,  “Growth status of Korean schoolchildren in Japan”, states that Korean children born and raised in Japan are very similar, in height and weight by age, to Japanese children, based on school health surveys conducted in Tokyo and Osaka in the end of the 1970s and found taller in height and heavier in weight than their counterparts in S. Korea at the same time.
2.1. Child Height in Japan and South Korea
School Health Examination Statistical Surveys   1, which provide stature by age from 1st graders in elementary school (6 years old2) to senior students in high school (17 years old2) have been conducted by the Ministry of Education every year since 1900 in Japan, except for a few years during WW II. Similar government surveys have been conducted in S. Korea since 1960. In view of the fact that the ratios of girls enrolled in high school for higher education in the 1960s and 1970s were substantially lower in S. Korea than in Japan  , there is a possibility that statistics pertaining to mean height of girls in high schools in
(a)Notes: 1965 = average (1964: 1966), for example. Sources:  .
(b)Sources:  .
Table 1. Secular changes in mean height of Jp. and Kr. school boys by age, from 1960 to 2010.
Sources: compiled by the author.
the earlier period of our investigation may not represent the entire population. That is the reason why this paper focuses on school boys for the comparative analyses of child height between Japan and S. Korea.
The school health surveys in both countries do not cover infants from 1 to 5 years of age and near-adults from 18 to 20 years of age. In the arena of human biology, the first two years of life, or 1000 days, including pregnancy, are crucial periods for future adult height     . Analyzing the secular changes in child height in Japan and S. Korea since 1965,  states, “Most of the height increment seen in adults had already accrued by age 1.5 years” (p. 12). Thus, the data used here which lack the height development for infants from 1 to 5 years may carry some limitations. On the other hand, the school health surveys in both countries are based on very large nationwide samples conducted every year from 1960 to 2015.
Statistics pertaining to mean height by age groups fluctuate from year to year, particularly in S. Korea, possibly because statistics published in the survey reports depict mean height of school children only in Seoul in some years and averages of a few major districts in other years and also the months when the surveys were conducted may have varied from year to year. In order to smooth the annual fluctuations, we employed 3 year moving averages for all age groups. Table 1 provides secular changes in mean height of school boys in Japan by age (from 6 to 17 years old) in Japan and in S. Korea, by 5 year intervals from 1960 to 2010. With possible statistical errors disregarded, at age 16 - 18 years, school boys in Japan were 3 cm taller than S. Korean boys in the 1960s and still slightly, say, 1 cm taller in the 1980s. Japanese boys ceased to grow any taller in the early 1990s and onward, whereas S. Korean boys kept growing in height onward to overtake their Japanese peers by 3 cm in the mid-2000s and ceased to grow any taller since then.
Based on the statistics pertaining to mean height of children from 1 to 20 years of age, mentioned above, Mori presumed that S. Korean children, in accordance with their ethnic traits, tend to grow faster during their late adolescent years than their Japanese peers   . Cole and Mori, however, based on the same data, analyzed by SITAR, concluded that most of the increment in height seen in adults had already accrued by age 1.5 years, either in Japan or S. Korea, as mentioned above. To determine a child growth curve, it is preferable to take the cohort aspects into consideration. A child born in 1960, for example, aged to 11 years old in 1971 and 20 years old in 1980. Using these birth cohorts, the magnitude of growth in height from 6 to 17 years has been nearly constant at 55 cm, almost the same for both Japanese and S. Korean school boys, over the 1960 to 2010 (Figure 1).
To reprise Cole and Mori, most of the height increment seen in senior boys in high school had already accrued by 1st graders in elementary school, age 6 years.
Based on the data provided in Table 1, mean heights of male 1st graders in elementary school in Japan and S. Korea from 1960 to 2010 are compared, to prepare Figure 2, which clearly demonstrates that at age 6 years, Japanese boys
Figure 1. Changes in height growth velocity from 6 to 17 yr old, by cohorts, Jp. and Kr. school boys, from 1960 to 2010.
Figure 2. Changes in mean height of 1st graders in elementary school, Jp. and Kr. boys, 1960 to 2000.
were 112 cm in 1960 and grew steadily to 115 cm in 1975, 1 cm taller than Korean boys, and then grew more slowly to 116.8 cm in 1990 and leveled off afterward, whereas S. Korean boys grew more sharply from 111 cm in 1960 to 117.7 cm, 1 cm taller than Japanese boys in 1990, and still kept growing afterward to 121.8 cm, 5 cm taller than Japanese boys in 2010.
2.2. Food Consumption in Japan and South Korea
This article will try to explore underlying environmental factors, particularly food supply, (“inputs to health”,  ) which may have given rise to distinct differences in secular changes in child height development between Japan and S. Korea since 1960 presented above.
Culturally, Japanese and Koreans share basically the same dietary patterns, i.e., shushoku (staple foods) and fukushoku (side dishes). Traditionally, they take principal energy from rice, noodles and/or bread with miso-soup and a small amount of okazu (side dishes), composed of fish/meats mixed with large amounts of vegetables in their diet. As their standards of living improved, they first increased both shushoku and fukushokuand then gradually increased fukushoku at the sacrifice of shushoku, rice in particular. This tendency of dietary changes is more pronounced in Japan than in S. Korea: more concretely, per capita net supply of meat + eggs and fish was 63.9 and 71.1 kg/year, respectively in Japan, as compared to 47.6 and 50.1 kg/year in S. Korea in 1995, for example, whereas that of cereals was 122.9 kg/year in Japan, as compared to 168.5 kg/year in S. Koreain the same year (Table 3). In term of daily individual caloric intakes, S. Korea’s have been a few hundred kcal/day greater since the mid-1970s than Japan’s (Table 2). Traditionally, Asians do not drink much milk, as compared to
Table 2. Changes in per capita caloric intakes, from vegetal and animal products: Japan and South Korea, 1961 to 2010.
Sources:  .
most Europeans. However, Japanese consumed (in terms of net supply) 86.2 kg/year of milk in 1995, substantially more than Koreans who consumed 48.8 kg/year of milk3 in the same year (Table 3).
On the other hand, S. Koreans have consumed substantially more vegetables than Japanese since the mid-1970s, i.e., per capita net supply of vegetables in S. Korea was 222.3 kg/year, as compared to 116.6 kg/year in Japan in 1995, for example. People in Korea did not eat much fruits many years ago. Per capita net supply of fruit in S. Korea was 5.2 and 9.8 kg/year, respectively in 1961 and 1965, as compared to 29.7 and 39.0 kg/year, respectively in Japan in the same years. Per capita consumption (=net supply) of fruit in S. Korea steadily increased to 14.6 kg/year in 1975 and to 69.6 kg/year in 1995, whereas that in Japan hit the peak in 1975 at 61.9 kg/year and then gradually declined to 53.2 kg/year in 1995 (Table 3). As will be discussed in detail in the subsequent section, children and young adults in Japan began to reduce their fruit consumption radically in the mid-1970s. Young people in Japan also reduced their consumption of vegetables, not to the extent of fruit, though.
Comprehensive surveys of daily nutritional intakes, by major food groups, both in terms of kcal/capita and grams/capita have been conducted since the end of WW II in Japan, whereas a national health and nutrition survey  was first conducted in 1998 in S. Korea, followed by the second one in 2001, and the third one in 2005. In comprehending trends and differences in food consumption between Japan and S. Korea since 1960 onward, official nutrition surveys cannot meet our needs for this paper. This is the reason why we relied on  for the basic information concerning changes and differences in food consumption between Japan and S. Korea over the past half century, as presented in the preceding section. Regrettably, however, we used national per capita net food supply (as proxy for consumption) for the entire population, not for the growing children, in particular.
Japan’s Nutrition Surveys  started to publish per capita food intakes by age groups in 1995 for the first time, while we may need the age-wide data for the period three decades up to 1995.
Japan’s Bureau of Statistics has conducted household income and expenditure surveys  of approximately 8000 households across the nation, with 1/6 of the subjects replaced each month every year since 1950.  publishes the survey results by goods and services, such as rice, --, beef, pork, ---, fresh fruit, coffee, --, gasoline, eating-out, including expenditures for school lunch, at sushi restaurants, etc. in major cities across the country and by income classes of households. Starting in 1971,  published expenditures on major goods and services
Table 3. Changes in per capita net supply of selcetd product groups in Japan and S. Korea, 1961 to 2010.
Sources:  . Note: *Derived by total domestic supply/population, provided in  .
(prices paid and quantities, in the case of goods) by age groups of household head (HH). In the case of foods, eating-out has been on the increase lately and thus  does not cover most of the actual consumption of selected food items. In the case of meat, for example, FIES is estimated to cover about half of total meat consumption  , 60% - 80% of fresh fruit  , and 50% - 60% of vegetables  .
Focusing just on at-home consumption of selected food products, individual consumption by age groups can be derived from  data classified by HH age groups, by means of the TMI model which explicitly incorporates family age composition by HH age groups  .
M. Blum states, “The central determinants of human stature is the protein-intensity of a nation’s diet. Populations consuming larger amounts of animal protein reach higher average height levels than countries with less animal protein consumption. However, a high consumption of animal proteins alone does not result in increasing body heights if the overall consumption of calories and other essential nutrients is insufficient”  , p. 21). The author is inclined to contend that this may apply to the comparative case of child height developments between Japan and S. Korea over the past half century since the early 1960s. Ishibashi, based on the comprehensive analyses of large samples of  panel data since the early 1980s, Japanese youth have been consuming more meats than the older adults   . In Japan, vegetables are believed to regulate bowel movements and fruit is called mizikashi (jelly desert), implying fruit consumption has nothing to do with essential nutrients. The author, however, suspects that children in Japan consumed insufficient “essential nutrients” noted by M. Blum, due to the drastically reduced consumption of fruit and vegetables4 in the recent few decades.
 , Ministry of Agriculture drew public attention to the widespread tendencies of wakamono no kudamono-banare (steering away from fruit by the young), in the 1980s and the early 1990s by displaying household fresh fruit purchases classified by HH age groups divided simply by the number of persons contained in each group.  designed a quadratic programming model to derive individual at-home consumption of selected food products by age from FIES, classified by HH age groups, which was statistically refined by  .
Table 4 and Table 5 demonstrate that Japanese youth used to eat as much fruit and vegetables as older adults in the early 1970s but began to reduce their consumption of these products in the early 1980s, while the older adults kept constant their produce consumption at high levels: for example, nearly three times higher than young people’s consumption in the case of vegetables and more than five times in the case of fruit in 2000.
Table 4. Changes in per capita at-home consumption of fresh fruit by age groups, 1971 to 2010 in Japan.
Sources: derived from  by the authors, using the TMI model. Notes: Estimated by 5 year age intervals first, which were simply averaged into 10 year intervals.
Table 5. Changes in per capita at-home consumption of fresh vegetables by age groups, 1971 to 2010 in Japan.
4. Conclusions and Future Research Recommendations
At age 17 - 18 years, Japanese school boys were 2 - 3 cm taller in mean height in the 1960s through 1970s, barely 1 cm taller in the 1980s than their Korean peers and they ceased to grow any taller in and around 1990, whereas Korean boys kept steadily growing in height to overtake their Japanese peers by 3 cm in the mid-2000s and ceased to grow any taller since then.
After the end of WW II, Japan and South Korea made rapid and steady economic progress, with S. Korea some two decades behind Japan due to the Korean War (1950-53). As the economy grew, food consumption remarkably improved both in quantity and quality. Particularly, intake of animal-sourced products increased conspicuously in the two countries. Yet, animal product supply per person in Japan was considerably greater, as much as 30% higher, than S. Korea in the mid-1990s.
Increases in animal protein products alone cannot explain the differences in growth patterns of child height between the two countries. However, S. Koreans have been eating twice as many vegetables as Japanese on average and their per capita fruit consumption soared very rapidly since the mid-1970s, surpassing that of Japanese in the mid-1990s. Meanwhile, Japanese youth reduced their fruit and vegetable consumption from the mid-1970s on. This “steering away from fruit and vegetables” by Japanese youth could have contributed to the stagnation of child height growth since the early 1990s, and deserves scientific investigation.
The author is grateful to John Dyck, formerly an economist in Asian affairs at ERS, USDA, who advised to rearrange the structure of the article and thoroughly edited the manuscript. Thanks go to anonymous reviewers for considerate comments and careful proofs.
No research fund has been received from any sources.
Table A1. Per capita daily intakes of fruits and vegetables by age groups in South Korea, 1998 and 2001.
1In our previous studies, the data provided in  were used, which furnish child height by age from 1 to 20 years, in 1965, 1975, 1984, 1997, and 2005 in S. Korea. Accordingly, the similar data in the same years, 1965, 1975, 1984, 1997, and 2005, based on  , were employed.
2Depending on the month when the school surveys conducted, actual age of 1st graders in elementary school was 6 years + some months old.
3Per capita net supply (kg/year) of milk, excluding butter, for Republic of Korea seems to be extraordinarily under-calculated, particularly for the period after the mid-1980s, as compared to the estimates provided  .The author calculated per capita supply of milk, excluding butter in kg/year by dividing total domestic supply as providedin  in 1,000 tons/year by total populationavailable also from  .
4A few recent empirical studies report that consumption of fruit and vegetables contributes to greater bone mineral accrual/density in post-menopausal females   and also in young and adolescent children     . As shown in Appendix Table 1, S. Korean youth in their 20s and 30s consume as much vegetables and fruit as the older adults in the end of 1990s and the early 2000s  . In the presence of robust cohort effects in food consumption   , S. Korean children are supposed to consume as much vegetables and fruit as adults in the 1980s and 1990s.
 Kim, J.-Y., Choi, J.-M., Jin-Soo, M., Shin, S.-H., et al. (2008) Anthropometric Changes in Children and Adolescents from 1965 to 2005 in Korea. American Journal of Physical Anthropology, 136, 230-236.
 Tanaka, M., Mori, H. and Inaba, T. (2004) Re-Estimating per Capita Individual Consumption by Age from Household Data. Japanese Journal of Rural Economics, 6, 20-30.
 Ishibashi, K. (2006) Statistical Investigations of Structural Changes in Household Consumption: Household Categories and Age Factors. National Research Center for Agricultural Research, Tsukuba, 57. (In Japanese)
 Sugiura, M., Nakamura, M., Ogawa, K., Ikoma, Y. and Yano, M. (2012) High Serum Carotenoids Associated with Lower Risk for Bone Loss and Osteoporosis in Post-Menopausal Japanese Female Subjects: Prospective Cohort Study. PLoS ONE, 7, 1-9.
 Nakamura, M., Sugiura, M., et al. (2016) Serum β-Carotene Derived from Satsuma Mandarin and Brachial-Ankle Pulse Wave Velocity: The Mikkabi Cohort Study. Nutrition, Metabolism & Cardiovascular Diseases, 26, 808-814.
 McGartland, C.P., Robson, P.J., et al. (2004) Fruit and Vegetable Consumption and Bone Mineral Density: Northern Ireland Young Hearts Project. The American Journal of Clinical Nutrition, 80, 1019-1023. https://doi.org/10.1093/ajcn/80.4.1019
 Vatanparast, H., Baxter-Jones, A., Faulkner, R.A., Bailey, D.A. and Whiting, S.J. (2005) Positive Effect of Vegetable and Fruit Consumption and Calcium Intake on Bone Mineral Accrual in Boys during Growth from Childhood to Adolescence: The University of Saskatchewan Pediatric Bone Mineral Accrual Study. The American Journal of Clinical Nutrition, 82, 700-706.
 Prynne, C.J., Mishra, G.D., et al. (2006) Fruit and Vegetable Intakes and Bone Mineral Statues: A Cross Sectional Study in 5 Age and Sex Cohorts. The American Journal of Clinical Nutrition, 83, 1420-1428. https://doi.org/10.1093/ajcn/83.6.1420
 Li, J.-J., Huang, Z.-W., et al. (2012) Fruit and Vegetable Intake and Bone Mass in Chinese Adolescents, Young and Postmenopausal Women. Public Health Nutrition, 16, 78-86.
 Mori, H. and Saegusa, Y. (2010) Cohort Effects in Food Consumption: What They Are and How They Are Formed. Evolutionary and Institutional Economics Review, 7, 43-63.
 Mori, H., Inaba, T. and Dyck, J. (2016) Accounting for Structural Changes in Demand for Foods in the Presence of Age and Cohort Effects: The Case of Fresh Fish in Japan. Evolutionary and Institutional Economics Review, 13, 363-379.