Height and Nutrition

Bernard Harris. Cambridge World History of Food. Editor: Kenneth F Kiple & Kriemhild Conee Ornelas, Volume 2, Cambridge University Press, 2000.

Scientists have long recognized that there is a close association between the height of a population and its nutritional status, and by the end of the nineteenth century governments were coming under increasing pressure to measure their citizens as a means of determining that status. Many have done so in the twentieth century, as height and weight statistics have played an ever increasing role in health assessment programs (Tanner 1981). This chapter employs the data generated thus far to construct a general picture of changes in height and nutritional status in various countries during the course of the twentieth century.

The studies on which this chapter is based have taken a variety of forms. Most of the information about changes in average heights has been derived from school surveys, but a great deal of data on adult heights is also available. Many of these data have been derived from military recruiting records and are primarily concerned with the heights of adult men (Floud, Wachter, and Gregory 1990).

In addition, a number of investigations have focused on changes in the timing of growth spurts and in overall rates of growth. These investigations have tended to focus on the age of peak-height-velocity (PHV) in both males and females and the onset of menarche in girls (e.g., Matsumoto et al. 1980; Matsumoto 1982; Danker-Hopfe 1986). Some observers have also examined changes in the relative dimensions of different parts of the body, such as leg length and sitting height (Himes 1979; Tanner et al. 1982).

The relationship between social and economic changes and changes in height is both simple and complex. On the one hand, there is a clear association between the average height of a population and its standard of living. This is reflected in the unprecedented improvements in average height that have occurred in many parts of the world during the past one hundred years, and in the differences that continue to exist between the average heights of people from different socioeconomic backgrounds (e.g., Rea 1971; Rona, Swan, and Altman 1978; Jones, Nesheim, and Habicht 1985).

On the other hand, it is much more difficult to define the precise significance of different components of the “standard of living.” Most observers, for example, would agree that the two most important environmental determinants of height are diet and disease, but it has proven extremely difficult to establish their relative importance (Roche 1979). Some have suggested that “health care may be more important than nutrition in the first year of life” and that “nutrition may be more important than health care” after the first year, but the evidence on which these suggestions are based is far from conclusive (Floud et al. 1990: 249).

Another area of difficulty concerns the relationship between environmental and genetic influences. There are evident differences among the heights and rates of growth of different ethnic groups, but the extent to which these differences are genetic or environmental in origin is not clear (Eveleth 1979: 388; Eveleth and Tanner 1991). However, it is reasonable to argue that changes in the average height of members of the same ethnic group over time reflect environmental and nutritional changes rather than changes in genetic potential (Floud et al. 1990). We can also seek to isolate the genetic determinants of growth by comparing the heights and weights of members of the same ethnic group in different circumstances and in different parts of the world (e.g., Moore 1970; Greulich 1976; Bindon and Zansky 1986).

It is important to remember that changes in social and economic conditions can affect heights and rates of growth in a number of different ways. In general, undernourished children who are subjected to repeated bouts of infection grow more slowly than children who are well nourished and live in a largely disease-free environment. Undernourished children experience the adolescent growth spurt at a later age, and although they continue to grow for a longer period, their final height is shorter. These differences mean that the gap between the heights of the two populations is at its greatest when the more favored children reach adolescence. The increase in the average value of children’s heights has, therefore, been considerably greater than the increase in the average heights of adults (Floud et al. 1990; Eveleth and Tanner 1991).

The structure of this chapter follows that laid down in P. Eveleth and James Tanner’s invaluable surveys of Worldwide Variation in Human Growth (1976, 1991). The first two parts examine the main trends in the average height of Europeans in Europe and the heights of Europeans outside Europe. This is followed by sections on the heights of Africans in Africa and the Americas; the heights of Asiatics in Asia and the Americas; the heights of Indo-Mediterraneans in the Near East; the heights of peoples in North Africa and India; and finally the heights of Australian Aborigines and Pacific Island peoples. The last section summarizes the evidence and considers some of the implications of that evidence for future health policy.

Europeans in Europe

Europeans living in Europe have been the subject of a particularly large number of anthropometric surveys. Many of the earlier investigations were concerned with the heights of adult male army recruits, but the present century has seen a growing concentration on the heights of schoolchildren. We also possess a considerable amount of data on the age at which European girls reach menarche, and there is some information about changes in European birth weights. The long history of anthropometric monitoring in Europe, combined with the availability of large amounts of data, has encouraged a number of investigators to use the heights of Europeans as “standards” against which to assess the growth of other population groups (e.g., Eveleth and Tanner 1976, 1991).

Studies of European heights provide some of the clearest evidence of the increase in average height that has occurred in most parts of the world during the course of the twentieth century. In 1991, for example, Eveleth and Tanner reported evidence of a secular trend toward increased adult height in

Table VI.5.1 Changes in the heights of Eupropean army recruits circa 1900-1975 Source: Floud (1984).

Change in
Country Period height (cm)
Belgium 1920/4-69 8.1
Denmark 1904/5-66/75 8.4
France 1900-60 7.5
Germany (Bavaria) 1900-58 4.2
Italy 1900-52 3.6
Netherlands 1907-75 11.1
Norway 1900-60 6.7
Spain 1903/6-55 3.1
Sweden 1900-74 8.1
Switzerland 1908/10-57 6.2

Sweden, Finland, Norway, France, the United Kingdom, Italy, Germany, Czechoslovakia, Poland, Hungary, the Soviet Union, the Netherlands, Belgium, Switzerland, and Austria (Eveleth and Tanner 1991). Similarly, in 1984 Roderick Floud showed that the average height of army recruits in 10 European countries increased by between 3.1 and 11.1 centimeters (cm) during the years 1900 and 1975 (Table VI.5.1).

Some investigators have attempted to infer changes in average height by comparing the average height of men and women of different ages at the same point in time.Thus in 1971 Erik Bjelke demonstrated that Norwegians who were born between 1920 and 1929 were between 1.7 and 1.8 centimeters taller than Norwegians who had been born a decade earlier (Bjelke 1971). In 1988, Mary Walker, A. Shaper, and G. Wannamethee compared the heights of 7,735 British men who were born between 1919 and 1939 and concluded that the average heights of successive birth cohorts had risen by approximately 1.3 centimeters per decade (Walker, Shaper, and Wannamethee 1988).

There have also been a large number of investigations into the heights of children.The average height of 5-year-old boys in London increased by 7.4 cm between 1905-12 and 1966, whereas the average height of 5-year-old girls increased by 7.5 cm (Cameron 1979). Even more dramatic was the average height increase of 12-year-olds in Oslo, which was 12 cm for boys and 12.5 cm for girls between 1920 and 1975 (Brundtland, Liestøl, and Walløe 1980). In Sweden, the average height of 12-year-old children increased by a similar amount between 1883 and 1965-71 (Ljung, Bergsten-Brucefors, and Lindgren 1974), and the average height of 13- to 14-year-old children in Holland increased by more than 20 cm between the mid-nineteenth century and 1965 (van Wieringen 1979).

In addition, there have been striking changes in the heights of children in central and eastern Europe. For example, the average height of Czechoslovakian 12-year-olds increased by between 12.8 and 15.5 cm between 1895 and 1968-70 (Suchy 1972), whereas the average height of 14- to 18-year-olds in Moscow increased by between 3.6 and 5.7 cm between 1925 and 1965 (Vlastovsky 1966). Moreover, the average heights of 12-year-old girls and 14- to 15-year-old boys in Zagreb increased by 10 cm and 13 cm, respectively, between 1951 and 1982 (Prebeg 1984).

Secular changes in birth weights and rates of physical maturation have also been examined by a number of authors. Of the two, the history of birth weights has received rather less attention, but new data are slowly becoming available (e.g., Ward 1988, 1990). In 1966, V. Vlastovsky demonstrated that the average length of newborn boys in Kursk City rose between 1930 and 1939 and rose again after the end of the Second World War. However, babies born in 1959 were only 0.1 cm longer than boys born 20 years earlier.The average weight of newborn boys increased from 3.43 kilograms (kg) in 1930 to 3.49 kg at the end of the 1950s. In 1988 Margit Rosenberg revealed that the average birth weight of children in Oslo, Bergen, and Trondheim increased by approximately 0.2 kg between 1860 and 1980. However, these increases are much smaller than the increases in the heights (and weights) of older children and adults (Malina 1979a). In 1981 Tanner estimated that the average birth weight of full-term males born at the end of the nineteenth century was only 0.1 kg less than that of European children today.

By contrast, changes in the rate of maturation and in age of menarche have been much more dramatic. In 1962 Tanner suggested that mean menarcheal age had declined by 4 years in Norway between 1840 and 1950; by 1.7 years in Sweden between 1890 and 1950; by 2.3 years in Finland between 1860 and 1940; and by 3.0 years in Germany over the same period (Tanner 1962, 1966; Steendijk 1966). There is also evidence after the Second World War of a continuation in this secular trend in Austria, Belgium, Bulgaria, Denmark, Finland, France, Germany, Greece, Hungary, Italy, the Netherlands, Poland, Romania, Spain, Sweden, the United Kingdom, Yugoslavia, and the former Soviet Union (Danker-Hopfe 1986; see also Venrooij-Ijsselmuiden, Smeets, and van der Werff Ten Bosch 1976; Laska-Mierzejewska, Milicer, and Piachaczek 1982; Helm and Helm 1984, 1987).

Some writers have suggested that the trend toward earlier maturation may be slowing down and that in some areas it may have ceased or even been reversed (Eveleth and Tanner 1976, 1991; Vercauteren and Susanne 1985). However, in 1986 Heidi DankerHopfe compared the median menarcheal ages of girls in eight countries in the 1950s and 1960s with the median menarcheal ages of girls in the same areas in the 1970s and 1980s and found an average decline in menarcheal age of approximately three and a half months (Table VI.5.2).

Table VI.5.2 Median menarcheal age of girls in various European countries 1950s-1960s and 1970s-80s Source: Danker-Hope (1986).

Change in median
Country Period age at menarche (yrs)
Belgium 1965-80/1 -0.10
Croatia (Zagreb) 1964-82 -0.45
France (Paris) 1965/6-74 -0.22
Hungary (Szeged) 1958/9-81/2 -0.43
Italy (Florence) 1955/7-77 -0.58
Netherlands 1965-80 -0.15
Poland (Warsaw) 1965-76 -0.28
United Kingdom
(N.E. England and
Newcastle-on-Tyne) 1967-71/9 -0.06

Europeans Outside Europe

Despite a wealth of information about the heights of Europeans in Europe, we know much less about the secular trend in the heights of European descendants in other parts of the world. The one obvious exception to this rule is the history of changes in the heights of European descendants in the United States. But there is also some information about long-term trends in the heights of European descendants in Australia, the Caribbean, and parts of South America. There have, however, been very few studies of the heights of European descendants in southern Africa.

The main trends in the heights of white Americans in the United States have been summarized on many occasions, including the work of Thomas Cone (1961), Howard Meredith (1963, 1976), Robert Fogel (1986), and Richard Steckel (1989, 1992). In 1986, Fogel revealed that the average heights of native-born white Americans rose steadily from the beginning of the eighteenth century up until the middle of the nineteenth century. This was, however, followed by a sharp decline in the heights of those born during the second half of the nineteenth century, although there was a resumption of the general upward trend in the first third of the twentieth century. Some writers have suggested that the secular trend ceased after World War II, but this suggestion was subsequently contradicted (Bakwin and McLaughlin 1964; Damon 1968). In 1989 R. Bock and R. Sykes demonstrated that the heights of people attending the Fels Institute in Ohio continued to increase after 1945, and T. Greiner and C. Gordon (1992) claimed that the average height of successive birth cohorts of white Americans increased by 0.7 cm per decade between 1910-14 and 1970-4 (Bock and Sykes 1989; Greiner and Gordon 1992).

In addition to studying the heights of adult Americans, a number of authors have also attempted to examine the heights of children. Cone, for example, showed that the average height of Boston children increased by between 3 and 7 centimeters between 1877 and 1930-56, and Harry Bakwin suggested that this trend continued into the early 1960s (Cone 1961; Bakwin 1964). The most comprehensive survey of heights in the United States was carried out by Meredith in 1963, who found that there had been very little change in either the length or weight of newborn infants since the end of the nineteenth century, but that there had been substantial increases in the heights of children from the age of 1 onward.

It is unclear whether these increases have continued toward the present. Eveleth and Tanner found no evidence of any increase in the heights of children in the United States between 1963-70 and 1976-80, but D. Jones, M. Neshiem, and J.-P. Habicht claimed that there was an increase in the heights of poor children over the same period (Jones et al. 1985; Eveleth and Tanner 1991).

There have also been a number of investigations into changes in the age of menarche among U.S. girls. In 1962 Tanner suggested that the mean age of menarche in the United States fell from 14.2 to 13.0 years between 1900 and 1950. These results have been confirmed by Grace Wyshak and Madeline Goodman. In 1983 Wyshak demonstrated that the mean menarcheal age among women born before 1920 was 0.86 years greater than the mean menarcheal age of women born after 1940 (Wyshak 1983). Goodman studied the mean age of menarche of 3,205 Caucasian, Japanese, and Chinese women in Hawaii who had been born between 1900 and 1940. The menarcheal age of the Caucasian women declined by 2.28 months per decade, whereas that of the Japanese and Chinese women declined by 6.36 months and 5.24 months, respectively (Goodman 1983).

In view of the available data, it is probably inevitable that the anthropometric history of European descendants outside Europe should tend to concentrate on the history of those in the United States. There is, however, some evidence of secular changes in the heights of European descendants in Australia, Cuba, Argentina, and Venezuela. In 1983, for example, Antonio Gordon found that the average height of Cuban men had increased by between 5 and 6 cm since the start of the century, although he denied that the nutritional condition of the Cuban population had improved during the course of this period.

In Argentina, H. Lejarraga and colleagues (1989) showed that the average height of 15-year-old children in Buenos Aires had increased by between 6.5 and 8.2 cm during the interval between 1938 and 1981, and that the average height of 15-year-olds in Entre Rios had increased by between 4.5 and 6 cm between 1950 and 1981. In 1981 N. Farid-Coupal, M. Contreas, and H. Castellano discovered that the mean menarcheal age of girls in the Carabobo district of Venezuela fell by 1.7 years between 1937 and 1969. Moreover, the average height of children in Carabobo increased by 1.6 cm between 1978 and 1987 (Blanco, Landaeta Jiménez, and Castellano 1989).

We also possess a limited amount of information about the heights of European descendants in Australia and Mozambique.The average height of children in western Australia increased by between 2 and 6 cm between 1940 and 1971, and that of children in Sydney increased by between 2 and 4 cm between 1937 and 1965 (Blanksby et al. 1974). These increases are somewhat greater than those recorded among white children in Mozambique.The average height of 7-year-old children in Lourenço Marques increased by only0.5 cm (boys) and 1.5 cm (girls) between 1930 and 1965 (Martins 1971).

Africans at Home and Abroad

There have been a number of investigations into the heights of African populations in recent years, but very few of these provide any clear opportunity for chronological comparison. The main source of information about secular trends in the heights of Africans and people of African ancestry has been the United States. The lack of information about secular changes in the heights of Africans in Africa, combined with the poor social and economic circumstances of most of the continent’s countries, has made it difficult to construct normal standards for assessing the growth of African children. However, studies of well-off Africans and black populations in the United States have shown that black children grow more quickly than white children between the ages of circa 8 and 14 years and experience puberty at an earlier age. These studies also suggest that Africans and African-Americans are more long-limbed than Europeans and white Americans, but that there is little difference in their final heights (Eveleth and Tanner 1991).

In view of the long history of anthropometric measuring in the United States, it is not surprising that we should possess quite a lot of information about the heights of black Americans. In 1963 Meredith suggested that the average height of 1-year-old boys rose by 5.7 cm between 1918-19 and 1944-56, and that the average height of 6-year-old boys rose by 6.3 cm between 1896-8 and 1957-8. In a separate study, William Moore discovered that the average height of 12-year-old boys increased by 17.41 cm between 1890 and 1968, whereas that of 12-year-old girls increased by 19.57 cm over the same period (Moore 1970). It is not clear whether the heights of black Americans continued to increase after this date. In 1979 Eveleth, E. Bowers, and J. Schall claimed that the average height of black adolescents in Philadelphia had increased by approximately 3.3 cm between 1956-65 and 1977, but the most recent national survey suggests that there was no increase in the heights of black children between 1963-70 and 1976-80 (Eveleth and Tanner 1976, 1991; Jones et al. 1985).

The availability of separate information about the heights of blacks and whites in the United States provides a valuable opportunity to study the secular trend in the heights of different population groups in the same national environment, but it must be admitted that the overall results are somewhat unclear. In his 1963 paper, Meredith argued that black males experienced greater increases in average height between the ages of 0 and 3, whereas white children experienced greater rates of increase between the ages of 6 and 17 (Meredith 1963). Yet Meredith also argued that black men experienced a greater increase in average height during early adulthood, and Moore has claimed that the average height of 14-year-old black boys increased by 2.88 cm per decade between 1890 and 1968. Previous estimates, however, had placed the rate of increase among black and white boys at 1.95 cm and 1.92 cm, respectively (Moore 1970). By contrast, Greiner and Gordon’s study of the heights of successive birth cohorts of U.S. army recruits contradicted both studies. They argued that the average height of black Americans born between 1910-14 and 1970-4 increased at less than half the rate achieved by white Americans during the same period (Greiner and Gordon 1992).

Despite such an abundance of information about the heights of African-Americans, we know rather less about secular changes in the heights of Africans in other parts of the world – including Africa itself. The most detailed information comes from southern Africa, where there have been a number of studies of the heights of South African blacks and Kalahari Bush-men. P. Tobias (1962, 1974) found that the average height of adult Bushmen (San) had increased by approximately 3.5 cm since the start of the century, and in 1985 A. Hausman and E. Wilmsen revealed that the heights of San children continued to increase between 1967-8 and 1979-80 (Tobias 1962, 1974; Hausman and Wilmsen 1985).

The situation regarding black populations in other parts of southern Africa is more unclear. Tobias (1975) has argued that most southern African populations “show either the absence of the secular trend …or a frankly reversed secular trend” (145), but A. Walker and B. Walker have claimed that the average heights of black children in South Africa rose by up to 6.9 cm between 1938 and 1976 (Tobias 1975; Tobias and Netscher 1976; Walker and Walker 1977).

A number of other studies have focused on the secular trend in the heights of Africans and people of African ancestry in Gambia, Mali, Nigeria, and the Caribbean island of Aruba. In 1981 Elisabeth van Wering demonstrated that the average height of children aged 5 to 14 on Aruba had increased by between 0.7 and 5.8 cm between 1954 and 1974. In a 1979 study suggesting improving nutrition in Nigeria, G. Ucho and A. Okorafor found that the mean menarcheal age of girls had fallen by approximately four months per decade since the early 1960s. However, no evidence of a secular trend toward greater height has been found in either Mali or Gambia. In Mali, the average height of adult men increased by only 0.2 cm between 1902 and 1985 (Prazuck et al. 1988). A longitudinal study of the heights of adult men and women in Gambia between 1951 and 1975 indicated that there was a slight tendency for height to decrease with age, but this was attributed to improvements in measuring techniques rather than to any real improvement in the heights of later-born subjects (Billewicz and Macgregor 1982).

Asiatics In Asia and the Americas

The term “Asiatic” describes a wide range of peoples who are believed to have originated in what is now known as the “Far East.” Thus present-day Asiatic populations include Mongols, Arctic Eskimos, American Indians, and Indonesian-Malays. The latter group alone includes the indigenous populations of Japan, China, Thailand, and the Philippines (Eveleth and Tanner 1976, 1991). So far as the secular trend in height is concerned, we have much information about the heights of Japanese people in Japan and the United States and about Chinese people in China and Hong Kong. There is a smaller amount of information about the heights of other Asiatic populations in the United States, South America, and Indonesia.

The Japanese are arguably the most-measured population group in the world. This is largely because Japanese authorities instituted a national program of anthropometric monitoring at the start of the twentieth century and have continued to publish the results ever since (Tanner 1986). In 1966 Eiji Takahashi showed that the average height of Japanese adolescents increased slowly but steadily between 1900 and 1937. Then there was a sharp fall in average heights during World War II, followed by an even sharper increase between 1945 and 1960.The average heights of Japanese adolescents continued to increase after 1960, and in 1982 Tanner and his colleagues found that the average heights of 12-year-old girls and 13-year-old boys had increased by 7.9 cm and 9.7 cm, respectively, between 1957 and 1977. The average height of Japanese adults increased by 4.3 cm (men) and 2.7 cm (women) over the same period (Takahashi 1966; Tanner et al. 1982; see also Meredith 1976).

Students of the growth of Japanese children have also examined changes in the rates of their ages of peak-height-velocity. In 1982 Kenji Matsumoto demonstrated that the age at which Japanese children achieved their peak-height-velocity had fallen steadily over the course of the twentieth century (Matsumoto 1982; see also Nagai et al. 1980). Matsumoto also noted that the age at which girls achieved their peak-height-velocity had fallen particularly rapidly since the end of World War II, and he attributed this change to the dramatic improvement in the nutritional status of Japanese women during that period (Matsumoto et al. 1980). These conclusions correspond closely to those reached by students of the age of menarche. The mean menarcheal age of Japanese girls fell from 15.0 years for those born between 1896 and 1900 to 12.5 years for those born between 1966 and 1970 (Hoshi and Kouchi 1981; Nakamura et al. 1986).

In addition to the Japanese data, there is also a growing body of information about the height of Chinese people in both China and Hong Kong. In 1984 Alan Piazza found that the average height of Chinese boys remained roughly constant between 1915-25 and 1951-8, but rose very sharply between 1951-8 and 1979. He was unable to locate representative data for Chinese girls in the earlier period, but he was able to show that their heights rose equally sharply from 1951 onward. There have also been significant changes in the heights of Chinese children in Hong Kong. J. Ling and N. King found that the average heights of 12-year-old boys increased by 6.7 cm and those of 12-year-old girls increased by 4.2 cm between 1961-3 and 1982-4 (Ling and King 1987: 187; see also Leung et al. 1987). The mean menarcheal age of South Chinese children in Hong Kong fell from 12.85 years in 1961-3 to 12.59 years 16 years later (Low, Kung, and Leong 1982).

We also have some information about the heights of Asiatic peoples in the southern United States, South America, and Indonesia. Pierre van der Eng has argued that the average heights of Indonesian and IndoEuropean children probably increased between 1911 and the late 1930s, but he was unable to find evidence of any further change between the late 1930s and 1973 (Eng 1995). Robert Malina and A. Zavaleta (1980) have indicated that the average heights of Mexican-American children aged 6 to 8 in the southern United States remained unchanged between 1929-31 and 1968-72, but small increases were recorded in the heights of those aged 8 to 15. The average height of Mexican-American children increased at all ages and at a greater rate between 1968-72 and 1982, but the overall rate of increase still lagged behind that of other U.S. groups. Malina, R. Martorell, and F. Mendoza (1986) concluded that the low rate of secular increase among Mexican-Americans reflected their failure to benefit from social and economic improvements, but their data may have been distorted by the inclusion of an influx of new migrants from Mexico and other South American countries.

It is interesting to compare the heights of Mexican-Americans in the United States with those of Asiatic peoples in South America. In 1974, Eveleth, F. Salzano, and P. de Lima studied the heights of 363 Xingo Indians in Brazil between the ages of 20 and 50 and found no evidence of any increase in the average height of successive birth cohorts (Eveleth, Salzano, and de Lima 1974). In 1982 G. Gonzales, I. Crespo-Retes, andR. Guerra-Garcia claimed that the average height of Puno Indians between the ages of 7 and 20 had increased substantially between 1945 and 1980, but they could find no evidence of any increase in adult heights (Gonzales, Crespo-Retes, and Guerra-Garcia 1982). The findings for adults, but not for children, were echoed by John McCullough in his study of the heights of Yucatec Maya between 1895 and 1968. McCullough concluded that the majority of Mesoamerican populations had failed to experience the secular increase in height that had been observed in most parts of the world since the beginning of the twentieth century (McCullough 1982).

Indo-Mediterraneans of the Near East, North Africa, and India

The term “Indo-Mediterraneans” is used to describe the indigenous populations of the “Near East,” North Africa, and India. It includes a wide range of diverse groups, including Hamites, Indo-Dravidians, Egyptians, Kuwaitis, and Libyans, but we know comparatively little about the secular trend in their growth or development. The majority of the relevant published studies concentrate on the populations of different parts of India, but it is possible to supplement these with some observations about possible trends in the heights of people in Ethiopia, Egypt, and Turkey.

Although there have been a number of studies focusing on different aspects of growth and stature in various parts of India, no consistent picture has emerged. It would appear that the pattern of change has varied considerably both between periods and between regions. In 1976 P. Ganguly examined secular changes in the height of 60 population groups and concluded that the dominant trend was toward a diminution of height rather than an increase (Ganguly 1977). This conclusion was reinforced by a study carried out in Hyderabad in 1977, but other investigations contradict it (Shatrugna and Rao 1987). In 1982 L. Sidhu, L. Bhatraga, and A. Dubey argued that the average height of well-off boys attending the Yadvindera Public School in Patiala increased by 7.5 cm between 1948-52 and 1973-7, and S. Singh and P. Malhotra contended that the mean menarcheal age of girls in Patiala fell by 0.65 years between 1974 and 1986 (Sidhu, Bhatraga, and Dubey 1982; Singh and Malhotra 1988). In 1981 D. P. Kaur and R. Singh compared the heights of Gujarati parents with those of their adult offspring and found that the children’s heights exceeded their parents’ by an average of 2.2 cm (Kaur and Singh 1981).

In view of the uncertainty surrounding the secular trend in the heights of Indians, it is interesting to compare the results of a number of different surveys carried out at different points in time between 1938 and 1982-3. In 1938 A. Chatterjee examined the average heights of more than 33,000 Bengali schoolboys, and the results were republished by W. M. Krogman in Growth of Man in 1941. These figures can be compared with the heights of the Indian boys whose measurements were reproduced in Eveleth and Tanner’s two volumes in 1976 and 1991. The results suggest that the average height of Indian boys may have declined between 1938 and 1956-65 and then increased sharply between 1956-65 and 1982-3 (Table VI.5.3). However, it is important to note that these surveys were carried out in different (although partly overlapping) areas and that they may have included different social groups.

One should also be mindful of such caveats when examining the secular trend in the height of other Indo-European populations. In the case of Egypt, information is available about the heights of 10- to 12-year-old boys in the whole of Egypt from 1956 to 1965 and about 10- to 12-year-old boys in East Cairo in circa 1981.The data show that the second group of children were between 3 and 4 cm taller than the first group, but this does not necessarily mean that the East Cairo children were taller than children from the same area or social background in the earlier study.

There is a similar problem with the data for Ethiopia, which show that “Ethiopian and Eritrean” children were between 5.2 and 6.7 cm shorter in 1958 than children in Addis Ababa and other urban areas 7 years later. In the case of Turkey, we can see that 13-year-old girls and 15-year-old boys in Ankara in 1950 were between 3.4 and 14.3 cm shorter than “poor” and “well-off” children in Istanbul in 1970.The figures suggest that average heights may have increased over time, but the difference between Ankara children in 1950 and poor children in Istanbul in 1970 is much less than the difference between the poor children and their well-off contemporaries.

Australian Aborigines and Pacific Island Peoples

This section reviews the available evidence regarding secular changes in the heights of the indigenous inhabitants of Australia, New Zealand, and the Pacific Islands. There have been a number of such studies of Aboriginal children, and in addition, a comparative study has been conducted of the heights of different groups of Samoan children. Insights can also be gained into the pattern of secular change by comparing the results of various surveys of the heights of people in Papua New Guinea.

The most detailed published effort examining Australian Aborigines in order to investigate the extent and nature of secular changes was that carried out by M. Barrett and T. Brown in 1961-9. They found that the adult heights of Aborigines in the central-western part of the Northern Territory were about 5.5 cm greater than the heights of comparable groups of Aborigines 30 years earlier (Barrett and Brown 1971). It is not clear, however, whether the experience of this group was typical of that of other Aborigines or whether the trend toward increased height has continued up to the present day.

In 1991 Eveleth and Tanner compared the heights of children in Barrett and Brown’s survey with those of Aboriginal children in western Australia in 1983, and they found that the heights were almost identical. But, as they themselves pointed out, it was impossible to say whether this meant that the secular trend had been halted or merely obscured by cross-sectional regional variations (Eveleth and Tanner 1976, 1991; see also Abbie 1967, 1968).

Table VI.5.3. Average heights of selected groups of Indo-Mediterranean children at different periods

Country People or place Period of study Source Height at age 13 (cm)
1. Males
India Bengal <1938 Krogman 1941 147.9
India India 1956-65 Eveleth & Tanner 1976 141.5
India Calcutta 1982-3 Eveleth & Tanner 1991 151.5
Egypt Egypt 1962-3 Eveleth & Tanner 1976 146.3
Egypt East Cairo <1981 Eveleth & Tanner 1991 149.0
2. Females
Ethiopia Ethiopia & Eritrea 1958 Meredith 1969 142.8
Ethiopia Urban 1965 Eveleth & Tanner 1976 148.0
Ethiopia Addis-Ababa 1965 Eveleth & Tanner 1976 147.1
Turkey Ankara (urban) 1950 Meredith 1969 144.5
Turkey Istanbul (well-off) <1973 Eveleth & Tanner 1976 155.6
Turkey Istanbul (poor) <1973 Eveleth & Tanner 1976 147.9

 

Country People or place Period of study Source Height at age 15 (cm)
1. Males
India Bengal <1938 Krogman 1941 161.3
India India 1956-65 Eveleth & Tanner 1976 153.0
India Calcutta 1982-3 Eveleth & Tanner 1991 162.4
Egypt Egypt 1962-3 Eveleth & Tanner 1976 159.7
Egypt East Cairo <1981 Eveleth & Tanner 1991 167.5
Ethiopia Ethiopia & Eritrea 1958 Meredith 1969 152.2
Ethiopia Urban 1965 Eveleth & Tanner 1976 158.5
Turkey Ankara (urban) 1950 Meredith 1969 152.7
Turkey Istanbul <1973 Eveleth & Tanner 1976 167.0
Turkey Istanbul (poor) <1973 Eveleth & Tanner 1976 159.9

In addition to these investigations, we can also examine a more recent inquiry into the heights of three different groups of Samoans. In 1975-7 J. Bindon and S. Zansky compared the heights of children in a “traditional” community of Western Samoa with those of children in two more “modern” communities in American Samoa and Hawaii. They found that there was little difference between the American Samoan children and the Samoan children on Hawaii, but both were consistently taller than the “traditional” children in Western Samoa (Bindon and Zansky 1986). The heights of the Hawaiian Samoans were also considerably greater than those of native Hawaiian children who were measured almost half a century earlier.The average height of native Hawaiian children who were measured before 1930 was approximately 7 cm less than that of Hawaiian Samoan children in the mid-1970s (Krogman 1941).

Insights into the pattern of secular change among Pacific Island peoples can also be gained by comparing the results of a number of recent surveys of the heights of the Bundi and Manus peoples of Papua New Guinea. The Bundi, for example, were among the smallest people in the world when they were first measured by L. A. Malcolm in 1958-60, but their heights had increased considerably by the time a second survey was carried out in 1983-4 (Malcolm 1970; Zemel and Jenkins 1989).

There is also some evidence of an increase in the heights of Manus girls between 1966-8 and 1982, although the heights of Manus boys remained virtually unchanged (Heath and Carter 1971; Schall 1989). In view of the discrepancy between the data for boys and girls and the small number of children involved, it would probably be unwise to reach any firm conclusions about trends in the height of this population at the present time (Eveleth and Tanner 1976, 1991).

Conclusions

This chapter has shown that the average height of people in most parts of the world has increased since the beginning of he twentieth century and that most population groups are probably taller now than at any time in the recent past (Kates and Millman 1990). However, the secular trend toward greater height has been neither universal nor rectilinear (Roche 1979). The average height of populations has increased much more rapidly in Europe and the United States than in many parts of Africa, Asia, and South America, and even those countries that have registered the greatest increases in height have also experienced periods in which the secular trend has either been arrested or reversed (Takahashi 1966; van Wieringen 1979). Moreover, the distribution of heights in virtually all parts of the world continues to be marked by social and economic inequalities (Eveleth and Tanner 1991). This finding suggests that even within the “developed” world large numbers of children do not experience the nutritional and environmental conditions necessary for the achievement of their full height potential (Jones et al. 1985; Whincup, Cook, and Shaper 1988).

A number of writers have attempted to account for the secular trend and for the variations within it. Malina (1979b) attributed the trend to improvements in nutrition and sanitation, urbanization, industrialization, reduced family size, genetic selection, and heterosis. Eveleth and Tanner (1991) have credited improvements in nutrition, the control of infectious diseases, reduced family size, better health care, and population mobility. Other scholars who have examined the consequences of the secular trend have associated it with the reduction of mortality, the lengthening of the female reproductive span, changes in the relationship between children and adults, and increases in physical performance and efficiency (Himes 1979; Malina 1979a, 1979b; Waaler 1984; Eveleth and Tanner 1991).

Several writers have examined the relationship between changes in height (and other anthropo-metric indicators) and economic growth. In 1984 Floud concluded that variations in Gross Domestic Product (GDP) per head and in infant mortality rates explained “about 96 per cent of the observed variation in heights between and within western European populations since 1880,” and that “an increase of one U.S. dollar (at constant 1970 prices) in GDP per capita has been accompanied by an increase in the average height of the population of 0.003 centimeters” (Floud 1984: 18). In 1988 H. Brinkmann, J. Drukker, and B. Slot argued that there was a close association between changes in GDP and the height of Army conscripts in Holland, and K. Liestøl found evidence of a similar relationship between changes of GDP and mean age of menarche in Norway (Liestøl 1982; Brinkmann, Drukker, and Slot 1988; but see also Mandemakers and van Zanden 1993). In 1988 L. A. Schmitt and G.A. Harrison concluded that “affluence” was the most important single cause of the differences that they observed in the average heights of 58 populations whose heights had been recorded in Human Biology and Annals of Human Biology over the previous 20 years (Schmitt and Harrison 1988).

Although economic growth is probably a necessary precondition for the long-term improvement of health and nutrition, it is not the only factor that has influenced trends in average height over the past 200 years. Both John Komlos and Floud have demonstrated that it is possible for the average height of a population to decline even when its GDP is increasing. In 1989 Komlos showed that soldiers who grew up in the economically more developed regions of the Hapsburg Empire during the second half of the eighteenth century were shorter than soldiers who grew up in the less developed regions, and Floud revealed that the average height of successive birth cohorts of British soldiers declined during the middle years of the nineteenth century (Komlos 1989; Floud et al. 1990). In 1983 Steckel concluded that differences between the average heights of 22 contemporary populations were attributable to variations in national income, income inequality, and welfare provision (Steckel 1983). These findings are reinforced by the dramatic increases in average height in China and those that have been achieved in some parts of India and Sri Lanka since the end of World War II (Drèze and Sen 1989; Kates and Millman 1990).