Community

and International

Nutrition

The Association between Wasting and Stunting: An International Perspective CESAR G. VICTORA Departamento de Medicina Social, Uniuersidade Pelotas, RS, Brazil

group (4), almost ten years later, expressed some caution regarding the interpretation of wasting and stunting, by stating that "terms such as 'acute' mal nutrition (for wasting) and 'chronic' malnutrition (for stunting) and 'acute-on-chronic' for the combination

ABSTRACT Wasting and stunting are widely used in dices for describing the prevalence of malnutrition in childhood. Studies from Latin America, however, consis tently show low prevalences of wasting in the presence of other indicators of poor health. A data base as sembled by the World Health Organization, including 175 studies with information on wasting and stunting among 1-y-old children from four regions (Africa, Americas, Asia and Eastern Mediterranean), was used for examining the prevalences of the two indices and their associations. Whereas there was little variation in the median prevalences of stunting among the four regions, those of wasting showed a sevenfold differen tial. In Africa and Latin America, the correlations be tween the two prevalences were low; nevertheless, wasting was about three times higher in Africa, for any given level of stunting, than in Latin America. In Asia and the Eastern Mediterranean, there were strong positive correlations between stunting and wasting, but wasting was about twice as common in Asia for any given level of stunting. To interpret low prevalences of wasting as being indicative of a low intensity of mal nutrition may be misleading; weight-for-height standards and cut-offs may have to be re-examined for use in different populations. J. Nutr. 122: 1105-1110, 1992.

of wasting plus stunting, are not direct observations but deductions which may not always be correct." Both indices are dependent upon a comparison with a growth reference—usually from a developed country population such as the National Center for Health Statistics (NCHS) standard (5)—because it has been assumed that the growth potential of young children is largely similar across ethnic groups and continents (6). After several years of being involved in the exami nation of tens of thousands of Latin American chil dren, this author has become increasingly puzzled by the low prevalence of wasting found in these popula tions at any age group. Child populations with ap palling social and environmental conditions, stag gering levels of morbidity and mortality, and high prevalences of stunting, had minor prevalences of wasting, often below that of the North American NCHS reference population. For example, in a survey (7) of 4513 children under 3 y of age in Ceará state in northeastern Brazil, the infant mortality rate was 96 per thousand, 74% of the families had no access to piped water, 12% of the children had diarrhea on the day of the interview and 27.6% were moderately or severely stunted. Yet, the prevalence of moderate or severe wasting was only 2.4%, almost exactly what would have been observed in the NCHS reference population (5). The low prevalences of wasting in Latin America have been noted by a number of researchers. Martorell and Ho (8) argued that "where malnutrition is

INDEXING KEY WORDS:

•protein-energy malnutrition •epidemiology •humans

The concepts of wasting and stunting were intro duced by Waterlow (1, 2) in the early 1970s to express two different forms of protein-energy malnutrition in childhood. Wasting, a deficit in weight relative to height, was originally defined as an acute situation, indicating recent weight loss. Stunting, a deficit in height or length relative to a child's age, was proposed as an indicator of longstanding or chronic malnutri tion. This idea was supported in 1977 by a World Health Organization (WHO)1 committee (3), which defined "weight for height as an indicator of the

Abbreviations used: NCHS, National Center for Health Statis tics; U5MR, under-five mortality rate; WHO, World Health Organi zation.

present state of nutrition and of height for age as an indicator of past nutrition." A second WHO working 0022-3166/92

$3.00 ©1992 American

Institute

of Nutrition.

Federal de Pelotas, CP 464, 96001

Received 26 April 1991. Accepted 19 November 1105

Downloaded from https://academic.oup.com/jn/article-abstract/122/5/1105/4754815 by University of California, Santa Barbara user on 17 March 2018

1991.

VICTORA

1106

mild to moderate, as in Latin America, children ex hibit moderate degrees of stunting but few abnormal ities in measures of wasting...where malnutrition is moderate to severe, as in India and Nepal, stunting is marked and moderate levels of wasting are observed." A similar argument was raised by Keller and Fillmore (9), who stated that "wasting appears not to be a major problem in these populations, which seems plausible in view of the generally advanced state of development in Latin America." Both groups of authors, therefore, interpreted the low frequency of wasting in Latin America as being indicative of a low intensity of malnutrition associated with reasonable socio-economic conditions. Another possible explanation is that Latin American children differ in body proportions and/or constitution from those of North America, where the NCHS standards were created. This hypothesis was presented by Trowbridge et al. (10), who showed that the increased weight for height of Peruvian children did not result from increased fat tissue, but rather from increased hydration of the fat-free body. These findings from an Andean population may not, how ever, be applicable to other areas of Latin America. The publication (11) by the World Health Organi zation of the prevalences of malnutrition in a number of countries provides an interesting opportunity to compare the distribution of three most often used indicators (weight-for-height, height-for-age and weight-for-age) in different continents. It also allows the study of the associations between these indica tors, which might contribute to the clarification of some of the above issues.

METHODS Cross-sectional data on the prevalences of wasting and stunting were obtained from the World Health Organization publication, Global Nutritional Status (11). Five percent of these studies are from the 1960s, 33% from the 1970s and the remaining 62% from the 1980s. The analysis was restricted to children age 12-23 mo of both genders, because this is the age range of highest prevalence of wasting (9) and was also the one with the largest number of studies. Similar patterns were observed for all children under 5 y of age. The units of analysis were individual studies. Only studies with 50 or more children in the 12- to 23-mo age range were included; the number of children per study ranged from 52 to 31,788, with a median of 341 children per study. Only regions with 15 or more studies were included in the analyses, and the few studies from developed countries were ex cluded. The administrative regions used by the WHO (11) were slightly adapted, as follows: Caribbean countries were separated from South and Central America; developed countries were excluded; coun Downloaded from https://academic.oup.com/jn/article-abstract/122/5/1105/4754815 by University of California, Santa Barbara user on 17 March 2018

tries from the Western Pacific in continental Asia (China, Laos, Malaysia and Korea) were grouped with those from Southeast Asia; two studies conducted in drought and famine areas (Somalia and Sudan) were excluded. The countries included in each regions were as follows (the number of studies, for countries with more than one study, are shown in parentheses): Africa (89 studies): Botswana (4), Burkina Faso (4), Burundi (4), Cameroon, Cape Verde (2), Congo (2), Ethiopia (9), Gambia (4), Ghana, Guinea-Bissau, Ivory Coast, Kenya (3), Lesotho (4), Liberia, Madagascar (3), Malawi (7), Mali, Niger (2), Rwanda, Senegal (3), Seychelles, Sierra Leone (2), South Africa (7), Swaziland, Tanzania (3), Togo (4), Uganda (2), Zaire (10), Zambia,America (37 studies): Belize, Bolivia (4), Brazil (10), Chile (2), Colombia (2), Costa Rica (2), El Salvador (2), French Guyana, Guatemala, Guyana, Honduras (2), Nicaragua (3), Panama (2), Peru (2), Venezuela (2); Asia China Nepal, Lanka

(32 studies): (2), India (4), Republic of (4), Thailand

Bangladesh (4), Burma (5), Indonesia, Laos, Malaysia, Korea, Singapore (4), Sri (3);

Eastern Mediterranean (17 studies): Democratic Yemen (4), Egypt, Iran (2), Jordan, Kuwait (3), Palestinian refugees (4), Tunisia, Yemen. The prevalences of stunting (deficit in length for age), wasting (deficit in weight for length) and under weight (deficit in weight for age) in each study are the percentages of children two or more standard devia tions below the NCHS median. Estimates of the under-five mortality rates (U5MR) for the countries listed above were obtained from The State of the World's Children 1990 (12). It was not possible to obtain mortality estimates for 22 areas (16 from Africa, two from the Americas and four from the Eastern Mediterranean). Median prevalences by region are shown, because these are less likely to be affected by extreme values than would the mean values. Pearson's correlation coefficient and linear regression were used in the analyses (13).

RESULTS Table 1 shows the median prevalences of mal nutrition by region. Prevalences of stunting tended to be higher than those of underweight, except for Asia. Prevalences of wasting were the lowest of the three indicators.

INTERNATIONAL

VARIATIONS

IN WASTING

AND STUNTING

1107

TABLE 1 Median prevalences of malnutrition and estimated under-ftve mortality rates (U5MR), by region malnutrition1RegionAfricaLatin

Prevalence of ofstudies89373217

41.133.847.038.6Underweight35.722.852.925.4U5MR146857299No AmericaAsiaEastern MediterraneanWasting10.12.718.85.8Stunting% ^Percentage of children below the National Center for Health Statistics median minus two so.

All three prevalences tended to be lowest in Latin America and highest in Asia. The prevalences of stunting showed less variability (medians ranging from 33.8% for Latin America to 47.0% for Asia) than those of underweight (from 22.8 to 52.9%) or wasting (from 2.7 to 18.8%). The median prevalence of stunting was 4.1 times higher than that of wasting in Africa. The corres ponding ratios were 2.5 for Asia, 6.7 for the Eastern Mediterranean and 12.5 for Latin America. Regions with similar median prevalences of stunting, such as Africa and Asia (41.1 and 47.0%, respectively), showed considerable differences in wasting (10.1 and 18.8%, respectively). On this regional basis, there were no clear associa tions between the U5MR and the nutritional indica tors. For example, Asia had a median U5MR of 72 and yet presented three times more wasting than did the Eastern Mediterranean, where mortality was higher. The African continent, despite having a median mor tality level considerably higher than the other regions, was not the worst in any of the nutritional indicators. Table 2 shows Pearson's correlation coefficients between the prevalences of wasting and stunting within each region, the units of analysis being the individual studies. There were strong direct correla tions of -0.7 between the prevalences of wasting and stunting in Asia and the Eastern Mediterranean.

There was a weak, not quite significant correlation of -0.3 for Latin America, and no association for Africa. The regression analyses of wasting on stunting are shown in Table 2 and Figure 1. Again, marked vari ation was seen. The intercepts (a values) were close to zero for Latin America, Asia and the Eastern Mediter ranean, suggesting that if it were possible to eliminate stunting, wasting would also be eliminated. This was not the case for Africa, where a prevalence of wasting of -10% would still be expected even if stunting could be eliminated. (These are obviously theoretical exercises because they imply considerable extrapo lation from the actual range of the data.) The slopes (ß values) also showed marked varia tion. For both Africa and Latin America, an increase of 0.5% or less in the prevalence of wasting was expected per an increase of 10% in stunting. For the Eastern Mediterranean and Asia, however, the corres ponding increases in wasting would be 3.1 and 4.4%, respectively. The regression and correlation coefficients were hardly changed when the analyses were restricted to studies with prevalences of stunting between 20 and 60%. Figure 2 combines the regression lines from the four regions. To avoid extrapolation, the lines in Figure 2 are bound by the interquartile range of stunting in each region, although the full data were used in their calculation. For any given level of

TABLE 2 Correlation and regression analysis of the prevalences of wasting on stunting, by region of studies89

RegionAfrica Latin America Asia Eastern Mediterraneanr0.044

0.311 0.700 0.652a10.185

2.097 2.093 -2.146ß0.020

Downloaded from https://academic.oup.com/jn/article-abstract/122/5/1105/4754815 by University of California, Santa Barbara user on 17 March 2018

0.053 0.440 0.082 0.310SE(ß)0.0500.028 0.071P0.7

0.06

The association between wasting and stunting: an international perspective.

Wasting and stunting are widely used indices for describing the prevalence of malnutrition in childhood. Studies from Latin America, however, consiste...
1MB Sizes 0 Downloads 0 Views