Haiti nutrition status survey, 1978*

Bulletin of the World Health Organization, 58 (5): 757-765 (1980) Haiti nutrition status survey, 1978* P. L. GRAITCER,1 M. A. GEDEON, 2 I. DE BEAUSSET, 3 & E. M. DUCKETT 4 The Haiti National Nutrition Survey, conducted during June-September 1978, quantified the magnitude and distribution of malnutrition in young children and their mothers. Ofthe 5353preschool children surveyed, 6.0% wereseverely wasted (less than80% ofthe reference median weight-for-height). Stunting (less than 90% ofthe reference median height-for-age) was mostprevalent in the 48-59-month agegroup. Over40% ofthe children in thisgroup were stunted. Although the causes ofundernutrition cannot be documented in a cross-sectional survey, the baseline data obtained constitute important evidence that acute undernutrition is a major problem in Haiti, that it is particularly pronounced in children 12-23 months old, and that it is evenly distributed through all rural departements in Haiti. Protein-energy malnutrition and other nutritional deficiencies have been cited among the most serious health problems in Haiti. A number of major nutritional surveys have been undertaken since 1958 (1, 2, 3).° Although these studies were conducted at different times of the year and used different meth- odologies, comparisons are possible. Between 24 and 52% of children aged 0-6 years of age had second or third degree malnutrition, using the Gomez classifi- cation of weight-for-age.b According to these estimates about 400 000 children under 5 years of age had second or third degree malnutrition. There have also been reported deficiencies in vitamin A, ribo- flavin, iron, and folic acid; however, the magnitude of these deficiencies does not approach that of the protein-energy deficit in Haitian children. The Bureau of Nutrition, Department of Public Health and Population, Republic of Haiti, with the assistance of the Center for Disease Control, Public * This study was conducted under the auspices of the Bureau of Nutrition, Department of Public Health and Population, Republic of Haiti, and was supported by the United States Agency for International Development and the Center for Disease Control, Public Health Service, Department of Health and Human Services. The opinions expressed herein are those of the authors and do not necessarily reflect the official views of the sponsoring authorities and organizations. I Epidemiologist, Bureau of Smallpox Eradication, Center for Disease Control, Public Health Service, Department of Health and Human Services, Atlanta, GA 30333, USA. 2 Director, Research and Evaluation Branch, Bureau of Nutrition, Department of Public Health and Population, Republic of Haiti. 3 Public Health Nutritionist, Bureau of Nutrition, Department of Public Health and Population, Republic of Haiti. 4 Statistician, Bureau of Smallpox Eradication, Center for Disease Control, Public Health Service, Department of Health and Human Services, Atlanta, GA 30333, USA. a Also: TOUREAU, S. ET AL. Assessment of the prevalence of xerophthalmia in Haiti, American Foundation for Overseas Blind, 1976 (unpublished document). b KING, J. M. Analysis and compilation of nutrition data and studies, US Agency for International Development, 1978(unpublished document). Health Service, Atlanta, Georgia, USA completed a national nutrition status survey between May 1978 and September 1978. The purposes of this survey were (1) to provide an estimate of the nutritional status of children of preschool age; (2) to determine the preva- lence of anaemia in these preschool children and their mothers; and (3) to determine the frequency of selected social, demographic, health, and dietary characteristics and their relationship to the nutritional status of children. This paper presents data on the anthropometric measurements of nutritional status, the prevalence of pedal oedema, the prevalence of anaemia, and the duration of breast-feeding for Haitian preschool children. METHODS The Republic of Haiti has a land area of approxi- mately 28 500 kM2, most of which is mountainous. The estimated 1978 population, based on extrapol- ation from the 1971 sample census, is 5.3 million persons; 16% of this population, about 840 000, are under 5 years of age. Approximately 24/o of the population live in urban areas, the largest of which is Port-au-Prince with a population of about 800 000. For the purposes of this survey, the rural sample was taken from five sampling universes that correspond to former geopolitical divisions (Departements). The urban centres of these rural universes were excluded from the sampling (see Fig. 1). The sixth sampling area represents the preschool urban population of metropolitan Port-au-Prince, which in this study is considered separately from the Departement de l'Ouest. Sampling A two-stage population proportional sampling method was used. In the first stage, 30 enumeration 4000 -757- P. L. GRAITCER ET AL. Fig. 1. Map of Haiti showing sampling universes. districts were selected in each universe by using a random start systematic sample from the 1971 census listing. The second stage involved random selection of initial households in each sample site. Data were collected on 30 consecutive children aged 3-59 months, beginning with the initial household. Nine hundred preschool children and their guardians were sampled in each universe (4). A special group of 730 children aged 3-59 months was identified in private nursery schools and from socially advantaged families in Port-au-Prince. Anthropometric indices for these children were used to provide an estimate of the potential growth attain- able by the Haitian preschool-age population. Survey administration Eight 3-man survey teams were trained to administer the survey questionnaires, weigh, and measure the children, and obtain blood specimens for haemoglobin analyses. The importance of accuracy and reproducibility of the measurements was emphasized during the 3-week training course. The survey team measured the height or length of each survey child to the nearest 0.1 cm with a portable measuring board. Weight was determined to the nearest 0.1 kg using a hanging spring scale. The presence or absence of pedal oedema was determined by firm thumb pressure for 3 seconds on the dorsal surface of both feet. Blood specimens were obtained for haemoglobin determinations on a 20% subsample of children and their mothers. Capillary blood was collected by finger prick. Haemoglobin content was determined by the cyanmethaemoglobin method using a Fisher 74D spectrophotometer. Anthropometric indices Three anthropometric indices are used to describe the nature and extent of malnutrition in preschool children: weight-for-height, height-for-age, and weight-for-age (5). Weight-for-height, an estimate of body proportions, provides an index of current 758 HAITI NUTRITION STATUS SURVEY nutritional status. Height-for-age is a measure of linear growth and therefore an index ofpast or chronic undernutrition. The weight-for-age index is a measure of the combined current and past components of nutritional status. It does not discriminate between the acutely undernourished thin child, the chronically undernourished child who is short for his age, and the short and overweight child. Data for all anthropometric indices are presented as percentages of the NCHS/CDC reference median rather than the Stuart Meredith reference for height and weight. The larger number of children, improved sampling techniques, direct weight-for-height cal- culations, and the complete statistical description used in the NCHS/CDC reference improve the stat- istical precision of the outlying percentiles of the anthropometric data (6). Children with a weight-for-height of less than 80% of the NCHS/CDC reference median are considered acutely undernourished (wasted). Children with a weight-for-height between 80 and 89.9'/o of the reference median are moderately undernourished. Use of this intermediate classification permits the identification of a wider spectrum of undernourished children and helps place the malnutrition problem in proper perspective. Children with a weight-for-height greater than 10% of the NCHS/CDC median are considered overweight. Children with a height-for-age less than 900o of the reference median are chronically malnourished (stunted). Children with a weight-for- age of between 75 and 90/o of the reference median are classified, using an index similar to the Gomez classification, as having first degree or mild malnu- trition (7). Those children with a weight-for-age of less than 75% of the reference median are classified as having second or third degree malnutrition. Waterlow has suggested a system of classification that permits assessment of the extent of under- nutrition in children by using the categories of wasting, stunting, and concurrent wasting and stunt- ing. Children less than 8007 of the reference median weight-for-height have wasting alone if they are above 90% of the median height-for-age. Children are considered to have stunting if they are below 90% of the median height-for-age but above 80% of the median weight-for-height. Children below the cut- offs for weight-for-height and for height-for-age are classified as having concurrent wasting and stunting, and are probably at greatest risk of morbidity and mortality (8). Graphic comparisons of anthropometric measure- ments with reference population values utilize the statistical parameter, standard deviation scores. The standard deviation score describes the approximate probability distribution of values and permits easy comparison with nutritional data from other countries. RESULTS Data from 5353 preschool children were analysed- questionnaires on 46 children were excluded because of obvious errors in age or anthropometric data. Age and sex distribution of the children are shown in Table 1. Table 2 presents the distribution of weight-for- height as percentages of median values for Haitian preschool children by age group. Ofchildren aged 3-5 months, 2%o are severely wasted. The prevalence of severe wasting is greatest (9.8 0/) in children 12-23 months of age. Table 1. Percentage distribution of preschool children by age and sex: Haiti 1978a Age (months) Male Female Total 3- 5 3.4 ( 179) 3.3 ( 170) 6.7 ( 349) 6 - 11 6.7 ( 341) 7.0 ( 383) 13.7 ( 724) 12-23 11.6 ( 610) 12.2 ( 645) 23.8 (1255) 24- 35 10.5 ( 577) 9.9 ( 530) 20.4 (1107) 36-47 10.4 ( 556) 9.3 ( 493) 19.6 (1049) 48-59 8.2 ( 443) 7.6 ( 426) 15.8 ( 869) Total 50.8 (2706) 49.2 (2647) 100.0 (5353) a Representative national sample. All percentages are weighted by universe population proportions. The actual numbers of children sur- veyed are given in parentheses. Table 2. Percentage distribution of preschool children by weight-for-height classes (percentage of reference median) and age: Haiti 1978w Over- Total Wasting Normal weight number of Age 80.0- 90.0- children (months) < 80.0% 89.9% 109.9% ) 110% surveyed 3 - 5 2.0 14.4 37.9 45.8 349 6- 11 5.1 26.7 39.0 29.2 724 12 - 23 9.8 38.0 38.9 13.2 1255 24 - 35 7.5 31.8 40.9 19.8 1107 36 - 47 3.7 22.6 48.7 24.9 1049 48- 59 3.7 26.4 48.1 21.8 869 Total 6.0 28.8 42.6 22.6 5353 a Representative national sample. All percentages are weighted by universe population proportions. 759 P. L. GRAITCER ET AL. Table 3. Percentage distribution of preschool children by weight-for-height classes (percentage of reference median) and geographical area: Haiti 1978 Wasting Normal Overweight Total number of Geographical areaa < 80.0% 80.0-89.9% 90.0-109.9% > 110% children surveyed Nord-Ouest 5.3 30.2 43.4 21.1 891 Nord 6.8 33.5 42.9 16.7 892 L'Artibonite 6.9 27.1 44.3 21.7 889 Ouest 5.1 27.6 44.8 22.4 895 Sud 7.3 32.4 38.5 21.8 893 Representative rural sample 6.4 30.1 42.6 20.7 4460 Metropolitan Port-au-Prince 3.8 21.7 42.9 31.7 893 Representative national sample 6.0 28.8 42.6 22.6 5353 Special group 0.1 7.3 76.8 15.8 730 8 See Fig. 1 for areas covered by sample universes. All percentages with the exception of the special group are weighted by universe population proportions. The data in Table 3 on the distribution of weight- for-height as percentages of median values show that in the representative rural sample (weighted data from all rural geographical areas), 6.4% of the children 3-59 months of age are severely wasted (less than 80% of the reference median weight-for-height). In metro- politan Port-au-Prince, 3.8% of the children are severely wasted. The differences in prevalence of wasting in the various geographical areas are not significant. To permit comparisons with other surveys, the distribution of weight-for-age values of Haitian chil- dren 3-59 months of age are shown in Table 4. The prevalence of second and third degree malnutrition is higher in the rural areas of Haiti than in metropolitan Port-au-Prince. No children in the special group have third degree malnutrition and only 0.5% in that group have second degree malnutrition. Table 4. Percentage distribution of preschool children by degree of malnutrition (Gomez classes) and geographical area: Haiti 1978 Gomez classesb 3rd: 2nd: 1st: Normal: Total number of Geographical areaa <60.OC 60.0-74.9 75.0-89.9 90.0+ children surveyed Nord-Ouest 3.0 22.8 48.8 25.4 891 Nord 5.7 28.5 46.2 19.6 892 L'Artibonite 2.5 28.3 45.4 23.7 889 Ouest 2.7 23.4 47.2 26.8 895 Sud 3.6 25.9 46.0 24.5 893 Representative rural sample 3.5 26.0 46.4 24.1 4460 Metropolitan Port-au-Prince 1.5 13.1 43.8 41.7 893 Representative national sample 3.2 24.1 46.0 26.8 5353 Special group - 0.5 15.1 84.4 730 a See Fig. 1 for areas covered by sample universes. All percentages with the exception of the special group are weighted by universe population proportions. b In terms of percentage of NCHS/CDC reference median weight-for-age. c Children identified as having pedal oedema were not automatically ciassified as third degree malnutrition. 760 HAITI NUTRITION STATUS SURVEY The weight-for-height standard deviation distri- bution curves shown in Fig. 2 show that the special group curve is in approximately the same position as the reference population curve, indicating that the distribution of heights-for-weights of this group is roughly similar to that of the healthy population from the United States of America used to develop the NCHS/CDC reference curve. The curve of the representative national sample of Haitian preschool children is shifted leftward, indicating that weight- for-height distribution values are lower than in the special group and the reference populations. tl 10 9 i8 ;.7 I 4 3 565 -5.0 *4.5 -4.0 56 *3.0 *2.5 *2.0 1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 6.5 Z-Score Fig. 2. Weight-for-height Z-scores: Haiti 1978. Height-for-age as a percentage of median distri- bution by age group is shown in Table 5. Little stunting is seen during the first year of life. Dramatic increases in the prevalence of stunting begin during the second year oflife and continue through the fifth year. The level of moderate stunting remains fairly constant from 24 to 59 months. Table 5. Percentage distribution of preschool children by height-for-age classes (percentage of reference median) and age: Haiti 1978° Stunting Normal Total number of Age 85.0- children (months) < 85.0% 89.9% 90.0 + surveyed 3 - 5 - 3.0 96.9 349 6- 11 0.6 5.1 94.2 724 12 - 23 4.2 15.1 80.8 1255 24 - 35 9.7 25.6 64.7 1107 36 - 47 12.1 25.0 62.9 1049 48- 59 16.6 25.4 58.1 869 Total 8.0 18.6 73.4 5353 8 Representative national sample. All percentages are weighted by universe population proportions. The prevalence of stunting (height-for-age less than 90% of the reference median) ranges from 15.7 to 33.4% among the geographical areas. Although there is some variation in the prevalence between the rural areas, it is not statistically significant. The height-for-age standard deviation values shown in Fig. 3 show that the distribution of values for the special group children follows almost exactly that of the reference population. The distribution of height-for-age of the survey group is shifted leftward indicating that this group is shorter than the reference population and the special group by an average of about 1.75 standard deviation units (Z scores). 12 11 10 -- Specil Group In= 73DI U- nrm I-VIin =55) Expe1d (Rle-orm) A AN'A .6.0 4.6 -4.0 -3.5 -20 -2.5 -2.0 -1.0 1.0 40.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Z-Score Fig. 3. Height-for-age Z-scores: Haiti 1978. The Waterlow cross-classification of height-for-age and weight-for-height is shown in Table 6 for each age group. The highest prevalence of concurrent wasting and stunting (5.4%) is seen in the 24-35-month age group. The dramatic increase in the prevalence of Table 6. Percentage distribution of preschool children by Waterlow classes and age: Haiti 1978° Total number of Age Wasting children (months) Wastingb Stuntingc & stunting Normal surveyed 3-5 2.7 1.7 0.3 95.3 349 6 -11 4.6 5.2 0.5 89.7 724 12 - 23 6.4 15.8 3.4 74.3 1255 24- 35 1.8 29.6 5.6 62.9 1107 36 - 47 0.8 34.2 2.9 62.1 1049 48- 59 0.8 39.0 3.0 57.3 869 Total 2.9 23.6 3.1 70.4 5353 See Fig. 1 for areas covered by sample universes. All percentages are weighted by universe population proportions. b Wasting = < 80% of median weight-for-height. c Stunting = < 90% of median height-for-age. 761 __ Specol Group n = 73W ... Unwre I-VI In = 553) A Expcted MRel. /I -X I I II I ' - . I I ....-UL 4 3 2 P. L. GRAITCER ET AL. wasting and stunting seen in the 12-23-month group corresponds with the traditional age at which Haitian children are weaned. The high nutritional require- ments for growth coupled with episodes of infection plus the change in food source from breast milk with some supplementation to complete reliance on the family diet are probably responsible for creating this high-risk nutritional situation. Bilateral pedal oedema, a sign frequently attribu- table to severe undernutrition, was seen in 2.7%o of all children surveyed. The prevalence of pedal oedema in the group nutritionally most at risk, those children with wasting and stunting, was 10.4%o. The prevalence of anaemia in a 20%o subsample of 1056 children is shown in Fig. 4. Wide statistical vari- ance prevented statistical testing; however, anaemia appeared to be least prevalent in l'Artibonite and most prevalent in urban Port-au-Prince. The 24-35-month Gaph1= - Fig. 4. Prevalence of anaemia8 in preschool children by geo- graphical area: Haiti 1978. 8 Anaemia defined as a haemoglobin value of< 100 g / litre for ages < 24 months and < 110 g/litre for ages > 24 months. age group in the representative national sample has the highest prevalence of anaemic children. Table 7 shows that the percentage of children being breast-fed is inversely proportional to their age. In the rural areas almost all children are breast-fed through the first year of life. Between 50 and 67% of children in the rural areas are breast-fed through the second year of life. In metropolitan Port-au-Prince less than 29% of the children are being breast-fed after 12 months of age. As regards age at weaning, Table 8 shows that in both the urban and rural areas, younger mothers tend to wean their children earlier than do older mothers. Differences in breast-feeding patterns are apparent between urban and rural areas: urban mothers wean their children earlier than rural mothers. In all age groups urban mothers are less likely to begin breast-feeding than rural ones. DISCUSSION The Haiti Nutrition Status Survey provides objective information on the magnitude and location of protein-energy malnutrition found in a represen- tative national sample of children aged 3-59 months. These data can be useful in planning nutritional policy and developing programmes to improve the nu- tritional status of the Haitian population. The data can also be used as a baseline in evaluating the effectiveness of future nutrition programmes. Undernutrition in Haitian preschool children is a general problem throughout the country, and no single anthropometric index can identify both the acute and chronic components of the problem. Acute undernutrition, as evidenced by substantial numbers of children who have low weight-for-height (wasting), is a major problem. During this survey, conducted during the summer months of 1978, 6.0% of the Table 7. Percentage distribution of preschool children being breast-fed at the time of the survey by age of child and geographical area: I Haiti 1978 Metropolitan Age (months) Nord-Ouest Nord L'Artibonite Ouest Sud Port-au-Prince Total 3 - 5 100.0 ( 44) 98.3 ( 60) 100.0 ( 55) 96.7 ( 60) 100.0 ( 56) 90.2 ( 61) 97.3 ( 336) 6-11 95.1 (103) 97.3 (113) 96.3 (108) 90.9 ( 99) 94.1 (135) 71.7 (113) 91.0 ( 671) 12 - 23 59.7 (181) 66.9 (175) 60.5 (195) 50.3 (177) 53.6 (237) 28.9 (166) 53.1 (1131) 24- 35 0.6 (157) 5.4 (184) 5.2 (153) 3.1 (162) 5.8 (121) 3.1 (161) 4.3 ( 938) 36- 59 0.4 (275) - (262) 1.1 (265) 0.4 (275) 1.2 (249) 1.6 (254) 0.8 (1580) Total 33.2 (760) 37.3 (794) 37.1 (776) 31.4 (773) 40.1 (798) 25.6 (755) 34.6 (4656) a See Fig. 1 for areas covered by sampling universes. All percentages are weighted by universe population proportions. The actual numbers of mothers questioned are given in parentheses. 762 HAITI NUTRITION STATUS SURVEY Table 8. Percentage distribution of survey children by stated age of weaning and by age of mother and by geographical area:" Haiti 1978 Age in months of child at weaning: Age of mother Never Mothers in (years) < 12 12 - 17 > 18 breast-fed Total age group (%) Representative rural sample < 25 19.5 42.8 36.9 0.9 100.0 13.2 52) (128) ( 113) ( 2) ( 295) 25 - 34 10.4 36.8 52.3 0.5 100.0 53.9 (127) (486) ( 698) ( 5) (1316) > 35 8.2 26.9 64.8 0.1 100.0 32.9 66) (243) ( 551) ( 1) ( 861) Total 10.9 34.3 54.4 0.4 100.0 100.0 (245) (857) (1362) ( 8) (2472) Urban sample (Port-au-Prince) < 25 51.9 20.7 12.6 14.8 100.0 24.6 70) ( 28) ( 17) (20) (135) 25 - 34 44.2 28.2 17.6 10.0 100.0 54.8 (133) ( 85) ( 53) (30) (301) > 35 32.7 34.5 29.2 3.5 100.0 20.6 (37) ( 39) (33) (4) (113) Total 43.7 27.7 18.8 9.8 100.0 100.0 (240) (152) (103) (54) (549) a Sea Fig. 1 for areas covered by sampling universes. All percentages are weighted by universe population proportions and are based on numbers of children with known age of weaning. Figures in parentheses are actual numbers of children surveyed. children were wasted (i.e., weight-for-height less than 80% of the reference median). An additional 28.8% of the survey children were classed as moderately undernourished (between 80.0 and 89.9% of weight- for-height reference median). The rural areas had higher prevalences of acute undernutrition than urban ones. All geographical areas had approximately the same prevalences of acute undernutrition. In Haiti the prevalence of acute undernutrition assumes major proportions during late infancy and in the second year of life. During this period, the nutritional needs of the child are high and the major source of food for the child begins to change from breast milk with some food supplementation to complete reliance on the family diet. Additionally, these children are more susceptible to acute infections and to the acute dehydration associated with these infections. In Haiti, there is little stunting during the first year of life, indicating that Haitian children are not born with a length/height deficit relative to the NCHS/CDC reference population. After 12 months of age, the deficit in height-for-age accelerates and by the second year of life (when Haitian women generally wean their children) over 19% of Haitian children show signs of stunting. Over the next 4 years of life the deficit of height-for-age continues for these children and affects many more children as a result of the continuing shortage-qualitatively and quanti- tatively-of food and to the infectious and parasitic disease processes. The special group of Haitian children taken from private paediatric practices and day care centres in Port-au-Prince show little stunting or wasting. Over 5'% of the children, however, were heavier than 120% of the reference median weight-for-height, an indi- cation of a potential obesity problem. The overall prevalence of bilateral pedal oedema is higher than that found in Nepal, Sri Lanka, and Togo (9, 10, 11) but is considerably lower than rates seen during famines or disasters, when rates as high as 20% have been reported (12, 13). If nutritional deficits are assumed to be responsible for all pedal oedema, then nearly 3% of Haitian preschool children had clinical evidence of kwashiorkor during the summer of 1978. By World Health Organization criteria (14), between 22 and 53% of Haitian children in various geographical areas are considered anaemic. The prevalence of anaemia is greatest in the 24-35-month age group. The specific etiology of this anaemia was 763 P. L. GRAITCER ET AL. not determined; however, nutritional deficiency of iron may be a contributing cause, as well as hook- worm, roundworm, and microbial infections. There are important differences in breast-feeding and weaning practices in rural and urban areas. Urban women, of all ages, tend to breast-feed for a shorter time than rural women and a higher percentage of urban women than rural women never begin to breast- feed their children. In both urban and rural areas younger women tend to wean their children earlier than older women. The causes of undernutrition cannot be docu- mented in a cross-sectional survey such as this one. The data collected do not enable us to distinguish between children who were undernourished because of decreased food intake or who were undernourished because of increased metabolic needs caused by infec- tions or parasitic infestations. The survey does, however, provide a baseline for quantifying the extent of undernutrition in Haiti. It also permits the quanti- fication of undernutrition in various geographical areas. For example, this survey has documented that contrary to local belief, nutritional problems are not significantly greater in the north and in the north-west than in the rest of the country. It has also demon- strated that acute undernutrition is a major problem in Haiti and that it is particularly pronounced in children 12-23 months old. A programme of nutritional relief directed nationwide at this age group will, it is hoped, reduce the dimensions of the undernutrition problem. As programmes are developed to respond to existing nutritional problems, surveillance indicators need to be identified that will predict short- and long-term changes in nutritional status within the country. Following the identification of patterns affecting nutritional status, it will be possible to make responsible decisions on the resources needed for short- and long-term intervention. RESUME ENQUETE SUR L'ETAT NUTRITIONNEL EN HAITI, 1978 Une enquete sur l'etat nutritionnel a ete effectu&e en Haiti de juin A septembre 1978. Elle a servi A mesurer quantitative- ment l'ampleur et la distribution de la malnutrition dans ce pays. La rapidite, la facilite et le faible cout eaient des facteurs pris en consideration dans la conception de cette enquete. Une importance particuliere a ete accordee A la qualite des techniques d'echantillonnage et A la bonne standardisation des mesures. Les parametres nutritionnels utilises dans l'enquete etaient les mesures corporelles du poids et de la taille, des renseignements fiables sur l'age et les taux d'hemoglobine. On a choisi dans tout Haiti 180 secteurs echantillons au moyen d'un echantillonnage systematique A point de depart aleatoire dans les listes fournies par le recensement de 1971. Dans chaque secteur, les donn6es ont e recueillies sur 30 enfants consecutifs Ages de 3 A 59 mois. Neuf cents enfants d'Age prescolaire et les personnes qui en avaient la charge ont e choisis dans chaque zone d'echantillonnage. Une d6nutrition aigue, definie comme un poids pour la taille inferieur A 80% de la mediane de reference, a e trouvee chez 6% de la population enquetee. Elle etait parti- culierement fr6quente chez les enfants de 12 A 23 mois. Plus de 40'! de la population 6tudi& presentait un retard de croissance ou une petite taille (moins de 90% de la mediane de reference relative A la taille pour l'Age). Ce phenomene etait le plus frequent dans le groupe d'Age 48-59 mois. Les taux d'hemoglobine ont e determines sur environ un cinquieme des enfants etudies. L'anemie etait la plus fr& quente dans la zone metropolitaine de Port-au-Prince. Elle etait maximale dans le groupe d'age de 24 A 35 mois dans l'ensemble du pays. Sur un <groupe speial> compose d'enfants appartenant A des familles urbaines de Port-au-Prince ayant les moyens de leur fournir une bonne nutrition et les soins d'un medecin exercant en clientele privee, les donnees anthropometriques ont e recueillies, analysees et comparees A celles de l'enquete rurale. Dans ce groupe, les parametres anthropo- metriques etaient de toute evidence nettement meilleurs. Ces valeurs representent un etat nutritionnel susceptible d'etre atteint pour tous les enfants d'Haiti. Bien qu'il soit impossible de reunir une documentation sur les causes de la denutrition dans une enquete transversale, les donnees de base recueillies prouvent nettement que la denutrition aigue constitue un probleme majeur en Haiti et qu'elle est particulierement prononcee chez les enfants de 12 A 23 mois et uniformement repartie dans tout le pays. 764 HAITI NUTRITION STATUS SURVEY 765 REFERENCES 1. JELLIFFE, D. B. & JELLIFFE, E. P. The nutritional status of Haitian children. Acta tropica, 18: 1-45 (1961). 2. BEGHIN, I. ET AL. L 'alimentation et la nutrition en Haiti, Paris, Presses Universitaires de France, 1970. 3. SEBRELL, W.H. ET AL. Appraisal of nutrition in Haiti. American journal of clinical nutrition, 7: 1-8 (1959). 4. SERFLING, R. E. & SHERMAN, I. L. Attribute sampling methods, Atlanta, Department of Health, Education, and Welfare, Public Health Service, Center for Disease Control, 1965. 5. MILLER, D. C. ET AL. Simplified field assessment of nutritional status in early childhood: practical suggestions for developing countries. Bulletin of the World Health Organization, 55, 79-86 (1977). 6. National Center for Health Statistics. Growth charts, Rockville, US Department of Health, Education, and Welfare, Public Health Service, Health Resources Administration, 1976 (Serial No. HRA 76-1120, 25, 3). 7. GOMEZ, F. ET AL. Mortality in second and third degree malnutrition. Journal of tropical pediatrics, 2: 77 (1956). 8. WATERLOW, J. C. & RUTISHAUSER, I.H. Malnutrition in man. In: Cravioto, J. et al., ed., Early malnutrition and mental development. Proceedings ofthe Symposium of the Swedish Nutrition Foundation, No. XII, Sweden, Almquist & Wiksell, 1974, pp. 13-26. 9. BRINK, E. W. ET AL. Nutrition status of children in Nepal, 1975. Bulletin of the World Health Organization, 54: 311-318 (1976). 10. BRINK, E. S. ET AL. Sri Lanka nutrition status survey, 1975. International journal of epidemiology, 7: 41-46 (1978). 11. Togo nutrition status survey, Washington, DC, Office of Nutrition, Agency for International Development, 1978. 12. KLOTH, T.I. ET AL. Sahel nutrition status survey, 1974. Americanjournal ofepidemiology, 31: 383-390 (1976). 13. HOGAN, R. C. ET AL. Sahel nutrition surveys, 1974 and 1975. Disasters, 1: 117-124 (1977). 14. WHO Technical Report Series, No. 503, 1972 (Nutritional anaemias: report of a WHO Group of Experts).