Contents - Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Sampling and data collection
All women who were pregnant or lactating and all children from birth to seven years of age living in the study villages between 1 January 1969 and 28 February 1977 were included in the original design of the study. Some data were also collected for fathers, but no supplementation information was recorded for them. Not all of those identified by the censuses and their fortnightly updates were actually measured. Most often, this was because the family remained in the village only a short time. The other reason, refusal to participate in the study, was rareless than 2%. Thus, anthropometric data are available for 517 women for one or more of their pregnancies and corresponding lactation periods.
Table 4 shows the distribution of the 1,992 children with one anthropometric examination or more by birth-date cohort and village type. The cohorts identify children with different exposures to supplementation during the critical periods of gestation and the first three years of life.
TABLE 4. Distribution of subject children by birth cohort and type of supplementation
|Birth cohort||Exposure period||No. of children|
|1 Mar. 1974 or after||All||Partial||260||280||540|
|1 Mar. 1969-28 Feb. 1974||Partial||All||374||395||769|
|1 Jan. 1966-28 Feb. 1969||None||Partial||185||194||379|
|Before 1 Jan. 1966||None||None||151||153||304|
Data collection and reliability
Census information and data on supplementation, home diet, morbidity, clinic visits, anthropometry, and mental development in pregnant and lactating mothers and in their children 0-84 months old were collected longitudinally from 1969 to 1977 (table 5). Most data collection began 1 January 1969 in the large villages and somewhat later (March-May) in the small villages; it ceased in August 1977 for most data but in March for some. The data collection is described in detail in a manual of standard operating procedures , in which the data-collection forms used are also presented. The table also lists data collected in cross-sectional surveys conducted in collaboration with the Rand Corporation in 1974-1976; additional details are given elsewhere .
Short descriptions of the data-collection methods are presented here. Originally all scheduling for longitudinal data collection was programmed in the village itself using a simple data sheet that represented all children and pregnant women chronologically according to birth date or date of conception respectively. The dates for scheduled examinations were then noted in a line for each examinee. Visual inspection of the dates crossed out with different colours according to the timeliness of the examinations gave immediate information about who had not been examined on time and examinations missed. Later the programming of examinations and the quality control of coverage were done by computer. This permitted easier central compilation but entailed less immediate feedback about missing examinations.
Census and socio-economic data
A census was conducted in the four villages at the end of 1968. This was updated whenever the dietary-morbidity interviewers in their fortnightly home visits identified changes in household composition, new families (new marriages/unions or in-migrants), changes in residency within the village, and out-migration. This updating of the census was verified by a cross-sectional census in 1974.
The informant was the mother or other primary caretaker. The data collected included information about family structure, marital status, religion, number of pregnancies, number of live children, relation of the nuclear family to the head of the extended family, and ownership of items such as radios, sewing machines, refrigerators, bicycles, and motor vehicles. Information about parental literacy was obtained through interviews and also through tests. Status (e.g., alive, dead, immigrated), birth order, kinship (e.g., father, son, adopted), parity (for mothers), schooling, and occupation (for those over ten years old) were recorded for each household member. Observations were noted about the house, such as the types of walls, floor, and roof, the availability of electricity, the type of water source, and grey-water and faeces disposal. Whenever a change occurred in the location of the home or in the status or kinship of a family member, the date of the change was noted.
Full descriptions of the psychometric tests are given elsewhere [21, 22]. Neonates were tested within 10 days of birth with the Brazelton neonatal scale, and then at 6, 15, and 24 months with a scale composed of items compiled from the Bayley, Cattell, Gesell, and Merrill-Palmer infant scales. These data were collected from July 1969 through December 1972.
From 3 to 7 years of age, the children were tested annually on a battery of 24 tests chosen to tap memory, language, perceptual reasoning, learning, and abstract reasoning ability. These data were collected from March 1969 to February-May 1977 depending on the village.
The components of variance approach was applied to the standardization  and to description of data quality  of continuous indicators of diet and nutrition status in the study, using an approach that had the advantage of well-developed statistical techniques and a well-defined definition of reliability (square of the correlation coefficient between replicate estimates in the same children over a time span of nutrition relevance). As a result of careful attention, the data collected proved to be of a quality equal to or better than that achieved elsewhere [25, 26]. The approach proposed by Habicht  differentiated between components of unreliability due to measurement imprecision and those due to day-today variability, and allowed for comparisons of statistical power, or the ability to identify a difference, among indicators of nutrition status . This conceptualization also fostered the differentiation  of biases likely to cause confounding (associated with but not due to nutrition status) from those likely to hide actual associations because of unreliability and other random influences.
Supplementation and measurement of attendance and consumption began on 1 January 1969 in the large villages and 1 May 1969 in the small villages. Attendance at the feeding centre was recorded for all sample persons. The supplement was poured into cups calibrated to 180 ml. The cups were filled as often as requested. Individual intake was measured by recording the number of cups given and subtracting any leftover supplement measured to the nearest 10 ml.
All leftovers were poured into a large vessel. At the end of the serving period the total amount recorded as ingested was compared with the total amount served minus the volume of leftovers. Calculations based on this comparison showed that the reliability of the records of the amount of supplement ingested was better than 99% per cup served. The reliability of concern in this study is of intakes over a week or longer periods; that reliability is almost perfect.
The home diets of the children and the pregnant and lactating women were collected according to the schedule indicated in table 5. Because of financial constraints, information on the diets of infants was not collected from 1970 to 1973; this was considered a serious omission at the time despite the difficulty of interpreting such data in mostly breast-fed infants. Fortunately, funds were obtained from the Rockefeller Foundation to allow for the collection of data on infants' diets from 1973.
TABLE 5. Data collected in the INCAP longitudinal study, 1969-1977
|Type of data||Targets||Frequency or time of collection||Where collected||Collector||Dates|
|Supplement intake||Children 0-7 yrs; pregnant & lactating women||Daily||Feeding centre||Supplementation supervisors||1969-77|
|Morbidity, breast-feeding, menstruation recall||Children 0-7 yrs; pregnant & lactating women||Every 15 days||Home||Dietary and morbidity interviewers||1969-77|
|Diet: 24-hr recall in large villages; 72-hr recall in small villages||Children
Children 15-36 mo
Children 42-60 mo
Every 3 mo
Every 6 mo
Every 3 mo
Every 3 mo
3, 6, 9, 18 mo postpartum
|Home||Dietary and morbidity interviewers||1973-77
|Birth weight, birth process, Apgar scores||Newborns||At birth||Home||Perinatologist||1969-77|
|Anthropometric indicators of nutrition status||Children
15 days-24 mo
Children 30-48 mo
Children 60-84 mo
Pregnant & lactating women
Every 6 mo
|Clinic||Child anthropometrist Maternal anthropometrist||1969-77
|Hand-wrist x-ray||All children||Age
Every 6 mo, ages 6-48 mo
60, 72, 84 mo
Children 48-84 mo
|Test room in village||Psychometrist||1969-77|
|Physical examination||Children 0-7 yrs||15 days; 3 mo; 1, 3, 7 yrs||Clinic||Paediatrician||1971-77|
|Prenatal examination||Pregnant women||Each trimester of pregnancy||Clinic||Auxiliary nurses||1969-77|
|Reproductive history||Pregnant women||First prenatal examination||Clinic||Auxiliary nurses||1969-77|
|Record of visits to outpatient clinic||Children 0-7 yrs and pregnant & lactating women attending the medical clinic for any reason||All visits for therapeutic care, Mon-Fri||Medical clinic||Auxiliary nurses||1971-77|
|Census and socio-economic information||Families in community||Twice||Home||Census interviewers||1968-69
|Changes in household composition through census updates||Families in community||Every 15 days||Home||Dietary and morbidity interviewers||1968-77|
|Fathers' anthropemetry||Fathers of children in study||Once (cross-sectional)||Home||Child anthropo-metrist||1973-75|
|Retrospective life history of women (fertility, infant mortality, employment)||Women 15-49 yrs ever in a marriage/union or ever a mother||Once (cross-sectional)||Home||Interviewer||1974-75|
|Income and wealth||Heads of households||Once (cross-sectional)||Home||Interviewer||1974-75|
|Attitudes and expectations of women||Women 15-49 yrs ever in a marriage/union or ever a mother||Once (cross-sectional)||Home||Interviewer||1975-76|
|Attitudes, expectations, and retrospective life history of men||Half of husbands of respondents to women's retrospective life history, and half of all single men||Once (cross-sectional)||Home||Interviewer||1975-76|
The daily home diet was estimated from 24-hour-recall surveys in the large villages and three-day-recall surveys in the small villages. The amounts of food ingested were recorded in grams according to the usual INCAP recall method  and were converted to energy and nutrients using the INCAP food composition tables [29, 30], at first by hand and then by computer.
There was concern that the dietary-recall method might be unreliable as well as systematically or proportionately underestimating intake, because intake data often seemed lower than appeared physiologically compatible with survival. Poor estimation of the home-diet data would matter when adding the correctly measured contributions of the supplements to estimate the total diet because the sum would be unreliable and decrease the power of analyses. However, the decrease in power could be avoided by using the supplement and the home diet as distinct variables so long as the home diets were reliably measured and either correctly estimated or proportionally or systematically underestimated .
The reliability of the dietary data was, however, found to be poor [9. 11, 24, 31] . Energy and protein had the highest reliabilities, but these were only of the order of .15-.30. The reliability for vitamin A was no different from zero even when the non-Gaussian distribution was normalized. The other nutrients examined, riboflavin and iron, had intermediate reliabilities. These findings are similar to those reported elsewhere subsequently .
Reliability analyses over intervals of three months showed that, for energy and for the most reliable nutrient, protein, at least twice as many children would be needed as would ever be available to show stable associations between home dietary protein and growth in length , the latter being the second most reliable indicator measured in the study . (Supplement intake was more reliable.) Length was also the most responsive indicator to variations in supplement intake  and in the prevalence of diarrhoea . Psychometric indicators could be expected to be much less sensitive to nutritional effects than length, and therefore even more children would be necessary to demonstrate nutritional effects of diet on mental development.
Improvement of the precision of the dietary method was considered. At the time, the dietary registry method was considered the gold standard for large-scale surveys because it produced data similar to those obtained by weighing . The registry method is much more labour-intensive than recall because it entails weighing foods in the same portions as eaten, albeit after the fact. However, a comparison of recall and registry methods showed no differences in variability over a six-month interval . Therefore, it was concluded that the registry method was no more reliable than the recall method.
Inspection of the sources of this unreliability  showed that at most 52.5% could be explained by imprecision of the measurement, and that presumed that the registry method was absolutely precise. In fact, one can be sure that imprecision accounted for much less than half of the unreliability. When half or more of unreliability is not due to imprecision, improving precision must be very inexpensive to be cost effective . In this case, improving the method to perfection would not have made much difference in the number of children required to show the associations of interest.
Improving reliability by more frequent measurement is the preferred strategy when unreliability is not due to imprecision. Extending the recall periods from one to three days did not, however, improve reliability because of carry-over recall from day to day. Increasing the frequency of the surveys four to nine-fold would be necessary to obtain adequate power for statistical analysis of home diets' contribution to nutritional outcomes, a desirable but financially impossible option. Thus, an understanding was reached that the home-diet data were unlikely to give much insight about the relationship between energy or nutrient intake and any outcome variables.
The data were useful to decide that the dietary intake of a nutrient was adequate if most of the children's intakes were above the recommended daily allowances. However, it was concluded that one could make no inferences about dietary inadequacies in this population from dietary information and the published recommended dietary allowances alone. Fifteen years later the US National Academy of Sciences , using the same methodology, came to the same conclusion for all populations.
The mean home-diet data were also useful for descriptive purposes and for face validity analyses of the study and its reports. That is because mean data from a group of people may be reliable even though the individual data for home diet were unreliable. This is due to the relationship of the reliability of the mean (Rm) to the reliability of the individual measurements (Ri) and to the number of individual measurements (n) that contribute to the mean such that Rm=1-[(1-R)/n]. Thus, the mean of 10 persons, each measured with a reliability, Ri, of .15, has a Rm of better than .90. If the mean is of 100 persons, the Rm is better than .99.
Records of complaints, diagnoses, and treatments were kept for all visits for use in the quality-control system but, unfortunately, are no longer available. As of 1971, symptomatology was recorded for each visit on the same form as the fortnightly morbidity data collected in the home, and those data are available.
Morbidity, breast-feeding, and menstruation
Morbidity data were gathered every 14 days through retrospective interviews of mothers in the home by four home visitors, one for each village. The visitors were rotated periodically among the villages to offset interviewer bias. Interviews took place Monday through Friday, the families being so divided that the entire population of each village was interviewed routinely every two weeks. During the interview, the mother was asked to recall any symptoms that she and any of her children younger than seven years had in the previous two weeks. Each subject's information was recorded on a separate form. The beginning and ending dates of a symptom were always noted. A routine quality-control system was applied, allowing the method to be standardized, using a supervisor, and validated, using a physician. Information on menstruation (beginning and ending) were recorded or noted as absent on the form for the youngest child. Also noted was whether or not the child was breast-fed.
To validate the morbidity survey, a physician examined children half a day after the morbidity visit without previously informing the interviewer. This generally showed satisfactory sensitivities and specificities for the symptoms recorded. Sensitivity and specificity were 66% and 99% respectively for diarrhoea, and 75% and 99% for fever .
A study of the prevalences of recalled symptoms over the 14 days between periodic surveys showed a fall in prevalence with respect to the day of interview, indicating memory loss over the two-week period . The consequent average underreporting varied from 4% for long-term skin infections to 37% for fever. Forgetting was closely associated with the duration of the episode, with a correlation coefficient of -.81 (p < .01) between average underreporting and the duration of the episode. Despite an average underreporting of 22%, diarrhoea was reliably enough reported to reveal statistically significant associations between percentage of time ill and growth .
Body measurements were taken at specific ages by trained and standardized anthropometrists. Originally anthropometry was the responsibility of the auxiliary nurse in each village, but early in the project two individuals were hired to measure mothers and children respectively. A single person measured the mothers throughout the study, but three measured the children at different times (1970-October 1973, October 1973 - October 1976, October 1976-1977). All changes in personnel were preceded by rigorous standardization sessions.
Each week the data collected in the field were brought to the INCAP headquarters, computerized, and analysed. All children with values beyond two standard deviations from the age-specific means were remeasured for all variables to determine whether or not errors had been made in measuring, recording, or punching. There were weekly calibration of instruments, frequent standardization exercises for the anthropometrist at an urban orphanage, and field replications. The exercises permitted monitoring of precision and reliability .
The anthropometry standardization method  is widely used today. Comparison of the reliability and precision achieved during the longitudinal study with results from other studies [35, 25] speaks favourably of the quality of the INCAP data.
The anthropometrist also took a radiograph of the left hand and wrist of children using a General Electric machine (model 100-15) set at 15 mA and 65 kV and using power from a gasoline generator. The x-ray head was set at 76 cm above the third metacarpal of the left hand with the fingers moderately splayed and the forearm placed at right angles to the x-ray beam. Exposure was for one forty-eighth of a second for children under two years of age and one second for older children. Roentgenograms were taken at the ages specified in table 5 concurrently with anthropometry.
Great care was taken to protect the children from stray x-rays. The film was placed in a lead-lined box attached to the head of the radiograph machine. The seated child placed his or her hand into the box through a lead curtain. A film was placed on the child's seat and developed every month to be sure that there was no stray radiation.
The films were read for the number of ossification centres and the thickness of compact bone. Initially, bone age was also determined according to the Tanner-Whitehouse and Greulich and Pyle methods. These detailed assessments were discontinued when analyses showed that the simple counting of ossification centres provided as much information .
As mentioned above, Dr. Carlos Beteta examined children at 15 days and at 3, 12, 36, and 84 months of age to identify developmental and other problems. He gave special attention to neurological function and minor anomalies indicative of congenital mental retardation or neurological impairment. This information would help to identify children whose association between less adequate nutrition and impaired mental development was probably due to the latter, impaired behaviour leading to inadequate bonding and poor coping, both of which might result in malnutrition. It was also thought that the neurological data might reflect improved nutrition from the supplement. None of these data have been analysed except for an unpublished study (Beteta, personal communication) showing that a high prevalence of apparent phimosis at 15 days of age becomes negligible by the age of seven years.
Contents - Previous - Next