The general objective of the 1972–1974 National Growth and Development Study was to obtain information on the physical growth and development of the Cuban population from birth to 20 years of age, through a random and representative sample of the population of the entire country that would allow comparisons among different population groups, localities, geographic areas and other variables.
The specific objectives of the study included obtaining national standards for growth distance and velocity with respect to the following elements: weight; length in the supine position and standing height; vertex-ischion length; sitting height; head, arm, thigh and leg circumferences; three skinfolds: triceps, subscapular and suprailiac; biacromial and biiliac diameters; and foot length.
In addition, stages of sexual development were investigated in both sexes, as well as the age of menarche in girls. In 10% of the sample, a hand X-ray was performed to obtain data on bone maturation. Information on genetic influence was collected by measuring the height of both parents, and information on environmental factors was gathered by recording economic and social variables, such as income, the child’s birth order, family size, parents’ occupation and educational level, among others.
All of this was intended to relate anthropometric variables to genetic and environmental factors. Based on this information on the epidemiology of growth in different areas of the country, priorities could be established when allocating resources for development.
The study was planned in two stages: a cross-sectional stage, from May 1972 to April 1973, and a longitudinal stage, from May 1973 to January 1974. The design consisted of a probabilistic, multistage and stratified sample of non-institutionalized individuals. The 1970 National Population and Housing Census was used as the sampling frame.
The theoretical sample size was 55,000 individuals, in order to establish, within relatively narrow limits, with a standard error of ±0.3 cm in height, sufficient precision for the extreme percentiles of 3 and 97. These percentiles are the most important for establishing the cutoff points that define supposedly normal or typical cases and those in the risk zone.
Between the ages of 3 and 9 years, average annual growth in height is about 5 cm. For each 5 cm increase, a sample composed of 1,000 boys and 1,000 girls is required in order to obtain a standard error of ±0.3 cm.
Before the age of three and during adolescence, annual growth is greater than that length. Therefore, whenever the children in the sample experience an approximate growth of 5 cm, regardless of the time in which they achieve it, there must be 1,000 children of each sex to maintain the same precision in the extreme percentiles.
Since from birth to adulthood human beings grow in length by about 125 cm, 50 cm at birth and 175 cm at the end of growth in males, those 125 cm contain 25 segments of 5 cm each. Since each segment requires a sample size of 1,000 for each sex, the total sample size would be 50,000 individuals: 25,000 boys and 25,000 girls.
If a 10% increase is added to that figure to account for possible non-attendance, the final theoretical sample is reached: 55,000 individuals. The strata were: province, urban or rural locality and ethnic group. The primary sampling units were areas, sectors or individuals under 20 years of age residing in them.
A total of 440 sectors were selected at random and were visited, as part of voluntary work tasks, by members of mass organizations: the Federation of Cuban Women, the Committees for the Defense of the Revolution and, in rural areas, the Association of Small Farmers.
In this way, a total list of 126,720 individuals was prepared, from which the 55,000 who were to be called for measurements were randomly selected with the help of a computer. The population response was 93% attendance and 4% non-attendance with justified reasons. These percentages demonstrated good work by the popular organizations.
The study used the measurement technique and instruments recommended by the United Nations International Biological Program and the International Children’s Center in Paris for recording anthropometric dimensions.
For sexual development, the system of pubertal stages proposed by J. M. Tanner was used. Menarche was investigated in girls using the status quo method with logit analysis; cumulative percentages were calculated and the corresponding percentiles were adjusted.
For bone maturation, nearly 5,000 X-rays of the left hand and wrist were taken and later evaluated by the same observer following recommended techniques. Preliminary tests were carried out by placing dosimeters on the gonads, sternum and hand. Radiation was found to be zero in the first two locations, and an average of about 30 to 40 mrad was observed in the hand. This operation was carried out with authorization from the National Committee for Radiological Protection.
One measurement team was assigned to each of the country’s six provinces, plus an additional team for the two most populated provinces, Havana and Oriente. Each team was composed of a chief, a statistician, two anthropometrists and an X-ray technician. Each had a laboratory truck, a vehicle for personnel, an X-ray machine and a portable power plant.
One year before the start of the survey, a pilot study was carried out to evaluate the operation of the system. That study showed the need for laboratory trucks, appropriate vehicles for personnel, portable power plants and radiographic films without chassis or cassette. It also defined the average number of children who could be examined daily.
An itinerary calendar was prepared, and it proved to be accurate, since it was completed in full during the study. Over the 20 months during which the field stage of the research lasted, 11 quality controls were carried out, approximately one every two months.
To carry out these controls, all teams met in the same place and measured the same group of children. This made it possible to compare the results obtained by the measurers within each team and with those from each province, in order to determine precision, consistency over time and differences in results that required adjustments.
It should be mentioned that the results of the controls were always very satisfactory. In any large-scale anthropometric study, it is necessary to review the data to avoid errors of different kinds. For this purpose, a computer program was prepared and rational limits were established for each of the anthropometric dimensions recorded.
Any value outside those limits was listed as an error, and was analyzed and verified against the primary data from the corresponding form. Around 200 cards were found with one or more punching errors. In others, the primary data recorded a value that was impossible to accept.
The total number of cards that made up the error-free file was 50,260. The data were transferred to the computer’s magnetic tapes and, from there, all the tables and curves were constructed.
