During the late 1950s the Kariba hydro-electric dam was constructed on the border of Zambia and Zimbabwe forcing the relocation of 57,000 people, mainly Tonga. As part of a larger study to assess the effects of the relocation, research into the human biology of the Tonga people was conducted. The research reported here provides a basis for comparison with long-term follow-up data on growth and physical status of Gwembe Tonga to determine the effects of resettlement. The sample consists of 303 schoolchildren, 7 to 13 years, from two schools that were not relocated and three schools that were to be relocated. Homogeneity of the two groups led us to combine them as a single baseline sample before relocation. Comparisons with NCHS (National Center for Health Statistics) reference data and with contemporary urban data from southern Zambia indicate sub-optimal nutritional status. After ages 6 and 7, height-for-age and weight-for-age Z-scores of boys decline steadily towards -2.0 SD throughout the 12th year, whereas mean Z-scores of girls decline markedly from 8 years on. Mean HAZ (Height-for-age-Z-score) of girls falls below -2.0 SD by 11 years and approaches -3.0 SD by 13 years. Thirty-nine percent of males and 47% of females in the baseline sample showed effects of moderate or severe protein energy malnutrition at the time of relocation. Because these children were from a school sample, gender differences in opportunity to attend school may be a factor in what appears to be gender bias, favoring the status of boys. This sample provides a baseline for assessing the long-term impact of forced relocation on the Gwembe Tonga.

In situations where birth records are unavailable and stated ages are unreliable, the emergence of the permanent dentition can serve as an indicator of age. Due to substantial variation in the timing of tooth emergence, a sample (n = 721) of Zambian school children, with known ages, was examined to provide a tooth emergence reference standard for the area. Three methods for assigning ages were utilized and their accuracy assessed. A random test sample was withheld from the original study in order to further evaluate the methods' accuracy. The three methods-1) number of teeth, 2) regression and 3) probit analysis-were applied to Zambian children, and estimates of age were made. Predicted ages were compared to actual ages to determine the percentage of accuracy in three categories-(+/-) .5, +/- 1.0 and +/- 2.0 years- and paired t-tests were conducted. Each of the three methods was then applied to the test sample, and their accuracy was evaluated in the same manner. Methods 1 and 2 were found to provide the higher percentage of correct ages within +/- .5 years, assigning roughly 39% of both male and female children within this increment. This was also the case at the next increment, with methods 1 and 2 assigning a higher percentage (66-76%) of children to the +/- 1.0 year category, while the accuracy of method 3 was quite a bit lower. The results for the test sample were very similar to those of the main sample. The overall accuracy of methods 1 and 2 was very similar in both the main and test samples, while method 3 had lower accuracy and t-tests indicated significant differences. Therefore, due to ease of application in the field setting, method 1, mean age per number of teeth emerged, is the method of choice.

There is sufficient variation in the timing of permanent tooth emergence to warrant population specific standards. Due to the lack of reliable ages in rural Zambia, a tooth emergence standard is derived from an urban sample of 543 Zambian schoolchildren. The standard allows assignment of an age to a child based on the total number of permanent teeth emerged. As a way to evaluate the ability of the standard to assign correct ages to Zambian children, a randomly chosen test (10%) sample was withheld from the formation of the standard. Accuracy of the standard in assigning ages to children in both the test and main samples was assessed in ±0.5, ±1.0, and ±2.0 year intervals. Both the main and test samples were able to accurately assign correct ages within ±0.5 year ∼40% of the time and within ±1.0 year ∼65% of the time for both males and females. To further facilitate use in field settings, the standard was visually smoothed. Once again the accuracy of the standard was very similar when applied to the main and test samples. The smoothed standard was as accurate as the original standard. The smoothed standard is thus recommended for application to rural Zambian children of unknown or unreliable age. Am. J. Hum. Biol. 10:45–51, 1998. © 1998 Wiley‐Liss, Inc.

No abstract