Caries risk factors in the permanent dentition of Tanzanian children: a cohort study (1997-2003)

The aims of the cohort study were to study the association between permanent dentition caries and malnutrition and other risk factors or indicators in a group of children with little or no access to restorative and preventive dental care.

The study was conducted at two primary schools in Dar es Salaam, Tanzania in the period 1997-2003. One school recruits its children from affluent families and the other school is attended by the children of poor families. The children attended grade 1 at the first examination and had a mean age of 7.6 years (SD = 0.4); when the study was completed, the mean age was 13.3 years. Two-hundred and eighteen children examined in 1997 were re-examined in 1999; 147 of the children were examined in 1997 as well as in 2003, and 122 children were examined in all three years. An age- and sex-based body mass index (BMI) was computed to determine the nutritional status of each child. Each year, the same standard methods were used to determine the count of lactobacilli and mutans streptococci, stimulated flow rate and buffering capacity of saliva, and caries. Risk ratios were computed with generalized linear models using the tooth as the unit of analysis.

Mean annual caries increment was 0.27 in the period 1997-1999 and 0.80 in the period 1999-2003 with most children developing no caries at all. Malnutrition at baseline in 1997 was insignificantly predictive for the development of caries, whereas a low stimulated flow rate of saliva (< or =0.7 ml/min) and a high count of lactobacilli (> or =100,000/ml) at baseline in 1997 were significantly associated. However, the generalized linear models for the two time periods 1997-1999 and 1999-2003 presented a confusing picture with different risk ratios and without consistency of the associations between the exposure variables and the development of caries. In accordance with this finding, the consistency of the exposure variables over time for the individuals was very low.

The results were inconclusive and left us with more questions than answers. The findings do, however, support the view that our methods for predicting caries are inappropriate or nonexistent.