Nutritional status and dental caries in a large sample of 4- and 5-year-old South African children
P Cleaton-Jones, BDS, MB BCh, PhD, DSc (Dent)
Ruth Sinwell, Dip Home Econ
Mirriam Mogotsi (Formerly, Research Institute for Nutritional Diseases, Medical Research Council, Johannesburg)
Barbara D Richardson, DSc
Medical Research Council/University of the Witwatersrand Dental Research Institute, Johannesburg
Lars Granath, DDS, DOdont
Department of Pedodontics, Lund University, School of Dentistry, Malmo, Sweden
L Paul Fatti, MSc, PhD, DIC
Department of Statistics and Actuarial Science, University of the Witwatersrand, Johannesburg
Alexander R Walker, DSc
Human Biochemistry Research Unit, School of Pathology, South African Institute for Medical Research and the University of the Witwatersrand, Johannesburg

S A J Clin Nutr 2000 August Vol. 13 No 3.

Background
Evidence from studies involving small samples of children in Africa, India and South America suggests a higher dental caries rate in malnourished children. A comparison was done to evaluate wasting and stunting and their association with dental caries in four samples of South African children.

Design
Cross-sectional study based on random sampling of birth records of two age bands.

Methods
A total of 2 728 4- and 5-year-old South African children from one rural community and three urban communities were examined for nutritional status and dental caries.

Results
In the total sample prevalences of wasting were mild (28%), moderate (4%) and severe (2%). For stunting the prevalences were mild (13%), moderate (3%) and severe (1%). For both conditions rural children showed higher proportions than the other groups. Statistical analysis showed statistically significant differences for wasting and stunting between the study groups. No significant association was found between the prevalence of caries and stunting or wasting, but an association was noted between wasting and decayed, missing and filled (dmf) surfaces (P = 0.003).

Conclusions
In the series of children studied, nutritional status was not found to be clinically relevant to dental caries prevalence and experience.

S Afr Med J 2000; 90: 631-635

Alvarez and Navia1 reviewed the interaction between nutritional status, tooth eruption and dental caries and highlighted the difficulty of demonstrating a clear correlation between malnutrition and increased dental caries, in contrast to the clear effects demonstrated in animal experiments. They showed that delayed tooth eruption was associated with malnutrition in industrialised as well as in developing countries. The authors provided a hypothetical caries prevalence versus age curve that might be seen as a cross-sectional survey of 6- to 15-year-old children in a population in which normal growth and mild-to-moderate stunting of growth coexist. They postulated that due to delayed eruption of the permanent teeth, more caries would be seen in the primary dentition at a particular age in malnourished children, and less in the permanent dentition, when compared with their well-nourished counterparts. Evidence to support this concept has come from cross-sectional surveys undertaken on Peruvian children2,3 as well as from a longitudinal survey of the primary teeth, also done on Peruvian children.4 The most recent study in the latter group5 has confirmed the higher caries rate in malnourished children.

In India, salivary flow was reported to be reduced among malnourished children,6 probably from alteration in the salivary glands due to chronic protein energy malnutrition (PEM), an effect shown in animal studies.7 Children with this reduced salivary flow had decreased salivary buffer capacity and increased dental caries.

In Africa, 1-5-year-old children classified clinically as malnourished had more caries than those regarded as normally nourished;8 however, among 10-12-year-old children those below and above the 5th percentile of weight for age had similar caries rates.9

Of the studies reported so far, sample sizes have been small, or when moderately large, they have been spread over a wide age range. This paper reports findings on the relationship between nutritional status and caries in a large series of South African children within a narrow age band.

Methods
The sample studied consisted of 2 728 children aged 4 or 5 years. One rural group was from the Gelukspan district in North West, approximately 350 km west of Johannesburg, an area with a fluoride concentration in the drinking water of 0.13 - 0.22 ppm.10 In Gauteng urban groups came from three areas, namely Soweto, Lenasia and Johannesburg. The water supply for these three groups was the same and contained 0.20 - 0.33 ppm fluoride.11 The four groups came from areas that under apartheid legislation had corresponded with residential areas for blacks (Gelukspan, Soweto), Indians (Lenasia) and whites (Johannesburg). In the current study the areas were used as proxy indicators of broad social conditions.

Following clearance by the Committee for Research on Human Subjects of the University of the Witwatersrand, random samples of children were drawn from birth records at well-child clinics (where all birth records were stored), the target being 700 for each of the four groups. Gender sub-samples were approximately equal in each group.

Oral examinations were carried out by calibrated dentists12 in good natural light using plane mirrors, probes (explorers) and World Health Organisation (WHO) caries diagnosis criteria.13 Dental caries was diagnosed when there was clinically detectable loss of tooth substance and when such loss had been treated with fillings or extraction. The sum of decayed, missing and filled (dmf) tooth surfaces for an individualÕs primary teeth constituted the dmfs score, with a maximum possible score of 88. The height of the children, erect and barefoot, was measured with a height stick to the nearest 0.5 cm, and weight was measured using a bathroom-type scale to the nearest 0.5 kg. For calibration, this scale was checked at frequent intervals against an accurate beam balance laboratory scale (Detecto Personal Weigher, Detecto Scales Inc.; Brooklyn NY, USA).

For stunting (chronic malnutrition) and wasting (acute malnutrition) the classification of Waterlow14-16 was used together with the National Center for Health Statistics (NCHS) tables.17 Using a percentage of the NCHS median (50th percentile) the classification was as follows. For stunting, the height-for-age ranges were: normal (³ 95% of the median), mild (³ 90% - < 95%), moderate (³ 85% - < 90%) and severe (< 85%). In the case of wasting, weight-for-height groupings were: over-nutrition (³ 110%), normal (³ 90% - < 110%), mild (³ 80% - < 90%), moderate (³ 70% - < 80%) and severe (< 70%).

For a within-study comparison using a common scale and common median, the median height and weight were determined for the combined study sample. Each individualÕs height and weight were then expressed as a percentage of these sample medians.

The data for both age groups were pooled for analysis using the Statistical Analysis System (SAS).18 In analysing the stunting and wasting categories, the chi-square test was used and BonferroniÕs significance levels18 were included when pairwise comparisons of study groups were made. The critical levels of statistical significance were set at P < 0.05 for the simultaneous comparison of all groups and at P < 0.01 for the six pairwise comparisons.19 The individualÕs percentage of median height and percentage of median weight were analysed using the general linear model method. For the association between dental caries, stunting and wasting, the Catmod Procedure was used for presence or absence of dental caries, and the Kruskal-Wallis non-parametric test was used for dmfs scores. The null hypotheses set were: (i) that malnutrition is not different in the four study groups; and (ii) that malnutrition is not associated with dental caries.

Results
Detailed anthropometric results for each age, gender and group have been provided in a reference data base10 that is available from the authors.

The frequency of wasting in each study group is shown in Table I. Over-nutrition was seen in almost 15% of the Johannesburg sample and in 9% of the Soweto children but was uncommon in the other two groups studied. Mild wasting was seen in 28% of the total sample. It was higher among Lenasia (42%) and Gelukspan (43%) children, with the lowest frequency being seen in the Soweto group (16%). Moderate wasting was present in 7 - 8% of Lenasia and Gelukspan children compared with 1% in the other groups. Severe wasting was uncommon in all of the groups, with the highest frequency occurring in the Lenasia children (2%).

Stunting was uncommon (Table I); only 18% of the children showed some degree of stunting, most of which was mild. Within the study groups, stunting was most frequent among the Gelukspan and Soweto groups.

Chi-square tests for overall differences in the frequency distributions of wasting and stunting among the four study groups revealed highly significant differences among them for all groupings. When the classifications of wasting and stunting were combined, a pattern similar to wasting alone was seen. In general, Soweto and Johannesburg childrenÕs results were similar.

When the within-study comparisons were done with the general linear models analysis for percentage of median weight, the study group showed a highly significant effect (F = 232.98, P < 0.001) as did percentage of the median height (F = 186.97, P < 0.0001). TukeyÕs multiple comparison test showed that the Johannesburg and Soweto children differed significantly from the Lenasia and Gelukspan children for percentage of median height, while for percentage of median weight all study groups differed significantly (data available from authors).

The percentage of children with caries and mean and standard deviation of dmfs scores were cross-tabulated by wasting and stunting (Table II) to look for trends. Only groups containing at least 10 children were included. No consistent trends were seen within the study groups, but when the results for all the groups in the study were combined a clear but statistically not significant trend of increasing caries prevalence with increasing wasting was seen. The same trend was noted for dmfs scores, but this was statistically significant (c2 = 16.24, P = 0.003). For stunting no clear trends were seen. When dmfs scores were analysed, the Kruskal-Wallis test showed a weakly significant effect for stunting (c2 = 8.80, P = 0.03) but a more strongly significant effect for wasting (c2 = 16.24, P = 0.003).

Discussion
In nutrition studies choice of malnutrition definition is important for comparison of the findings of different researchers. There is no one accepted definition. Waterlow14 clearly motivated for the general requirements of a classification system. We used WaterlowÕs classification because it is practical, sound, and has been used by other workers such as Alvarez and colleagues.1-5

Comparison of wasting and stunting in our study groups with groups elsewhere was not the main purpose of the current article. However, to give some perspective a recent African study by Aldana and Piechulek,20 which examined children aged 0 - 59 months in Cameroon, found that prevalences of wasting were 6% for rural areas and 4% for urban areas. Stunting was found to be 22% in rural areas and 15% in urban areas, all of which figures are considerably lower than those in our South African samples.

Among South American children Alvarez et al.3 have shown clearly that when dmft teeth are plotted against age, a bell-shaped curve results that shifts to the right by 2.5 years in malnourished groups compared with normal children, a statistically significant shift. This shift was associated with a delay in both the eruption and exfoliation of the primary teeth in malnourished children. Peak caries activity was significantly higher in wasted and stunted-wasted children when compared with normal controls. The authors concluded that malnutrition delayed tooth development, affected the age distribution of dental caries and resulted in increased caries experience in the primary teeth. Wasting, and stunting as well as wasting, were associated with a greater delay than stunting alone, suggesting that acute malnutrition (i.e. low weight for height) has a more pronounced effect on tooth exfoliation than chronic malnutrition (i.e. low height for age). Alvarez and colleagues1-4 concluded that significant changes in the age distribution of dental caries in the primary teeth have important implications for epidemiological studies of dental caries. Firstly, comparisons of age-adjusted dental caries data between different countries or between different regions within a country cannot be made without the nutritional factor (i.e. skeletal growth) being taken into account. Secondly, infected carious molars stayed 2 or 3 additional years in the oral cavities of children with malnutrition at an age when most of the permanent teeth emerge (8 - 11 years), particularly the first permanent molar. This may increase the level of cariogenic bacteria in the mouth, and thereby increase the risk of caries development in the permanent dentition. A question that they consider remains unanswered is whether malnutrition in children leads to an increased caries experience.

Extensive animal experimentation has shown that protein-calorie malnutrition increases the caries rate in rats fed a moderately cariogenic diet.21 It is likely that the effect of nutritional status on the timing of tooth eruption and caries development described above has been an important confounder that has prevented epidemiologists and clinicians from observing a clear-cut effect of malnutrition on caries experience in children.

Not all delay in permanent tooth eruption is nutritional; there is also a genetic influence. For example, in studies in our communities we have shown that permanent teeth erupt 3.,5 - 7 months later among South Africa Indian children compared with their urban black counterparts.22 We do not know if this delay in eruption has influenced the caries rates among South African children.

In India, Johansson et al.6 studied salivary flow and dental caries in 8 - 12-year-old children suffering from chronic malnutrition as defined by weight-for-age ratios.23 They found that 7% were severely, 25% moderately, and 38% mildly stunted, whereas 30% were of normal height for age according to Indian anthropometric standards.24 Thirty-four children with severe-to-moderate PEM were selected for the study, and were compared with normal weight-for-age children from the same sample, as well as with well-nourished Swedish children. Stimulated saliva secretion rate decreased as malnutrition increased. The Swedish children had a significantly higher secretion rate of saliva than the age-matched malnourished Indian groups. Chronic malnutrition did not affect unstimulated secretion rate; however, the well-nourished Swedish children had a much lower buffer capacity than all four of the Indian PEM groups. The average amount of sucrose consumed per day was low. The Johannson study6 showed a significant correlation between degree of chronic malnutrition and impairment of chewing-stimulated saliva secretion; furthermore secretion rate and buffer capacity were inversely correlated with malnutrition and dental caries. Diet may have influenced the secretion rates and possibly the immune system.

In another study done on our sample, black children were found to secrete saliva at approximately twice the rate of the white and Indian children.10 Re-examination of those data according to wasting and stunting criteria showed a slight reduction in secretion rate of saliva which was weakly significant due to wasting (N = 2 691, F = 3.29, P = 0.01) and stunting (N = 2 683, F = 3.04, P = 0.03). No relationship between salivary secretion rate and dental caries was seen in our sample.

Walker et al.9 concluded that a prejudicial effect of malnutrition on caries seen in 11-12-year-olds appears to concern principally the primary teeth. In their study, which was undertaken on a low-caries group of black schoolchildren, no disadvantage to the permanent teeth was apparent in the slower-growing moiety compared with the normal children.

The message of the current study is clear -- a significant association was found between wasting and dmfs scores but not between stunting and dmfs scores. The strengths of the current study were the large sample size and the inclusion of groups known to vary in caries rate and degree of malnutrition. This conclusion differs from the work of Alvarez and colleagues1-3,5 and that of Johannson et al.6 in other countries. It is possible that confounding variables not considered in any of the studies are responsible for the differences between them.

This study report is the final one in a series dating from 1991 using data collected in 1984. The massive data set, which took almost 6 years to code, to store in a computer and to ÔcleanÕ, is on file for comparison with data collected some 10 years later, in the same urban areas, as part of the Birth-to-Ten study.

We are most grateful to the children, parents, guardians and teachers who took part in this study.

References

1. Alvarez JO, Navia JM. Nutritional status, tooth eruption, and dental caries: a review. Am J Clin Nutr 1989; 49: 417-426.
2. Alvarez JO, Lewis CA, Saman C, et al. Chronic malnutrition, dental caries and tooth exfoliation in Peruvian children aged 3 - 9 years. Am J Clin Nutr 1988; 48: 368-372.
3. Alvarez JO, Eguren JC, Caceda J, Navia JM. The effect of nutritional status on the age distribution of dental caries in the primary dentition. J Dent Res 1990; 69: 1564-1566.
4. Alvarez JO, Caceda J, Woolley TW, et al. A longitudinal study of dental caries in the primary teeth of children who suffered from infant malnutrition. J Dent Res 1993; 72: 1573-1576.
5. Caceda J, Alvarez JO, Aguayo H. Early childhood caries and nutritional status: A longitudinal study in Peru. J Dent Res 1995; 74: (Special issue): 79, IADR abstract.
6. Johansson I, Saellstršm A-K, Rajan BP, Parameswaran A. Salivary flow and dental caries in Indian children suffering from chronic malnutrition. Caries Res 1992; 26: 38-43.
7. Johansson I, Alm P. Effect of moderate protein-deficiency on ultrastructure in parotid and submandibular acinar cells in the adult rat. Scand J Dent 1989; 64: 37-43.
8. Sawyer DR, Nwoku AL. Malnutrition and the oral health of children in Ogbomosho, Nigeria. J Dent Child 1985; 52: 141-145.
9. Walker ARP, Walker BF, Dison E, Walker C. Dental caries and malnutrition in rural South African Black ten- to twelve-year-olds. J Dent Ass S Afr 1988; 43: 581-583.
10. Cleaton-Jones P, Granath L, Richardson BD, eds. Dental caries, nutrient intake, dietary habits, anthropometric status, oral hygiene, and salivary factors and microbiota in South African black, Indian and white 4-5-year-old children. Descriptive data from a survey in 1984. South African Medical Research Council Technical Report. Parowvallei: Medical Research Council, 1991.
11. Dreyer AG, Grobler SR. Die fluoriedgehalte in die drinkwater van Suid-Afrika en Suidwes-Afrika. J Dent Ass S Afr 1984; 39: 793-797.
12. Cleaton-Jones P, Hargreaves JA, Fatti LP, Chandler HD, Grossman ES. Dental caries diagnosis calibration for clinical field studies. Caries Res 1989; 23: 195-199.
13. World Health Organisation. Oral health surveys. Basic methods. 2nd ed. Geneva: WHO, 1977.
14. Waterlow JC. Classification and definition of protein-calorie malnutrition. BMJ 1972; 3: 566-569.
15. Waterlow JC. Classification and definition of protein-energy malnutrition. In: Nutrition in Preventive Medicine. WHO Monogr. Ser. No. 62. Geneva: WHO, 1976: 530-555.
16. Neumann CG. Reference data. In: Jelliffe DB, Jelliffe EFP, eds. Human Nutrition 2. Nutrition and Growth. (Alfin-Slater RB, Kritchevsky D, general eds.) New York: Plenum Press, 1979: 299-327.
17. Hamill PVV, Drizd TA, Johnson CL, Reed RB, Roche AF, Moore WM. Physical growth: National Centre for Health Statistics percentiles. Am J Clin Nutr 1979; 32: 607-629.
18. Statistical Analysis System Institute Inc. SAS/STAT UserÕs Guide, version 6, 5th ed. Cary, NC: SAS Inc, 1989.
19. Ingelfinger JA, Mosteller F, Thibodeau LA, Ware JH. Biostatistics In Clinical Medicine. New York: MacMillan, 1983: 169-170.
20. Aldana JM, Piechulek H. Situation nutritionelle des enfants de 0 ˆ 59 mois en zone urbane et rurale du Cameroun. Bull World Health Organ 1992; 70: 725-732.
21. Shaw JH, Griffiths D. Dental abnormalities in rats attributable to protein deficiency during reproduction. J Nutr 1963; 80: 123-141.
22. Blankenstein R, Cleaton-Jones PE, Maistry PK, Luk KM, Fatti LP. The onset of eruption of permanent teeth amongst South African Indian children. Ann Hum Biol 1990; 17: 515-521.
23. Gomez F, Ramos R, Frenk S, Cravito J, Chavez R, Vazquez J. Mortality in second- and third-degree malnutrition. J Trop Pediatr 1956; 2: 77.
24. Wahi PN. Growth and Physical Development of Indian Infants and Children. Tech Rep Ser. No. 18. New Delhi Indian Council of Medical Research, 1984.

Reprinted from the South African Medical Journal (2000; 90: 631-635).

Last updated: 17-Feb-2004