[Digital morphometric and stereologic analysis of small intestinal mucosa in well-nourished and malnourished children with persistent diarrhea].

Objective: To test the hypothesis that the proximal small intestines of children with persistent diarrhea present morphometric and stereologic changes proportional to their nutritional status, using microscope images stored in a computer.

Methods: Cross-sectional study with 65 pediatric patients, whose ages ranged from 4 months to 5 years, with persistent diarrhea for over 14 days. The nutritional assessment was performed according to the z-scores for weight/age (W/A), weight/height (W/H) and height/age (H/A) ratios, divided into: well-nourished = z > or =2SD and malnourished = z<2SD; well-nourished = z > or =2SD, nutritional risk = z<1SD and malnourished = z<-2SD; and continuously, in descending order, using the NCHS charts. After obtaining the computer images using the software Scion Image, villous height, crypt depth, mucosal thickness, total mucosal thickness, and villous/crypt ratio were measured in the fragments of the small intestinal mucosa, enlarged 100 times. When images were enlarged 500 times, enterocyte height, nuclear height and brush-border height were measured. Stereologic analysis was performed using cycloid arcs.

Results: For W/A, W/H and H/A z-scores, divided into two nutritional status categories, no statistically significant difference was observed in regard to villous height, crypt depth, mucosal thickness, total mucosal thickness and villous/crypt ratio. Enterocyte height presented the most significant difference between well-nourished and malnourished groups, for W/A and W/H ratios, with a 500x enlargement, although this difference was not statistically significant. When z-scores were subdivided into three nutritional status categories, a digital morphometric analysis showed a statistically significant difference for villous/crypt ratio between the well-nourished and slightly malnourished group and the well-nourished and mild to severe malnourished group (p=0.048). The villous/crypt ratio was higher among well-nourished children. Using the Spearman coefficient, the variables enterocyte height, height of enterocyte nucleus and brush-border height presented a clear association with the W/A ratio (r=0.25; p=0.038), W/H ratio (r=0.029; p=0.019). The height of the enterocyte and the brush-border height were associated with W/H ratio.

Conclusion: The observed associations between nutritional status and the analyzed small intestinal mucosa variables showed a positive correlation with patients' weight. Although these associations were of a slight to moderate magnitude, we observed a tendency of enterocyte size reduction, as well as a reduction in the size of its nucleus and brush-border, as the level of malnutrition increases.