CHILDHOOD BLINDNESS 331 namely that boys get the "good food" such as meat and lentils, which are rich in proteins, while girls get the "cheap food," which consists mostly of vegetables. Ironically, the "cheap foods" contain more vitamin A, thereby protecting girls from xerophthalmia. The current Nepal Blindness Survey data do not shed any light on this hypothesis; however, the fact that the male:female sex ratio remains virtually the same at all ages well into adulthood suggests that whatever factors are operating are not only factors of childhood. If the expensive food argument is correct, one would expect to see similar nutritional patterns in adulthood-if, in fact, adult Bitot's spots are linked to food patterns at all. The urban/rural location of cases of active xerophthalmia and Bitot's spots (Table 8-11) parallels the predominantly rural distribution of the population as a whole. For example, of the 15,534 estimated cases of Bitot's spots in the country, 14,636 (94.2%) are located in rural areas. The prevalence rate for rural areas and urban areas is virtually identical (0.6%). The distribution of active xerophthalmia by urban/rural location follows the same pattern. TABLE 8-11. PREVALENCE, ESTIMATED NUMBER OF CASES, AND DIS- TRIBUTION BY RURAL/URBAN AREA, ACTIVE XEROPHTHALMIA (AND BITOT'S SPOTS), CHILDREN UNDER AGE 6, DECEMBER-APRIL (NBS 1981) Estimated Number of Children Affected Prevalence per 100 Children Percentage of Nepal's Total Sample N Area Rural Urban Subtotal Missing data 6,837 367 7,204 367 19,518 (14,636) 1,340 (898) 20,858 (15,534) 6 (0) 0.9 (0.6) 0.8 (0.6) 0.8 (0.6) NA(NA) 93.6 (94.2) 6.4 (5.8) 100.0 (100.0) NA (NA) Total 7,580 20,858 (15,534) 0.8 (0.6) 100.0 (100.0) Source: FR8.002 AN.81, 91 M Note: Figures in parentheses refer to the estimated number of cases, prevalence, and distribution of cases of Bitot's spots. 8.5.3 Geographic Characteristics In this section the distribution of xerophthalmia is discussed with reference to certain geographic characteristics that are usually as- sociated with surveys. It may be helpful to organize these geographic variable into three subsets:
