This categorization of individuals into an "affected" and an "unaffected" group has been used to investigate the distribution of the morbid complications which often accompany diabetes. As expected, compared with unaf- fected individuals, affected ones showed greater frequencies of polyuria (75.0 vs 37.8%) and polydipsia (72.3 vs 45.1%). Recent weight loss was reported by 48.8% of the affected group and 22.1% of the unaffected group. Itching was reported by 60.5% of the affected group and 30.5% of the unaffected. In females, pruritis vulvae was reported by 20.8% of the affected persons and by none of the unaffected ones. Previous studies on other populations in the Pacific suggested that the more important complications of diabetes were very similar to those observed in Caucasian populations (74,77). The analysis of the affected population in the Marshall Islands does not completely support such a contention. Fundoscopic examination revealed diabetic retinopathy in only one of the 54 affected individuals examined. An additional four had exudates in the fundi, but so did three of the 263 unaffected individuals examined. Exudates did not appear to be related to the degree of hypertension in this population. Cataracts were found in 24.1% of the affected individuals and 13.6% of the unaf- fected, but this difference might reflect age since the affected group was older. Abnormalities of the extremities were very unusual in this population, and no difference was found between affected and unaffected groups except for amputations of the leg in three previously diagnosed individuals. tations were reported, but were not seen during the 1974 survey.) (Other ampu- Peripheral pulses were absent in only 5% of the affected group (2% of the unaffected). Absence of dorsalis pedis pulse occurred in only one affected and five unaf- fected persons (not significantly different in frequency). Absence of poste- rlor tibial pulses was somewhat more common, occurring in 12.7% of affected persons but only 3.1% of unaffected ones. The presence or absence of reflexes was also used to investigate the prevalence of peripheral neuropathies. jerks were not elicited in 68.3% of the affected population tested, with 20.9% of the unaffected. Ankle compared Knee jerks were absent in 50.7% of the affected individuals but in only 17.5% of the unaffected ones. Brachioradalis reflex was absent in 28.6% of the affected group and 8.2% of the unaffected group. Kidney function could be evaluated crudely by measuring proteinuria. Some indication of proteinuria was seen in 61.7% of affected individuals and 15.94% of unaffected ones, as tabulated below. Group Proteinuria: Affected Unaffected l+ 2+ 3+ 4+ Total 17.0 12.8 6.4 25.5 61.7% 3.1 1.3 - 11.5 15.9% However, proteinuria did not appear to be correlated with the level of glucose intolerance as measured by fasting or post~prandial blood glucose values when corrected for age, sex, and level of obesity. Glycosuria was found in 80.6% of the affected population (at 4+ levels in 20.9%), compared with 14.4% of the unaffected population. Glycosuria was significantly correlated with fasting blood sugar and post-prandial blood sugar even with corrections for sex, age, and obesity. Symptoms of sexual dysfunction were reported in “13% of the entire population, and in 21.3% of affected individuals. Paresthesias, at least -~ 45 -