Underwater weighing is based on the assumption that fat-free body density is roughly constant among humans. This assumption should be examined, because fat-free body density may in fact depend on the bone mineral and water contents of the body, with fat excluded. The purpose of this study was to investigate the effects of bone mineral content (BMC) and density (BMD) on the accuracy of body fat measured underwater. The subjects were 12 young men (25.1±3.7 years, mean ± SD), some of whom were trained athletes. BMC and BMD were measured by dual-energy x-ray absorptiometry (DXA), as was body fat, as a percentage of body weight; this method is not based on the assumption that fat-free body density is the same in different individual. Body fat as a percen tage of body weight was measured underwater, also. Body fat measured by DXA was significantly correlated with that found by underwater weighing (r = 0.83, p<0.01), as expected, but the mean body fat found by DXA was 4.3% higher. The differences between results by the two methods for individuals were from -11.5% to 2.7%, and the differences were negatively correlated with BMC/fat-free weight (FFW ; r=-0.82, p < 0.01) and BMD (r=-0.85, p<0.01) . Fat-free body density ranged from 1.097 to 1.111 g/cm³because BMC/FFWs varied with the individual. We concluded that individual differences in BMC/FFW and BMD affected the fat-free body density. The variations in fat-free body density would give rise to systematic errors in body composition measured underwater.

Introduction
 During the course of primary health care activities in Khammouane province of Lao PDR by non- profit organization, international support and partnership for health, an exceedingly high infant mortality rate was recognized in some villages. As thiamine deficiency was strongly suspected for the main cause of the high mortality, the household survey with an emphasis on nutritional aspect and the measurement of thiamine level in the blood of mothers as well as their breast-milk were carried out for the confirmation.
Methods
 The survey was conducted at Sibounhouane sub-district (Group 1) with the highest infant mortality and Hatkhamhieng sub-district (Group 2) with the lowest mortality. Fifty families of the two groups, each consisting of randomly selected 25 families with less than one year old infant(s) were interviewed by using prepared questionnaire. The body weight of both mother and infant was measured and blood and milk were taken from the mother.
Results
 The ethnic minorities of the group 1 consisted of Lao Lum (44%) and Lao Theung (56%), whereas only Lao Lum in the group 2. The economical indicators and the mother's educational level in the group 1 were inferior to those in the group 2. There was no significant difference in the methods of preparing staple glutinous rice, likely to flux thiamine and intake of thiaminases between the two groups. High infant mortality rate was documented in the group 1 whose clinical symptoms were mostly consistent with those of infantile beriberi. The concentration of thiamine in the blood and milk of mothers in both groups were considerably lower and significant reduction of thiamine level was observed in the group 1 than the group 2. Nevertheless, both groups of those infants tended to be inadequately fed with the foods except for breast-milk during their early infantile phase.
Conclusions
 The low concentration of thiamine in the blood and milk of mothers was strongly suggestive that infantile beriberi due to thiamine deficiency was the principal cause of the infant mortality in the study region. We hereby surmise that thiamine deficiency could be extensively prevalent in Lao PDR than the currently studied areas because of their traditional preparing methods of staple dietary rice, postpartum food taboos and intake of thiaminases. For the prevention of infant death due to severe thiamine deficiency, necessary measures including the exchange of information among the institutions concerned, adequate thiamine supplementation and nutritional counseling have to be urgently adopted.

Enrolled in our study were 16 seriously handicapped persons (9 males and 7 females) institutionalized in our medical center. They were 20 to 58 years of age (mean 35 yrs). In the previous study, we found they had no evidence, serological or clinical, of osteomalacia but were verified as having significantly low bone mineral density (BMD) at the calcaneus and lumbar vertebrae. Serum levels of 25-OH-vitamin D were low and positively correlated with the lumbar vertebral BMD. Thus we postulated that low BMD was at least partly due to the lack of ultraviolet exposure. In the present study, the lumbar vertebral BMD of the above subjects were measured before and after increasing their exposure to sunlight. We found that mean lumbar BMD increased signigicantly from 770 mg/cm² to 786 mg/cm² (P<0.05, Wilcoxon, paired test). The Z-score also increased significantlyfrom 78 ± 13% to 80 ± 12%(P<0.05, Wilcoxon, paired test).
These findings suggest low BMD could be corrected to some extent by increasing sunlight exposure of institutionalized persons. Former members of farming communities who are retired, handicapped, or otherwise limited to indoor activity, can also recover their BMD and prevent osteoporosis by an increased daily exposure to sunlight.