In the previous poaper, the author reported the relationships of hemoglobin (Hb) values with growth and onset of menarche, using 6-17 year-old school children (2, 229 boys and 2, 158 girls) in Amakusa, Kumamoto Prefecture. The present report further examined the different growth pattern by sex with special reference to growth rate as a trigger of anemia in adolescent girls. Results are are as folows. 1) Growth spurt of boys in stature, weight, chest girth and sitting height was observed during 12-14 years of age, while girls' spurt was much earlier, during 10-12 years of age, when girls' values were higher than boys. 2) Sex-age means of anthropometric values and Hb level correlated highly, but between individual values, few significant correlation coefficients were obtained. 3) In boys, age-means of anthropometric values increased in accordance with that of Hb values. But in girls, Hb value showed its peak when growth rate was maximum. Thereafter, girls' Hb values decreased slightly irrespective of ageing, although physical growth rate was low. For this reason, girls' anthropometric values correlated with Hb values positively in 6-12 year-old group, and adversely in 12-17 year-old group. 4) Girls' growth spurt preceded their "Hb spurt" about 0.5-2 years, which might be interpreted as if growth triggered increment of Hb level in girls. 5) The age-distribution of girls anemia (Hb 12.0 g/dl) was V-shaped, and the proportion of anemia was the lowest when their growth rate was the highest. Girls' anemia was higher in groups with low annual growth rate of height and weight than those with opposite growth pattern.
In order to clarify the involvements of acceptance of favorites such as tabacco, alcohol, coffee and tea in women's complaining, the relationship between their complaint and factor structure of preference, which was calculated by using factor analysis from acceptances of favorites and preferences for food groups classified 13 according to food composition table, was examined in 236 women aged 20-39 . The explanation of its relationship was performed by canonical discriminant analysis with 7 explanatory variables; its 4 factor scores, age, systolic, diastolic blood pressure and weight index examined simultaneously. The results were as follows: 1) The 1st, 2nd, 3rd, 4th factor of preference by factor analysis was deduced to be that for volumed foods, side foods, modern foods and favorites, respectively. 2) The women with either of headache, constipation or coldness of extrimities, which were found to be specified complaints to women in another examination of comparing men's complaint was discriminated from women free from complaints by the 1st canonical discriminant variate, and the women with the other complaints was by the 2nd canonical discriminant variate. The size of explanatory variables for 1st canonical variate was the order of the 2nd, 1st and 4th factor of preference, and 2nd discriminant variate was large above all for age variable. These results suggest that women's complaining of headahe, constipation and coldness of extrimities has deep relationship to no preference for side foods as well as preferences for favorites, while the other complaints being related especially to aging.
Spot urine of 7, 507 females aged 40-69 years were collected by Filter Paper Method from 169 regions (88 urban and 81 rural areas) in Japan during March to December of 1985. Spot urine were analyzed for sodium (Na) and potassium (K) by flame photometry and for creatinine (Cr) by Jaffe's method. The ratios of Na/K (mEq/mEq), NaCl/Cr (gig), K/Cr (g/g) in spot urine and estimated salt and potassium excretion (g) were calculated. The results were as follows: 1. Mean Na/K, NaCl/Cr and estimated salt excretion were significantly higher in rural area than urban area (4.13 vs 3.75, 13.45 vs 12.65, 9.56 g vs 9.41 g respectively). K/Cr and estimated potassium excretion were not different in both areas. 2. Mean systolic and diastolic blood pressure were significantly higher in rural area than urban area (130.4 mmHg vs 128.6 mmHg and 77.9 mmHg vs 77.3 mmHg). 3. Partial correlations were calculated taking the factors of age, height and weight into account. Systolic and diastolic blood pressure were correlated with Na/K, NaCl/Cr or estimated salt excretion positively and with K/Cr or estimated potassium excretion negatively. 4. Regional estimated salt excretion was higher in Tohoku and Kanto-II regions and was lower in Kinki-I, II, Chugoku, Shikoku and North Kyushu regions than an average of Japan, and was correlated positively with salt intake by National Nutrition Survey. 5. Significant correlations were found between the 1978-1982 standardized mortality ratio for cerebrovascular disease and Na/K, estimated salt or potassium excretion (r=0.325; p<0.05, r=0.690; p<0.001, r<0.473; p<0.001, respectively).
The purpose of this study was to examine the regional differences in infant mortality rate in relation to socioeconomic conditions. Using the mortality statistics issued by the Kagoshima prefectural government, secular changes in cause-specific infant mortality and the relationships between infant mortality rate and indicators of selected socioeconomic conditions were examined. The following findings were made. 1) Infant mortality rate declined drastically over the period 1968-1985. The decline in the infant mortality rate was mainly due to the decline in infectious disease and premature mortalities. The changes in mortalities due to congenital anomaly and hypoxia made little contribution to the decline in the infant mortality rate. 2) Among the 96 cities, towns and villages, the infant mortality rate negatively correlated with indicators of increase in population, gross income, medical doctors and crude birth rate, while it positively correlated with aging-index and gross fertility rate. 3) When total infant mortality was subdivided into two categories, the infant mortality due to congenital anomaly and hypoxia and that due to the others, the former infant mortality rate negatively correlated with indicator of increase in population and crude birth rate and positively correlated with aging-index, whereas the latter one negatively correlated with indicators of increase in population and medical doctors and positively correlated with gross fertility rate. 4) The findings above-mentioned support the hypothesis that infant mortality rate may be higher among the communities of decreasing population, independently of socioeconomic conditions.