The monthly distribution of births before 1900 in the normal Japanese population was investigated by collecting the birth dates of 43, 108 individuals, mostly adults, from various sources. The birth dates of 6, 919 adults born between 1901 and 1905 were also shown to make possible a comparison with the data in the vital statistics which have been published annually since 1899 by the Japanese Government. Major sources of these data were: 1) 34, 115 adult survivors registered in a) registration books of residents in ward offices in Tokyo and Osaka, b) lists of inmates at nursing homes for the aged in Tokyo, Osaka and Kyushu, and c) lists of members of local societies, the Diet and an alumni-association; 2) 6, 988 death certificates of adults in a) the general population in Tokyo between 1965 and 1975 and b) a nationwide employees' association, and 3) 2, 005 birth records in a) old parishoners' census registration books, known as "Ninbetsu-Cho, " for the years between 1755 and 1867 and b) old diaries, documents and descriptions. In the latter half of the 19th century, the seasonal distribution of birth had an early spring peak between January and March and a trough between May and July, similar to the typical distribution of births in the 20th century. From the late 18th to early 19th centuries, a spring peak was not so evident. Instead, a fall peak was rather prominent inmost samples. Similarly, from the 9th to the early 18th centuries, a fall peak pattern was also seen in small samples. The seasonal distribution of births in the latter half of the 19th century in Tokyo, Osaka and Kyushu can be used as distributions of normal control populations for studies on the relationship between birth season and morbidity of some particular diseases.
In Japan, the decline in the age at menarche after the Second World War has been repeatedly reported, but the observed period in the reports has not been long enough to evaluate the secular trend of it. More than a hundred reports of age at menarche of Japanese have been published from the year of 1886. More than half of these populations in the reports consisted of students or young workers, some of whom had not attained menarche at the survey, and the menarcheal ages were represented by the arithmetric mean for the menarche attained girls. Thus, ages at menarche of these reports have biases toward younger menarcheal ages which depend on the proportion of non-menstruating girls. The authors aimed to correct these biased menarcheal ages on the assumption that (1) menarcheal ages of a population distributes normally when all of the girls are menstruating; (2) when some girls are not menstruating, the distribution is censored sample of normal distribution. After eliminating these biases of historical sources, the trend of the menarcheal age in Japan (from the late 19th century to the present) was analysed.
From a seroepidemiological point of view a survey was conducted on the effect of influenza vaccine over a representative epidemic period of influenza of type A and B extending from August, 1968, when Hongkong type A appeared, and March, 1979. The results obtained are summarized as follows.1. The rate of infection in each epidemic was 30.6% for the A/Kumamoto /1/72 strain, 46.1% for the B/Kanagawa/3/76 strain, and 46.4% for the A/Tokyo/1/77 strain. It was only 8.7% in the second epidemic of the A/USSR/92/77 strain.2. The rate of infection was analyzed by HI antibody titer before the epidemic. As a result, it was 8% or less, showing no marked fluctuation, in every epidemic, so long as the patients possessed antibody titers of 1:641:128 or more against the epidemic strain.3. When the patients were examined for the production of antibody higher than 1:128 in titer against the initial vaccine strain, only about 20% of them exhibited such production as this against Hongkong type A and USSR type A.4. On the other hand, only about 710% of the patients exhibited such production as this against type B.5. Patients of the population exposed to the first epidemic of the A/USSR/92/77 strain were inoculated with vaccine containing virus of this strain. Then they were involved in the second epidemic of this strain. As a result, they presented such a low rate of infection as 8.7%.6. Examination was carried out on medical workers in offices and plants where epidemics of Hongkong type A and USSR type A had confirmed. None of them were found infected when they were examined for HI antibody against Hongkong type A after inoculation with vaccine and for retained antibody against USSR type A. In conclusion, it can be expected that influenza vaccine will have an effect when the strain used for the vaccine is identical with that which has caused the respective epidemic. The existing method of inoculation with vaccine cannot be anticipated to be effective when an antigenic shift strain or antigenic drift strain showing a large variation appear in future. It can be well expected, however, from the present studies that influenza vaccine will display an effect in those cases by increasing the number of inoculation or by raising the amount of virus contained in the vaccine.
According to high mortality and specific risk factors following main diseases have been chosen for principal-factor solution in Japan, and England and Wales; cerebrovascular disease, ischemic hear disease, stomach cancer, cancers of trachea, bronchus and lung, accidents, poisonings and violence, and suicide. The factor analysis has been carried out using SMRs by occupation. The results are as follows; 1) In Japan first factors weighed are 0.97 for stomach cancer, 0.95 for cerebrovascular disease, 0.82 for accidents, poisonings and violence, and suicide, 0.75 for ischemic heart disease, and 0.57 for cancers of trachea, bronchus and lung. Among second factors weighed in Japan, -0.70 for suicide, and 0.69 for cancers of trachea, bronchus and lung are dominant. 67.9% of communality is attributed to the first factors, and 18.0% to the second factors. 2) In England and Wales, first factors weighed are 0.96 for stomach cancer, 0.94 for cerebrovascular disease, 0.88 for cancers of trachea, bronchus and lung, 0.87 for accidents, poisonings and violence, 0.76 for ischemic heart disease, and 0.19 for suicide. Among second factors weighed in England and Wales, 0.93 for suicide is dominant. 70.1% of communality is attributed to the first factors, and 18.4% to the second factors. 3) The disease having the bigger correlation coefficient between SMRs and stillbirth fates by occupation shows the bigger first factor weighed in both countries.