Abortion Trends in Japan, 1975-95.
Aya Goto; Chihaya Fujiyama-Koriyama;Akira Fukao; Michael R. Reich.
In this study, recent trends in the incidence of induced abortion are analyzed in order to identify the target population and its requirements for family planning policy in Japan. Abortion statistics from 1975 to 1995from the Ministry of Health and Welfare are reviewed. The abortion rate (the number of cases of induced abortion per 1,000women per year) for women younger than 20 increased during the study period. The abortion ratio (number of cases per 1,000 live births) remained the highest among women aged 40-44. An increase in the abortion ratio was seen in the two youngest groups (younger than 20 and 20-24), especially among those who were born after 1955.The proportion of abortions experienced by women younger than 25 increased from 18 percent between 1976 and 1980 to 30 percent between 1991 and 1995, and a slight increase was also observed among women aged 40-44. The proportion of abortions performed after eight weeks of a pregnancy for the two youngest groups remained higher than that for older age groups during 1975-95.The analysis demonstrates that women younger than
25 should be the principal concern of family planning policy in Japan. Further investigations on unintended pregnancy are recommended. Japan was one of the first countries to legalize induced abortion through the Eugenic Protection Law of 1948. This law was revised as the Maternal Body Protection Law in 1996. Abortion is widely accepted in Japan: According to a survey conducted in 1998,79 percent of unmarried and 85 percent of married women approved of abortion (Population Problems Research Council of the Mainichi Newspapers 1998).
After World War 11,inadequate knowledge of contraception and widespread poverty throughout Japan contributed to an increase in the number of abortions, which reached a total of 1,170,143in 1955.The abortion rate in that year was 50.2 per 1,000 women aged 15-49 years (Ministry of Health and Welfare, Minister's Secretariat 1998). The Japanese government started a family planning program in 1952 targeting married couples. Subsequently, the number of abortions and the annual abortion rate decreased dramatically for about a decade after 1955, as Japan entered a period of rapid economic growth. Since the mid-l960s, the number of abortions, the abortion rate, and the abortion ratio have all continued to decrease, but at a slower rate than in the previous decade.
Current abortion statistics reflect a more complex social picture. Although the number of abortions per year has continued to decline, the proportion of unplanned births is reported to be as high as 52 percent in Japan (1992), compared with 19 percent in France (1994) and 30 percent in the United States (1995) (Alan Guttmacher Institute 1999).Until recently, contraceptive choice remained extremely limited in Japan. Only in 1999 did the government approve low-dose oral contraceptives (Goto et al. 1999),the copper-bearing intrauterine device (IUD), and the female condom. Whether government sanction of these additional methods will reduce the rate of unintended pregnancies remains to be seen, and will depend, in part, on whether health policy and planning are based on recognition and analysis of women's needs. This study analyzes trends revealed by abortion statistics in order to identify target populations for family planning policy in Japan.
Methods
Data concerning the incidence of induced abortion were obtained by five-year age group of the mother and by gestational age of the fetus from the Maternal Body Protection Statistics reports (Botai Hogo Tokei Hokoku), annual reports issued by the
Ministry of Health and Welfare (MHW) (Ministry of Health and Welfare, Minister's Secretariat 1998).The authors examined the age-specific trends in abortion statistics from 1975 to 1995. Physicians are required by law to report the following information about the abortions they perform: age and address of the patient, gestational age of the fetus, reason for and date of the procedure, name and address of the facility where the procedure was performed, and name of the attending physician. Additional characteristics of women obtaining abortions and of procedures performed are not reported.
Two measures were used in the present analysis. The abortion rate (the number of abortions per 1,000women per year) shows the incidence of abortion for all women, and the abortion ratio (the number of abortions per 1,000live births per year) reflects
the likelihood that a woman will have an abortion when she becomes pregnant. The number of live births is reported annually, but population data are reported only every five years based on the census (Health and Welfare Statistics Association 1998).Thus, although the abortion ratio was calculated for each year, the abortion rate was calculated only for census years (1975, 1980, 1985, 1990, and 1995). Both the abortion rate and the abortion ratio were presented by women's age groups: [less than]20,20-24,25-39, and 40-44. A single category was created for women aged 25-39, because women of these ages had similar time trends. Both the abortion rate and the abortion ratio for women younger than 20 were calculated using the number of women and live births in the 15-19-yearage group as denominators. For the numerator of the abortion rate, average numbers of women obtaining abortions during the five years around those census years were used. The women whose ages and for whom gestational ages were unknown were excluded from the calculation. In addition, the abortion ratio was also analyzed for birth-cohort effects, by dividing abortion cases into seven five-year birth cohorts from 1965 to 1995 (MacMahon and Pugh 1970).
Time trends in the age distribution of women having abortions were examined by dividing the study period into four five-year periods (1976-80, 1981-85, 1986-90,and 1991-95). For each period, the proportion of abortion cases in each age group among
total cases was calculated. To assess differences in the timing of abortion by age, women in each age group were categorized by the gestational age of the fetus: less than eight weeks and eight weeks or more. Time trends were examined by dividing the 20 years into the same four five-year periods as for age distribution. For each period, the proportion of women in each age group who underwent the procedure after eight weeks of pregnancy was calculated.
The official Japanese government statistics on abortion that are
used for the calculations above are known to be incomplete. Physicians tend to underreport the number of abortions they perform as a way of avoiding income tax payments (Coleman 1991) and because of social pressures to protect women's
confidentiality, especially that of young women who are junior high or high school students. Because the administrative reporting system did not change during 1975-95, however, the trends over time in abortions reported during this period may be
considered reasonably accurate (Hayashi et al. 1992; Hodge and Ogawa 1991). The issue of underreporting is discussed further below.
Results
In 1995,the total number of abortions reported was 343,024, representing a 49 percent decrease from the number reported for 1975. The overall abortion rate changed from 22 to 11 abortions per 1,000women in 1975 and 1995,respectively; and the overall abortion ratio changed from 353 to 289 abortions per 1,000live births in the same 20-year period. In more than 99 percent of cases, the reason reported for performing an abortion was to protect the woman's health; this percentage remained constant during 1975-95. Of the 11,060,037reported cases during 1975-95,the woman's age was not known for 6,271 cases, and gestational week was not known for 8,363 cases. The numbers of cases for which this information was missing decreased from 775 (age) and 627 (gestational week) in 1975 to 17 and 38, respectively, in 1995.
Figure 1 shows the abortion rate for all women between 1975-95.The only age group for which the abortion rate increased was that of womenyounger than 20, increasing by 109 percent from 1975 to 1995.The decline in the abortion rate for women aged 20-24 was less (33 percent) than that of women aged 25-39 and 40-44 (50 percent and 44 percent decrease, respectively) during the study period.
Figure 2 shows that the likelihood that a pregnant woman will obtain an abortion (the abortion ratio) in-creased in two age groups: the under-20 age group (114 percent increase) and the 20-24 age group (77 percent increase). This measure decreased in the 25-39 age group (40 percent decrease) and the 40-44 age group (59 percent decrease), but the 40-44 age group maintained the highest ratio. Next to the oldest age group, the teenage group showed the highest ratio, indicating that pregnancies among women at either end of the reproductive age span are most likely to be terminated.
Figure 3 shows the birth-cohort analysis of age-specific abortion ratios. All age groups in each one of seven birth cohorts (1950, 1955, 1960, 1965, 1970, 1975, and 1980) were plotted vertically, and the same age groups in those cohorts were connected by lines. An increase in the abortion ratio was found among women younger than 25 who were born after 1955.By contrast, other age groups did not show an increase, even for those who were born after 1955.
Figure 4shows that the proportion of all abortion cases represented by women younger than 25 increased from 18 percent between 1976 and 1980 to 30 percent between 1991 and 1995. Among women in older age groups, the proportion decreased from 73 percent to 59 percent for 25-39-year-old women, but it increased from 9 percent to 11 percent for 40-44-year-olds.
Figure 5 shows that the proportions of women having abortions after eight weeks of pregnancy in the two youngest age groups ([less than120 and 20-24) remained higher (60 percent and 51 percent, respectively, in 1991-95)than did those in older age
groups during the last 20 years.
Discussion
Data for 1975 to1995 from the annual reporting of abortion in Japan from the Ministry of Health and Welfare were analyzed. The authors' primary conclusion is that young women (who are mostly unmarried) should be the focus of Japan's family planning policies. Nearly one-third of all abortions are performed on women younger than 25. The abortion ratio has increased for women younger than 25, suggesting that such women increasingly are choosing to have abortions when they become pregnant. The abortion rate increased among women younger than 20. The birth rate for teenage women was fairly constant at four children per 1,000 women throughout 1975-95, indicating that unwanted pregnancies increased for these women, leading to their greater reliance on abortion (Health and Welfare Statistics Association 1998). The results also showed that younger women seek abortion services later during pregnancy than do older women.
Although the analysis indicates specifically that young people's contraceptive needs are increasing in Japan, the contraceptive requirements of older people should be addressed also. The incidence of abortion decreases among women entering their 40s,
but the present analysis shows that women in that age group are most likely to choose abortion once they become pregnant (with a ratio of about three abortions to one live birth in 1995). Furthermore, the proportion of abortions for women aged 40-44 increased during the study period. As shown in a comparison of abortion statistics for Japan, the United States, and England and Wales (Goto et al. 1999), Japan was found to have a distinct pattern in that women in their late 20s and 30s continue to have abortion rates as high as those of younger age groups. If recent abortion trends continue in Japan (that is, if the abortion rate among teenagers continues to increase, if the reduction in the abortion rate for women in their early 20s remains low or stops, and if the rate for women in their late 20s and 30s remains as high as for those in their early OS), then Japan's overall abortion rate could increase in the next decade. Trends observed in the late 1990s support this projection: The reduction in the overall abortion rate stopped recently, in the four years from 1995 to 1998, and the rate remained at 11 abortions per 1,000 women
As mentioned above, Japanese abortion statistics are known to be inaccurate as a result of underreporting. Matsuyama and colleagues estimated in the 1980s that only 60 percent of abortions were reported in Japan (Matsuyama 1988). According to a national survey of women in Japan, the abortion rate (the number of abortion cases per 1,000 women aged 16-49) was estimated at 25 (34 for single and 21 for married women) from May 1995 to May 1996 (Population Problems Research Council of the Mainichi Newspapers 1996).The approximate average ages of single and married women in this survey were 23.8 and 38.3, respectively. The official government overall abortion rate was reported in 1995 to be 11 abortions per 1,000 women aged 15-49, a figure reflecting substantial underreporting by physicians (Ministry of Health and Welfare, Minister's Secretariat 1998).These data suggest greater underreporting for single women, who tend to be younger than 25. Therefore, the abortion rate and the abortion ratio are probably underestimated here for younger age groups more than for older age groups. Furthermore, the overall level of underreporting could be rising, because the proportion of abortions among women in the younger age groups is shown to be increasing, and these groups are assumed to have a high rate of underreporting.
Three main factors have contributed to recent trends in abortion in Japan. First, sexual behavior and fertility among young people are changing. Sexual intercourse begun at an early age and before marriage is more prevalent than it was two decades ago. The birth-cohort analysis of the abortion ratio showed that it increased among young women born after 1955.This increase is consistent with the change in sexual behavior. A recent survey showed that the proportion of women aged 16--49and 20-24 who approved of premarital sex increased from 45 percent and 56 percent, respectively, in 1973, to 88percent and 89 percent in 1993 (NHK Hoso Bunka Kenkyujo 1998). Similarly, according to the National Fertility Survey of 1997,about 80 percent of single women and men approved of premarital sex (Institute of Population Problems 1999).Sexual behavior reflects this change in attitudes. The proportion of single women who have experienced premarital sex increased from 35 percent in 1990 to 49 percent in 1998 (Population Problems Research Council of the Mainichi Newspapers 1998), and a similar change was reported in the National Fertility Survey (Institute of Population Problems 1999). For high school and university students, the proportion of female students who were sexually experienced increased sharply from 6 percent and 11percent, respectively, in 1974 to 16percent and 43 percent in 1993 (Japanese Association for Sex Education 1998).
Marital status and fertility among women younger than 25 also changed greatly in the two decades between 1975 and 1995 in Japan (Retherford et al. 1996).The mean age of women at first marriage increased from 24.7 in 1975 to 26.3 in 1995 (Health and WeIfare Statistics-n-1998). The proportion of single women in the 20--24 age group increased from 69 percent to 86 percent during 1975-95, and the birth rate for this age group during the same period decreased markedly from 107 to 40 live births per 1,000 women (Health and Welfare Statistics Association 1998). Although young people are likely to postpone marriage and childbearing until their late 20s, they become sexually active at an earlier age than ever before, and thus, their need for contraceptives is increasing. This situation leads to a greater number of pregnancies that result in abortions among women younger than 25. The low social acceptability of childbearing outside of marriage also accounts for the increasing number of abortions among young people in Japan. The proportion of births to single women is only about 1 percent in Japan, whereas it is about 50 percent in Sweden and Denmark and 30 percent in France and England (Ministry of Health and Welfare 1998). In addition, cohabitation of couples in Japan was reported among only 2 percent of single women and men in 1997 (Institute of Population Problems 1999).Therefore, when a young single woman experiences an unintended pregnancy, she is likely to terminate it. In a 1997 survey, 65 out of 1,083 single women younger than 25 reported that they had been pregnant, and 57 of them reported that they had obtained abortions (Institute of Population Problems 1999).
A second factor is the limited number of contraceptive methods available and the extensive reliance on condoms in Japan. As mentioned earlier, oral contraceptives and the copper IUD were not approved until 1999, and other hormonal methods and progestin IUDs are still not available. Table 1 compares the age-specific distribution of contraceptive methods of Japan and France (Institute of Population Problems 1999; Toulemon and Leridon 1998). The table shows that people in all age groups in Japan rely mainly on condoms and that proportions using traditional methods (withdrawal and periodic abstinence) are high, compared with those using such methods in France. In addition, in Japan little difference is found in the distribution of contraceptive methods by age group, other than the slight increase in the proportions using IUDs and female sterilization among older women. By contrast, in France, the most commonly used methods are oral contraceptives and IUDs; the proportion using the pill decreases with age, and proportions using the IUD and choosing female sterilization increase with age. In short, Japanese couples are not using effective contraceptive methods that are suitable for each stage of their reproductive lives. Limited availability of effective, reversible contraceptives such as the pill together with increasing sexual activity among young people have contributed to the increase in abortions among them. In addition, the rare use of IUDs and sterilization could help explain why Japan's abortion rates among women in their late 20s and 30s are as high as are those among women in their early 20s.
Contraceptive choices among teenagers are not shown in the table. The proportion of sexually active single women aged 16-19 who were using contraceptives is reported to have increased from 43 percent in 1990 to 66 percent in 1998 (Population Problems Research Council of the Mainichi Newspapers 1998). The increasing abortion rate among teenagers suggests that their contraceptive use is not sufficient to prevent unintended pregnancies, however. In one survey, nearly 20 percent of high school and 10 percent of university students answered that they were not using contraceptives at all during intercourse (Japanese Association for Sex Education 1998). The most frequently used method among students is the condom (Japanese Association for Sex Education 1998), but their use of condoms may not be effective. One survey reported that nearly half of male and female high school or university students did not know how to use condoms correctly, especially when to put one on and when to remove it (Ikeda et al. 1986).
Another survey revealed that more than half of female university students did not know when ovulation occurs (Kishida and Sato 1998), although the second most frequently used contraceptive method among single women in Japan is periodic abstinence or the basal body temperature method (Japanese Association for Sex Education 1998).These data suggest that young women are undergoing abortions because they lack access to effective contraceptive methods and do not know how to use the available methods properly or at all. Poor access to abortion services is a third factor contributing to delay in seeking abortions among young women. Although abortion is legal in Japan, before performing the operation, physicians are required to obtain consent from the husband if the woman is married, from the partner if the woman is not married, or from the parent if the woman is single and younger than 20. For a young single woman, obtaining this consent can be a difficult process. The cost of an abortion may also be a barrier, because the procedure is not covered by insurance in Japan. An abortion typically costs around 100,000 yen (about US $900). This cost can be prohibitive for young women or men, especially if they are financially dependent on their parents. The average annual expenses for university students in 1996 were 2,000,000 yen (about US$18,200), a figure that includes about 300,000 yen (about $2,700) for personal use (Monbusho Kotokyoikukyoku Gakuseika 1997). Japan's recent abortion trends demonstrate that family planning policy should target women younger than 25 (Alan Guttmacher Institute 1998 and 1999), and at the same time, strategies should also be directed to all other age groups. Japan's maternal and child health services have persistently focused on improving care for pregnant women and infants (Miyaji and Lock 1994). In the immediate postwar period, when rapid population growth was a serious concern for the Japanese government, effective family planning campaigns were conducted. Since then, however, no new methods of contraception were introduced until 1999. People's contraceptive needs require greater attention in Japan in order to prevent further increases in the number of unintended pregnancies and abortions. This process will involve a number of steps.
First, access to effective modem contraceptive methods should be expanded. Women need opportunities to be informed of their options so that they can choose appropriate and effective contraceptives. The newly approved modem methods have the potential to reduce unintended pregnancies. In 1996,Maruyama and colleagues calculated that 28 percent of abortions in Japan could be prevented by the approval of low-dose oral contraceptives. Oddens and Lolkema (1998) also predicted a reduction in unintended pregnancies from 13 percent to as high as 58 percent, depending on the rates of oral contraceptive use. Several obstacles bar an increase in the use of the pill in Japan, however. The first is Japanese women's reluctance to seek gynecological care (Coleman 1981). The low-dose pill requires a prescription from a gynecologist, but women are not accustomed to visiting gynecologists, except when they are pregnant. Even among women older than 30, who are advised by the Ministry of Health and Welfare to receive a cervical cancer screening test annually, only 15 percent received a pap smear in 1996 (Ministry of Health and Welfare, Minister's Secretariat, 1996). A second obstacle is cost. Guidelines recommend that physicians conduct a gynecological examination, blood tests, and tests for sexually transmitted diseases every six months. Neither these examinations nor oral contraceptives are covered by health insurance, however, and the cost varies according to the physician. A recent survey of gynecologists found that the median monthly cost of contraceptive pills was 2,700 yen (about $25) (Saotome 2000). For examinations required on the first visit, the cost ranges from no charge to more than 10,000yen (about $90); most first visits were in the range of 3,000-10,000 yen (about $27-90) (Saotome 2000). A third obstacle is physicians' lack of interest in providing family planning services. A survey conducted by the Ministry of Health and Welfare of some 200 gynecologists reported that only half said they always gave advice about contraception to women who had had abortions (Hayashi 1996). Another survey of women who had recently given birth reported that only half had received family planning advice after delivery (Kaneko et al. 1996).Gynecological practice in Japan needs to become more woman-friendly and family planning, an important part of primary care, should receive greater attention in gynecological and obstetrical practice. Japan also should improve its national programs for sexual and reproductive health education (Ishii et al. 1988; Matsumoto 1990 and 1995). In Sweden, the introduction of a new sex-education curriculum in 1975 resulted in a decreased incidence of abortion and a lowered fertility rate among teenagers (Santow and Bracher 1999). The new curriculum did not explicitly recommend abstinence, and it included free contraceptive counseling and provided oral contraceptives and condoms at schools. In Japan, wide differences exist among schools in emphasis on and methods of sex education, and schools and health organizations do not cooperate (Matsumoto 1995; Anonymous 1999). Young people in Japan today need practical reproductive health knowledge and skills, including information on the newly approved contraceptives and on sexually transmitted diseases. Japanese abortion statistics indicate that the parents of the younger generation may not know enough about reproductive health to educate their children at home or even enough to protect themselves from the risk of unintended pregnancy. Opportunities to provide reproductive health information to older people can be found in health-promotion information networks and health-guidance sessions. Parent-teacher association seminars on sex education and HIV infection are held at a few elementary, junior high, and high schools in Japan, and should be held more frequently and widely. The school sex-education and the health-guidance sessions in municipalities are organized separately by the ministries of education and health and welfare. These different organizations should collaborate to develop comprehensive family planning strategies at the municipal level. Japan should improve its collecting of epidemiological data on reproductive health and its analysis of the available data. Efforts should be made to reduce the high level of underreporting of abortions and to adjust the existing data for underreporting. The current insurance system in Japan does not cover abortions, so physicians set abortion fees at their own discretion. Thus, income from unreported abortions is not included in tax reports, creating a strong motivation for physicians to maintain the current reporting system. Moreover, the penalty for underreporting is less than 100,000yen. Abortion and contraceptive services should be covered by insurance in order to improve women's access to those services and also to monitor physicians' abortion practice. The importance of abortion surveillance should be explained to physicians, and the penalty for underreporting increased. In addition, research is needed on unintended pregnancy leading to abortion in Japan. According to
the Eleventh National Fertility Survey in 1997, married women reported that they had planned to have a child in only 34 percent of pregnancies after marriage (Institute of Population Problems 1999).Few epidemiological studies have been conducted, however, about the factors that contribute to the occurrence of unintended pregnancy in Japan. Such studies are important for developing appropriate policies to reduce the number of abortions. In 1999,the Ministry of Health and Welfare's Research Group for the Life-long Supports for Women's Health presented recommendations to reduce the number of induced abortions by half over the next ten years. In January 2000, the Ministry announced plans to develop "Healthy Family 21,"a maternal and child healthcare plan for the next decade. The plan will include an adolescent reproductive health section aiming to reduce abortions among teenagers. The Ministry's suggested strategies do not provide for improvement in the collection and analysis of reproductive health data, however. In order to prevent an increase in unintended pregnancies and abortions, Japan must develop a comprehensive family planning program, including improvements in gynecological care, reproductive health education, and policymaking based on an analysis of people's contraceptive needs.
Aya Goto is Public Health Fellow, Population Council. Akira Fukao is Professor, Yamagata University School of Medicine, Department of Public Health. Chihaya Fujiyama-Koriyama is Lecturer, Kagoshima University School of Medicine, Department of Public Health. Michael R. Reich is Chair, Harvard School of Public Health, Department of Population and International Health. Correspondenceshould be addressed to Dr. Goto at Tu Du Obstetrical and GynecologicalHospital, 284 Cong Quynh Street, Ho Chi Minh City, Vietnam.