In China, reproductive health care is widely available through National Family Planning services. In the last three decades, China has experienced dramatically social changes associated with rapid economic growth and reform. Traditional attitudes towards unmarried cohabitation and premarital sex have also been changed quickly [9, 13, 23]. More young adults with sexual relationship are not legally married. However, social responses to these changes are slow . The current Family Planning Programme still targets mainly on married couples rather than unmarried young people .
We found unmarried graduates’ average age was 29.4. It was much higher than the legal age for marriage (≥22 for male and ≥20 for female) in China. During the long period between their initiation of sexual activity and marriage, 31.0% of male graduates and 17.5% of female graduates reported a history of premarital sexual intercourse, which was higher than those among undergraduate students (males: 15.4%~20.2%; females: 9.4%~10.6%) [2, 4, 8] and among their out-of school counterparts (males: 20.2%~21.4%; females: 7.1%~8.7%) [14, 25]. According to the data, 70.0% of males and 70.2% of females had a steady lover. We believed that the real prevalence rate of premarital sexual intercourse was much higher than reported. It seemed that unmarried graduate students also need the family planning services.
The prevalence rate of unintended pregnancy among graduate students active in premarital sex was 19.8% (571/2890). It was lower than undergraduate students (23.1%~31.5%) [8, 26, 27] and their out-of school counterparts(50%~64.2%) [14, 28], which might indicate graduate students could prevent themselves from unintended pregnancy more effectively. But we also found nearly all pregnant graduates (96.7%, 552/571) chose abortion to end pregnancy, which was much higher than undergraduates (78.7%)  and their out-of school counterparts(25%) . The possible reasons for this might be as fellows: (a) they had no money to feed a baby. Most of them could only get poor living allowances from schools to feed themselves (graduates’ disposable income per month was 727 Yuan; Chinese Per Capita Month Disposable Income of Urban Households was 1315 Yuan and Engel's Coefficient of Urban Households was 37.9% in 2008 ); (b) they had no time to feed a baby. Most of them have to spend all their time studying or looking for jobs. In contrast, their out-of-school counterparts who had premarital pregnancy chose to have prompt marriage in most cases ; (c) the sexual relationship was not stable for some graduate students. From these, we believed that the convenient family planning services for graduate students was in demand.
We found students with a steady lover had a significantly lower rate of unintended pregnancy. It might indicate that students with a steady lover were much better in preparation for contraception. In contrast, “one-night-stand” students without a steady lover had difficulty in preparing contraception timely due to the risk of being stigmatized for their searching of contraceptive services . Hence, we believed that convenient contraceptive services offer for them might effectively decrease the rate of unintended pregnancy.
Students having sex at younger ages were more likely to report a history of unintended pregnancy in both genders in our study. We presumed that two factors might be important. The first was that students who initiated sex at early age were lacking in awareness of the risks of pregnancy and utilized contraception less often than late initiators did . The second was that the students with earlier initiation of sexual intercourse would be exposed to the longer periods of the risk of unintended pregnancies .
Our data showed both genders that used condom in the first sex had a significantly lower rate of unintended pregnancy. It might be that those who used condoms in the first sex has increased likelihood of subsequent condom use . It was regretted that the rate of condom use in the first sex was only 30.4% for females' partners and 36.9% for males. Condoms, OCs, withdrawal and periodic abstinence were considered to be the most common contraceptive methods among unmarried people [6, 8]. However, the last two methods often led to contraceptive failure . We found that 29.9% of male graduates and 33.8% of female graduates always used contraception (methods were looked as effective for preventing pregnancy by responders) in their sexual intercourse. But the behavior could not effectively prevent them from unintended pregnancy. We also noticed that OR for female was <1, but OR for male was >1 of being pregnant (always use contraception that were looked as effective for preventing pregnancy by responders VS never use). We presumed that one of possible reasons was that self-reported ‘use of contraception' was not necessarily reliable method. Another possible factor was that male students had more possible classify experience of unintended pregnancy to contraceptive failure instead of absence of contraception. All of these indicates that how to select and how to use contraception correctly might still confuse graduate students.
Unawareness of the conditions of conception and unawareness of that abortion endangers women’s future pregnancy were related to pregnancy for both genders. It indicated that improving the awareness of reproduction and contraception might decrease the prevalence rate of unintended pregnancy among graduate students.
Our current data unexpectedly showed that medical students were not less likely to have unintended pregnancy than nonmedical students. In China, nearly all school and college do not set up reproductive and contraceptive courses. And due to the constrained Chinese traditional concept of chastity, students can not obtain correct contraceptive knowledge from society or family. We presumed that medical graduates might know more about the function and structure of the reproductive system than nonmedical graduates, but still lacked the knowledge of how to prevent unintended pregnancy.
Based on the multistage, stratified, cluster design, our sample came from diverse regions of China involving various disciplines and university levels. It could reduce the selection bias and nondifferential misclassification in the greatest possible way. Therefore, we believed our findings could be generalized. However, our study might have some bias: (a) for the sensitive question of premarital sex, no special technique (such as randomized responses) was used in the study, which might lead to underestimated rate of premarital sex; (b) for those male students who were reported unintended pregnancy for their partners, we could not judge whether the unintended pregnancy was caused by themselves or not, particularly for those who did not mention a steady lover; and (c) cause-effect relation could not be established and recall bias could not be avoided due to the cross-sectional nature of the study.