A Retrospective Study On Fecundity Trend Among Population In Tongliao, China

BackgroundEvidence for a secular decline in sperm concentration in some places, together with the growing demands for assisted reproductive technology and increasing incidence of genital tract abnormalities, lead to the assumption that human biological fecundity has been declining in recent decades. However, findings from various population studies in Europe and USA were contrary to expectations, indicating that human fecundity has either approved or remains unchanged in last decades. Although heterogeneity of methodologies and spatial variation was plausible explanation, the true extent of and time trend in human fecundity remains inconclusive.China had begun implementing the world's most stringent family planning policy since around1980, which limited most couples to one child. Another cornerstone policy, economic reform and open door policy, was also established simultaneously, and has brought about overall socioeconomic changes. We anticipated that some of these changes might also have influences on fecundity. Therefore, our objective in present study was to examine whether there was a secular trend of fecundity among Tongliao northern population during1981-2008with a retrospective study design.MethodsThe survey was conducted in Tongliao from Dec.2008-May,2009. Sample size was calculated with the formula for prevalence investigation, parameters for calculation were determined with reference to previous studies. Staff of grass-level population and family planning commission was trained to conduct face-to-face interviews with eligible women. Information on reproductive history of first pregnancy and demographic characteristics was retrospectively collected after written informed consent was obtained. We calculated the cumulative pregnancy probability at the third month, sixth month and twelfth month by demographic characteristic to describe fecundity of population in different status, using life test technology. Discrete time survival model was used to identify the influence of covariates. The cumulative pregnancy probability was further adjusted by correction coefficient, which was determined by magnitude, direction of influence of each covariate, as well as distribution. We also introduced join-point regression model to describe changes of fecundity over time with the consideration that human fecundity may not decline or increase smoothly during a long-time period. We performed sensitivity analysis by excluding those who reported pregnancy in the first month from the main analysis with the attempt to detect and eliminate the potential biases.Descriptive analysis and discrete time survival model were carried out in SAS9.1. JPM was performed in Joinpoint Regression Program, Version3.4.3(National Cancer Institute,2010).ResultsOf40,508eligible women,38,367were interviewed, the response rate was94.7%, proportion of valid questionnaire was86.6%. A total of33,224subjects were included in the final analysis data. The adjusted cumulative pregnancy rates at third, sixth and twelfth month were61.74%,75.00%and85.99%, respectively. Infertility rate was14.01%.Mongolia women had a lower fecundity than Han women (OR=0.943, CI:0.917-0.969). Respondents aged21to25had the highest fecundity than other age groups. Women who used IUD for contraception before planning pregnancy had a lower fecundity than other contraceptive users. Education degree and occupation had no significant influence on fecundity.Fecundity increased1.24and1.53times during1990s and2000s respectively, compared with that in1980s. Two join points were observed at year1993and2002, which cut the whole line into three segments:1981-1993,1993-2002and2002-2008. Fecundity significantly increased at the first segment from77.40%to86.65%, followed by a slight increase in the second segment from86.65%to87.01%, and a steep rising in the third segment from87.01%to93.14%. The annual percentage change (APC) was1.07%,0.09%and1.62%during the three time periods, respectively.Conclusion:Fecundity among people in northern China has been increasing during the time period from1981to2008. Biases including truncation bias, protection bias, recall bias and planning bias are insufficient to explain this rise. It was also unlikely caused by STD and ART. We speculate that this was resultant from the dramatic social change during this period. However, large sources of biases have the potential to distort the results, so cautious must be highlighted when explaining and generalizing the conclusion.