Calibrated prevalence of infertility in 30- to 49-year-old women according to different approaches: a cross-sectional population-based study.

STUDY QUESTION: How does the estimated prevalence of infertility among 30- to 49-year-old women vary when using different approaches to its measurement?

SUMMARY ANSWER: The prevalence of women with difficulties in conceiving differed widely according to the measurement approach adopted.

WHAT IS KNOWN ALREADY: Establishing the true magnitude of infertility as a public health problem is challenging, given that it is not categorized as a disability or chronic condition and may be largely unreported. The time required to conceive is an increasingly frequent concern among couples of reproductive age. Population-based studies do not consider multiple approaches to infertility measurement in the same sample.

STUDY DESIGN, SIZE, DURATION: A face-to-face cross-sectional population-based survey of 443 women aged between 30 and 49 years residing in Huelva, southern Spain, was carried out. The sample size estimation was based on an assumed prevalence of infertility of 19%, a sampling error of ±4.84 percentage points, a design effect of 1.8 and a 95% confidence level. The information was collected in 2011.

PARTICIPANTS/MATERIALS, SETTING, METHODS: Self-reported information was gathered on socio-demographic data, pregnancy history, time required to become pregnant and perception of difficulties in becoming pregnant. Eight approaches to the estimation of infertility prevalence were considered: diagnosed infertility, subjective infertility, 1-year infertility, primary infertility, secondary infertility and subfertility based on the time taken to conceive (6, 12 or 24 months). Calibration estimators (indirect estimation techniques) were used to extrapolate the infertility prevalences to the whole of Spain.

MAIN RESULTS AND THE ROLE OF CHANCE: The response rate was 61.05%. Among 30- to 49-year-old Spanish women, 1.26% had a clinical diagnosis of infertility, 17.58% did not achieve pregnancy in 1 year (1-year infertility), 8.22% perceived difficulties in procreation (subjective infertility), 6.12% had not succeeded in having biological children (primary infertility) and 11.33% had not been able to have another biological child (secondary infertility). Finally, pregnancy was not achieved within 6, 12 and 24 months of starting to attempt conception in 19.98, 11.21 and 4.36% of women, respectively. These approaches to estimate the prevalence of infertility show similar socio-demographic patterns except for educational level. Calibration adjustments allowed extrapolation of these prevalences to Spain and a reduction of from 3.7 to 90.4% in their variances.

LIMITATIONS, REASONS FOR CAUTION: The response rate was moderate but acceptable in comparison to similar studies. We only asked whether the women had practiced intercourse without contraceptive methods for >1 year. Hence, we could only calculate the time for which a couple were trying to become pregnant when the woman became pregnant, and we do not know whether it was longer than 1 year for the women failing to conceive. Future research should avoid this study weakness by gathering quantitative data on the months during which vaginal sexual intercourse was practiced, with no time limit. The only prevalence with a high coefficient of variation was that for diagnosed infertility, and our estimation for this prevalence should be interpreted with caution.

WIDER IMPLICATIONS OF THE FINDINGS: Despite major differences according to the approach adopted, the prevalences of infertility estimated in our study are comparable with those obtained in other similar studies. A set of categories to measure infertility including subjective infertility, 1-year infertility and subfertility have been proposed here, which may be useful for cross-disciplinary comparisons of infertility in clinical and population-based studies. These measures may also assist health managers to tailor fertility resources and services to the real needs of the population and provide a more rapid and effective response to couples. Finally, the calibration adjustments (indirect estimation techniques) applied to the infertility prevalences help to maximize their generalization and improve accuracy. This technique may be considered as a model for application in other epidemiological studies.