Infertility, infertility treatment, and congenital malformations: Danish national birth cohort

Introduction

About 10-20% of couples who are trying to become pregnant have a waiting time to pregnancy longer than 12 months, the clinical definition of infertility in most industrialised countries.1 An increasing number of couples seek treatment for infertility, as reflected by the rising number of children born after assisted reproductive technology.2 Children conceived after assisted reproductive technology have an excess of congenital malformations compared with children conceived spontaneously,^3–5 but little research has been devoted to attempting to separate the effect of assisted reproductive technology from that of infertility itself. In a recent review, Rimm et al pointed out that previous studies on the topic have lacked a comparison group of children born of infertile couples who conceived naturally.5 Some of the determinants of infertility may share a common causal path with mechanisms that cause congenital malformations. Using a large cohort of Danish couples, we estimated the prevalence of congenital malformations as a function of infertility and its treatment.

Methods

We did the study within the Danish national birth cohort.6 From June 1997 to February 2003, 85 381 women (92 892 pregnancies) responded during the first or second trimester to the first of four scheduled telephone interviews. Women were asked if their pregnancy was planned and how long they had tried to become pregnant before succeeding. Possible answers were: “right away,” one to two months, three to five months, six to 12 months, and > 12 months. If they reported a time to pregnancy of six months or longer, participants were also asked if they or their male partner had received any infertility treatment. Couples who reported a time to pregnancy of less than six months were not asked about treatment, and we classified them as not treated. We identified three groups of planned pregnancies: 52 380 conceived by fertile couples (time to pregnancy ≤ 12 months), 5910 conceived naturally by infertile couples (> 12 months), and 5564 conceived by infertile couples after infertility treatment. The first group comprised all couples with a time to pregnancy of ≤ 12 months and no report of infertility treatment. The second group included all couples with a time to pregnancy of > 12 months and no infertility treatment. The last group included all couples with a report of treatment for infertility before the pregnancy. If a woman contributed with more than one pregnancy, we included only the earliest.

If a woman reported that she or her male partner had received infertility treatment, she was then asked “what kind of infertility treatment did you or your male partner receive?” She could choose one or more of six specified forms of treatment and describe other types of treatment in free text. The specific treatments included intracytoplasmic sperm injection, in vitro fertilisation, intrauterine insemination, hormonal treatment, surgery (such as removal of adhesions or fibroid in fallopian tube), and hysterosalpingography (rinsing of fallopian tubes). Women who reported more than one procedure were classified according to the above priority sequence. We combined surgery and hysterosalpingography into one group and categorised an additional group as alternative treatments (for example, acupuncture, zone therapy, diet, or vitamins), on the basis of the free text description.

We linked the cohort to pregnancy outcomes by using the national hospital register and the medical birth register,⁷⁸ by means of the unique civil registration number. We excluded pregnancies that ended in spontaneous abortion (n = 628), induced abortion (n = 116), hydatidiform mole (n = 11), or ectopic pregnancy (n = 1) and pregnancies with unknown outcomes because of emigration or other reasons (n = 44). No differences existed in spontaneous abortion or induced abortion among the studied groups (P = 0.13 for spontaneous abortion; P = 0.73 for induced abortion). We further excluded all stillbirths (n = 209; of these, 16 were of one twin, and four were of both twins) because of incomplete information on malformations and all triplets (n = 84, all live births) owing to small numbers. The rate of stillbirth differed among the three groups (0.3% for fertile couples, 0.4% for untreated infertile couples, and 0.5% for treated infertile couples; P = 0.01). After these exclusions, 61 249 singletons and 3156 twins remained available for analysis.

Diagnoses on congenital malformations (ICD-10 (international classification of diseases, 10th revision) Q00-Q99 codes) are recorded in the national hospital register.7 We excluded two minor malformations (accessory auricle (Q170) and pigmented nevus (Q825)) and three malformations with possible uncertain diagnoses (patent ductus arteriosus (Q250), undescended testis (Q53), and hip dislocation (Q650-6)).

Results

The median follow-up time was four years from the time of birth. Of the 293 children who died during the first year of life, 125 (42.7%) had a diagnosis of congenital malformations.

Compared with singletons born of fertile couples, singletons born of infertile couples had a higher prevalence of congenital malformations, regardless of treatment (table 1). Among singletons born of infertile couples who conceived naturally, we saw an increased prevalence of malformations of the nervous system, digestive system, and musculoskeletal system. Singletons born of couples who received infertility treatment had a similar pattern, as well as a higher prevalence of malformations of genital organs and “others” (see extra table on bmj.com for results on detailed diagnoses with two-digital codes in ICD-10). When we restricted the analysis to singletons born of infertile couples, we saw an excess only of genital organ malformations associated with treatment (table 1). We could not attribute this increase to any specific malformations of the genital organs.

Among singletons born of untreated couples, prevalence of congenital malformations increased with increasing time to pregnancy (table 2). We saw a similar, but not statistically significant, trend in singletons born of treated couples.

Table 3 and table 4 show congenital malformations in singletons born after different forms of infertility treatment. Compared with singletons born of fertile couples, children born after intracytoplasmic sperm injection, in vitro fertilisation, intrauterine insemination, hormonal treatment, or surgery had an increased prevalence of congenital malformations, with hazard ratios between 1.31 and 1.85 (table 4). When we restricted the analysis to infertile couples, only intracytoplasmic sperm injection was associated with a significantly higher overall prevalence of congenital malformations (hazard ratio 1.57, 95% confidence interval 1.11 to 2.23). We saw a significantly increased prevalence of genital organ malformations (4.39, 1.58 to 12.21) and musculoskeletal system malformations (1.67, 1.01 to 2.76).

Twins had an overall higher prevalence of congenital malformations than singletons (7.0% v 5.2%; P < 0.0001). However, we saw no significant association with infertility, infertility treatment (table 5), or type of treatment (data not shown).

We obtained similar results when we restricted our analyses to firstborn children or to children born at term (data not shown).

Discussion

Conclusions

Our results indicate that, in order to properly evaluate the effects of assisted reproductive technology, infertile patients who conceived spontaneously should be used as a reference. Infertile couples should be offered prenatal screening, and research should be devoted to finding the mechanisms behind the association between congenital malformations and infertility or subfertility.