3.1 Maternal Age

Maternal age has been shown to be a strong, independent risk factor for miscarriage in a large prospective register linkage study [6]. The UK census data from 2011 reports that the standardised average age of mothers at first childbirth was 27.9 years, compared to 23.7 years in 1971. The report also found an increased percentage of births to mothers aged 35 years or more, which almost trebled from 7.5% in 1971 to 20% in 2011. The age-related risk of miscarriage in recognised pregnancies increases sharply after the age of 35 years, rising from 11% at 20–24 years to 93% at 45 years and older. In women aged 40 years or older, only 41.7% achieved live birth within 5 years as opposed to 81.3% of women aged 20–24 years [5]. Women aged 20 years with 2 previous miscarriages have a 92% chance of success in the next pregnancy compared with only a 60% chance of success in women aged 45 years [7].

3.2 Number of Previous Miscarriages

While around three quarters of women with 3 miscarriages achieve a live birth within 5 years, only about half have successful pregnancy following 6 or more previous miscarriages [5]. The risk of further miscarriage increases by 45% after 3 or more pregnancy losses [6].

3.3 Environmental Factors

A matched case control study by Lashen et al. [8] demonstrated significantly higher odds of RM in obese women (OR 3.51, 95% CI 1.03–12.01). In couples with RM, maternal obesity has been found to be an independent factor increasing the risk of miscarriage in a subsequent pregnancy.

Frequent and excessive maternal alcohol consumption, passive smoking and caffeine consumption in the peri-conceptual period have been associated with an increased risk of RM. However, there is no evidence from randomised controlled studies about lifestyle adaptations to improve reproductive outcome in women with otherwise unexplained RM.

4.1 Genetic Abnormalities

It is estimated that only 30% of the embryos are successful in resulting in a live birth. The remainder is either lost prior to implantation (30%) or post-implantation, 30% as early pregnancy loss (before 6 weeks’ gestation) and 10% as clinical miscarriage [9]. The exceptionally high attrition rate of human embryos is attributed mainly to the high prevalence of chromosomal and genetic aberrations throughout all stages of pre-implantation embryo development. Chromosomal abnormalities account for about half of all the cases of sporadic miscarriages and amongst these abnormalities, about 50% are due to chromosomal segregation errors such as trisomy. The rest are non-trisomies, predominantly monosomy X and triploidy [10]. Although chromosomal segregation errors can lead to miscarriage, these have not been identified as risk factors for RM, but can provide useful information about the prognosis for future pregnancy.

Chromosomal translocations refer to chromosome abnormalities that are caused by the rearrangement of segments of DNA between non-homologous chromosomes. Translocations are of two types: reciprocal translocation in which there is an exchange of two terminal segments from different chromosomes, or Robertsonian translocation in which there is centric fusion of two acrocentric chromosomes with the loss of the short arms. In couples with two or more consecutive miscarriages, a carrier rate of 3–6% for either a Robertsonian or reciprocal translocation has been reported [11]. Carriers of a balanced reciprocal translocation themselves usually have no loss or gain of genetic information and hence are usually phenotypically normal. However, 50–70% of their gametes may have an unbalanced chromosomal pattern at meiosis resulting in embryos inheriting an unbalanced translocation that is in turn associated with an increased risk of miscarriage. The miscarriage risk for couples where one is a carrier of a reciprocal translocation can be up to 72.4% [12]. Children born with unbalanced translocation generally have multiple congenital malformations and/or mental disability [4]. The importance of recognising the risk is not that it can be treated but that there may be an opportunity to avoid the birth of an affected child e.g. through antenatal screening and termination of the pregnancy after appropriate counselling or in selected cases through pre-implantation diagnosis.

4.2 Thrombophilias

Pregnancy is naturally a hypercoagulable state which is protective to the pregnant mother especially during the intra-partum and post-partum period. However, when this natural pivotal change in the pregnancy physiology is affected by a thrombophilic defect, early and late adverse pregnancy outcomes can occur. Thrombophilias, both inherited and acquired, are known to cause a predisposition to venous and arterial thrombosis. The thromboses in decidual vessels are thought to be related to intrauterine growth restriction, fetal death and recurrent miscarriage [4].

4.3 Structural Abnormalities – Uterine and Cervical

Congenital uterine abnormalities, depending on the type and degree of anatomical distortion, are thought to impair reproductive outcome. They are associated with an increased risk of spontaneous miscarriage and other pregnancy complications such as preterm labour, fetal malpresentation, low birth weight and increased perinatal mortality rates [16]. Although the exact aetio-pathophysiology remains uncertain, they are found to be significantly more prevalent in women with miscarriage than in the general population. The incidence of congenital uterine anomalies may range from 2.7% to 16.7% in the general population or in fertile women and from 1.8% to 37.6% in women with two or more consecutive miscarriages [17]. The prevalence of all congenital uterine anomalies in women with two or more miscarriages appears to be similar to those with three or more miscarriages.

Fibroids are common, benign tumours of the uterus occurring in up to 77% of women approaching the age of 50. The prevalence of fibroids in women with RM is reported as 8.2% [18]. Although cavity distorting sub-mucous and intramural fibroids and large non-cavity distorting intramural fibroids may interfere with implantation [19], their effect on the risk of RM remains poorly understood. An observational study found that mid-trimester miscarriages were significantly commoner in women experiencing RM with either cavity distorting or non-cavity distorting fibroids compared to women with otherwise unexplained RM [18]. In this study, resection of fibroids distorting the uterine cavity removed the risk of mid-trimester loss and increased live birth rate by twofold in subsequent pregnancies [18].

Intrauterine adhesions (Asherman’s syndrome) have been thought to increase the risk of RM by endometrial fibrosis and inflammation and diminishing intrauterine volume [20]. The reported incidence of intrauterine adhesions on diagnostic hysteroscopy following three or more spontaneous miscarriages is around 10%, but a study found that 97.5% of women with three or more miscarriages have had two or more surgical management of miscarriage procedures [20]. However, there is no evidence from prospective studies to confirm the causal relationship of intrauterine adhesions to RM.

A diagnosis of cervical weakness is based on a history of recurrent mid-trimester miscarriages or recurrent preterm deliveries following painless cervical dilatation in the absence of contractions, bleeding or other causes of RM. The exact incidence of cervical weakness in women with RM is unknown and there are no objective tests that can reliably identify women with cervical weakness in the non-pregnant state [1].

4.4 Endometrial Causes

There is a growing body of evidence from in-vitro and animal studies to suggest that decidualising endometrial stromal cells serve as sensors of embryo quality upon implantation and that any perturbations in endometrial decidualisation may cause recurrent miscarriage [9]. It has also been suggested that in some women with RM the duration of uterine receptivity is prolonged, widening the implantation window. Decidualising stromal cells from women with RM may fail to discriminate between high and low quality embryos and thereby allow developmentally compromised embryos to implant resulting in miscarriage which may then be recurrent. There are no standardised investigations however, to assess endometrial dysfunction and the current treatments to improve endometrial function are empirical.

4.5 Endocrine Abnormalities

Uncontrolled diabetes mellitus is a known risk factor for miscarriages and congenital malformations, but well managed, diabetes alone is not found to be a risk factor for miscarriage and thus should not cause recurrent miscarriage.

Physiological changes of pregnancy demands increased thyroid hormone production which is adequately accommodated by a normal thyroid gland. An underactive thyroid gland may not be able to meet the demands of changes in pregnancy; however, there is no evidence to suggest that women with miscarriage or RM have more overt or subclinical hypothyroidism.

The role of anti-thyroid antibodies in RM is not completely established. The prevalence of thyroid peroxidase antibodies (TPO-Ab) in women with recurrent miscarriage is reported to be higher. In a systematic review, it was identified that the presence of thyroid antibodies was associated with an increased risk of RM (OR 2.3, 95% CI 1.5–3.5), when compared with the absence of thyroid antibodies [21].

The role of prolactin in RM is controversial. Prolactin is essential for female reproduction and high levels are associated with ovulatory dysfunction. Hyperprolactinaemia is thought to affect the hypothalamo-pituitary-ovarian axis, resulting in impaired folliculogenesis, oocyte maturation and/or a short luteal phase [2]. In one study bromocriptine-treated hyperprolactinaemic women with RM had a significantly higher percentage of successful pregnancies overall [22]. An observational study, however, identified that those women with RM and significantly lower serum prolactin levels were at an increased risk of miscarriage in subsequent pregnancy [23].

Polycystic ovarian syndrome (PCOS) is the most common endocrine condition in women of reproductive age. The reported prevalence of PCOS in women with RM ranges from 4.8% to 82%, and the wide range of apparent prevalence is the result of the huge variation in the diagnostic criteria used for PCOS before the establishment of the Rotterdam criteria [24]. PCOS has been associated with an increased risk of miscarriage, but the exact mechanism remains unclear. The markers of PCOS are not predictive of pregnancy loss following spontaneous conception amongst ovulatory women with RM. The increased risk of miscarriage in PCOS women may be due to hyperandrogenaemia, hyperinsulinaemia and insulin resistance. The prevalence of hyperandrogenaemia in RM women is 11% [25]. In RM women, increased follicular phase free androgen index is associated with an increased risk of miscarriage in subsequent pregnancy [25]. Insulin resistance is prevalent in 17–27% of women with RM [24]. Hyperinsulinaemia in PCOS is attributed to obesity as well as to insulin resistance independent of body weight. It is found to be an independent risk factor for miscarriage and is thought to play a key role in implantation failure by suppressing circulating levels of glycodelin and IGFBP-1. However, there is no strong association between hyperinsulinaemia and RM.

4.6 Infection

There is a paucity of evidence to suggest infections such as toxoplasmosis, rubella, cytomegalovirus, herpes, listeria and bacterial vaginosis cause first trimester RM. Therefore, routine screening for these infections as a part of RM evaluation is not recommended and neither is any routine use of antibiotics.

4.7 Immune Dysfunction

The survival of a genetically ‘foreign’ semi-allogenic fetus and the placenta requires significant regulation of the maternal immune system and this maternal tolerance is mediated by MHC proteins, endometrial leukocytes and the cytokines.

HLA class 1b alleles and polymorphisms (HLA –C, –E, -F and –G) seem to induce suppression of the maternal immune system, but their exact link to pregnancy complications such as RM has not been convincingly established [26].

Uterine natural killer (uNK) cells are the most abundant of all decidual leucocytes. They are a rich source of cytokines and growth factors and are thought to play a role in conferring immunotolerance towards paternal antigens, in decidualisation by regulating trophoblast invasion and vascular remodelling. There is no convincing evidence to suggest that uNK cell density is different in women with RM when compared to controls, and the uNK cell density does not appear to predict future pregnancy outcome.

Peripheral NK cells are phenotypically and functionally different from uNK cells and have no role in decidualisation, implantation and hence in RM.

4.8 Psychological

Miscarriage can be a distressing event and can induce obvious emotional responses, such as anxiety, depression, denial, anger, marital disruption, and a sense of loss and inadequacy [4].

4.9 Male Factor

Studies examining the role of sperm DNA damage in RM patients have shown mixed results. There is insufficient evidence to suggest that tests of DNA fragmentation have predictive value in the prospective identification of women at risk of RM. At present, sperm DNA testing is not recommended as a part of the clinical evaluation of RM.