Miscarriage: Vitamin D is the only easily changeable major risk factor

Top 10 risk factors for recurrent miscarriages

By Claude AI - Oct 2026

There's no official "top 10," so I ranked these by size of effect and strength of evidence. In as many as half of couples with RPL, no cause is ever found [Inversetti 2023].

  1. Maternal age. Across Norwegian registry pregnancies, miscarriage risk was lowest at ages 25–29 (about 10%), climbed quickly after 30, and reached 53% at age 45 and older [Magnus BMJ 2019]
  2. Number of previous losses. Age-adjusted odds of another miscarriage were 1.54 after one loss, 2.21 after two, and 3.97 after three consecutive losses [Magnus BMJ 2019]
  3. Random chromosomal errors in the embryo. Embryonic aneuploidy shows up in as many as 90% of losses at 0–6 weeks, roughly half at 8–11 weeks, and 30% at 16–19 weeks [Genetics review]. Women with more prior losses are a partial exception: for them, the lost pregnancy is more likely to be chromosomally normal [PMC9783143]. That points toward maternal causes.
  4. Antiphospholipid syndrome. RCOG calls it the most important treatable cause. The antibodies are found in 15% of women with recurrent miscarriage, compared with under 2% of low-risk women [RCOG GTG 17]. Under strict criteria the numbers are much lower: a 2025 Brazilian cohort found confirmed positivity in only about 1.5% [Oliveira 2025]. The standard treatment is low-dose aspirin plus heparin.
  5. Uterine anomalies. Reported prevalence in recurrent-miscarriage populations ranges from 1.8% to 37.6%, and one review pooled it at 16.7% [Yeung & Regan]. The main congenital anomaly is a septate uterus. Acquired problems include submucosal fibroids, polyps and adhesions, and ESHRE's 2022 update added a new recommendation on investigating adenomyosis [ESHRE 2022].
  6. Balanced parental translocation. In 2–5% of affected couples, one partner carries a balanced reciprocal or Robertsonian translocation [RCOG GTG 17]
  7. Thyroid dysfunction and autoimmunity. Thyroid autoimmunity affects 5–20% of reproductive-age women, making it the most common autoimmune condition in that group [Thyroid meta 2020]. A meta-analysis of 15 studies (1,911 women with RPL plus subclinical hypothyroidism or TPO antibodies) found that levothyroxine improved outcomes [Levothyroxine meta]
  8. BMI extremes. Being underweight (OR 1.2) and having a BMI over 25 (OR 1.21) both raised the odds of RPL. Among women who already had RPL, a BMI over 30 raised the odds of another loss (OR 1.77). The evidence quality was rated low or very low. [Ng 2021]
  9. Male factor. Pooled data from 13 studies showed higher sperm DNA fragmentation in male partners of women with RPL (mean difference about 12) [McQueen meta]. An umbrella review rated that link as highly suggestive (class II) [Umbrella review 2024]. Male age over 40 is also a risk factor [Lancet 2021].
  10. Low vitamin D. Women below 50 nmol/L had nearly twice the odds of miscarriage of women above 75 nmol/L (OR 1.94). Pooling deficient and insufficient women gave OR 1.60. The authors noted it is still unproven whether correcting deficiency before conception prevents loss [Tamblyn 2022]. ESHRE does not recommend testing vitamin D in RPL, but says preconception counselling could include advice to consider prophylactic supplementation [ESHRE via Dutch adaptation]

Near misses. These have weaker or less RPL-specific evidence:

  • The Lancet series also lists smoking, alcohol, stress, night-shift work, air pollution, pesticide exposure and Black ethnicity [Lancet 2021].
  • Caffeine at 300 mg/day or more [Ichikawa 2024].
  • PCOS/insulin resistance, poorly controlled diabetes, and chronic endometritis.

Inherited thrombophilias such as factor V Leiden are often tested, but ESHRE advises against routine screening unless other thrombosis risk factors are present.

Keep in mind that the list mixes true causes (APS, translocations, septum) with risk markers (age, BMI, vitamin D) that raise the odds without being a single explanation.

Sources:

Reduce chance of first miscarriage

Claude AI - Oct 2026

I'll pull effect sizes for sporadic (first) miscarriage, which draws on a somewhat different evidence base than RPL.

About 15% of recognized pregnancies end in miscarriage [Lancet 2021]. For a first loss, the picture is different from RPL. Most first losses come from a chance chromosomal error in the embryo, and most other risk factors raise the odds only modestly (around 1.2–1.4×). I ordered the list roughly by effect size. Items 5–10 are close enough that their order among themselves doesn't mean much.

  1. Chromosomal errors in the embryo. Aneuploidy is found in up to 90% of losses at 0–6 weeks and about half at 8–11 weeks [Genetics review]. This is mostly bad luck rather than a parental condition.
  2. Maternal age. Risk is lowest at ages 25–29 (about 10%), rises quickly after 30, and reaches 53% at 45 and older [Magnus BMJ 2019]. The Lancet series also flags mothers under 20 [Lancet 2021].
  3. Thyroid autoantibodies, even with normal thyroid function. Cohort studies showed more than triple the odds of miscarriage (OR 3.90) [Thangaratinam BMJ 2011]. Two small trials suggested levothyroxine cut miscarriages by about half [Journal Watch summary]. However, the larger TABLET trial (2019) found no benefit in euthyroid TPO-positive women, so the antibodies may be a marker more than a cause.
  4. Low vitamin D. Women under 50 nmol/L had nearly twice the odds of miscarriage of women above 75 nmol/L (OR 1.94). That result rests on just 4 studies with 3,674 women. [Tamblyn 2022]
  5. High caffeine intake. Compared with little or no caffeine, the risk was 1.40 at 350–699 mg/day and 1.72 at 700 mg/day or more. Each extra 100 mg/day (about one cup of coffee) added roughly 7%. [Chen 2016]
  6. Psychological stress. A meta-analysis of 8 studies found higher odds of miscarriage after stress exposure (OR 1.42) [Qu 2017]. Recall bias is a real concern here, since women who miscarry may report more stress after the fact.
  7. Smoking. Any active smoking raised risk (RR 1.23), more so when the smoking was during the pregnancy itself (RR 1.32). Risk rose about 1% per cigarette per day. [Pineles 2014] In a cotinine-verified study, secondhand smoke raised the odds of early miscarriage (OR 1.67) [George 2006].
  8. Obesity or underweight. Obese women miscarried at 13.6% versus 10.7% for normal-weight women (OR 1.31) [Boots & Stephenson 2011]. Very low BMI is also a risk factor [Lancet 2021].
  9. Paternal age. Fathers aged 40–44 had a pooled risk estimate of 1.23 versus men aged 25–29, adjusted for maternal age [du Fossé thesis]. See also the paper itself [du Fossé 2020].
  10. Alcohol. Any drinking during pregnancy raised the odds of miscarriage (OR 1.19). Up to 5 drinks a week, each additional weekly drink added about 6%. [Sundermann 2019]

Near misses:

  • Night work. Two or more night shifts in the previous week raised miscarriage risk after week 8 (HR 1.32) [Begtrup 2019].
  • Earlier pregnancy complications. A previous preterm birth (OR 1.22), stillbirth (1.30), C-section (1.16) or gestational diabetes (1.19) modestly raised the risk in the next pregnancy [Magnus BMJ 2019].
  • Other exposures. Black ethnicity, air pollution and pesticide exposure are also listed [Lancet 2021].
  • RPL causes that can also cause a first loss. Uncontrolled diabetes, uterine anomalies and antiphospholipid syndrome all apply, but they account for a small share of first losses.

Sources:

Related in VitaminDWiki

Created October 2026
URL: http://vitamindwiki.com/pages/miscarriage-vitamin-d-is-the-only-easily-changeable-major-risk-factor/
Printed October 16, 2026