Research report: Non-criteria antiphospholipid antibodies and pregnancy outcomes in recurrent pregnancy loss patients

Non-criteria antiphospholipid antibodies were detectable in 95 of 247 women with recurrent pregnancy loss, but their presence was not significantly associated with adverse outcomes in a subsequent pregnancy, according to a retrospective cohort study from a single hospital in Chengdu, China. Overall, adverse pregnancy outcomes occurred in 53.68% of antibody-positive patients and 46.71% of antibody-negative patients, a difference that was not statistically significant. The unadjusted odds ratio was 1.322, with a 95% confidence interval of 0.791–2.211 and a P value of 0.286, meaning the data did not establish that antibody positivity increased the risk of a poor outcome.

Recurrent pregnancy loss is defined in the study as the failure of two or more clinically recognized pregnancies before 20–24 weeks of gestation. The researchers focused on non-criteria antiphospholipid antibodies, a group that includes antibodies against phospholipids or associated proteins not used in conventional antiphospholipid syndrome testing. The panel covered 14 specific antibodies or isotypes: anti-phosphatidylethanolamine, anti-prothrombin, anti-annexin A5, anti-phosphatidylserine/prothrombin, anti-phosphatidylserine, anti-phosphatidic acid and anti-phosphatidylinositol, each tested for immunoglobulin G and immunoglobulin M where applicable.

Conventional tests included lupus anticoagulant, anticardiolipin antibodies and anti-β2-glycoprotein I antibodies. Antibody-positive women were more likely to have received intensive treatment than those who were antibody-negative. Low-molecular-weight heparin was used in 86.31% of the positive group compared with 74.34% of the negative group, glucocorticoids in 26.32% versus 14.47%, and hydroxychloroquine in 82.11% versus 62.50%.

The positive group was also more likely to have experienced three or more previous miscarriages: 38.95% compared with 23.68% in the negative group. These differences complicate interpretation because the study was observational, treatments were selected according to clinical risk, and medication could alter subsequent pregnancy outcomes. Most antibody results were negative, although positivity for at least one non-criteria antibody was common.

Anti-phosphatidylethanolamine immunoglobulin M was the most frequent finding, present in 35 women, or 14.17% of the cohort. Anti-prothrombin immunoglobulin M followed in 29 women, or 11.74%, while anti-annexin A5 immunoglobulin M occurred in 21, or 8.50%. Anti-annexin A5 immunoglobulin G was detected in 15 women, or 6.07%, and anti-phosphatidylserine/prothrombin immunoglobulin M in 13, or 5.26%.

No participants tested positive for anti-phosphatidylserine/prothrombin immunoglobulin G or anti-phosphatidic acid immunoglobulin M. Among the 95 positive patients, 63 had one positive antibody, while 32 had more than one. The study’s adverse-pregnancy-outcome composite included pregnancy loss, stillbirth, preterm delivery, preeclampsia, fetal growth restriction and intrauterine distress. Pregnancy loss itself was nearly as frequent in both groups, occurring in 42.11% of antibody-positive patients and 40.79% of antibody-negative patients, with an odds ratio of 1.056, a 95% confidence interval of 0.628–1.776 and a P value of 0.838.

Other individual outcomes also did not differ significantly between the groups. For example, preterm delivery occurred in 6.32% of positive patients and 1.32% of negative patients, while preeclampsia occurred in 4.21% and 1.97%, respectively; neither comparison met the threshold for statistical significance. Analyses of individual antibody subtypes and of single versus multiple positive results likewise found no significant association with the composite outcome.

One measured difference concerned newborn size. The median birth weight was 3,000 grams among babies born to antibody-positive patients and 3,170 grams among those born to antibody-negative patients, a statistically significant difference in the analysis. The difference remained apparent after birth weight was standardized for gestational age and sex using the INTERGROWTH-21st newborn growth standards: the median z-score was −0.39 in the positive group and 0.11 in the negative group.

The reported effect size was small, however, and subgroup analyses did not establish the same pattern consistently for individual antibody types or for the distinction between single and multiple positivity. Multivariable binary logistic regression produced its clearest association with treatment rather than antibody status. After adjustment for age, body mass index, prior abortion, number of previous pregnancy losses, pregnancy history and use of in vitro fertilization with embryo transfer, positivity for a single non-criteria antibody was associated with an adjusted odds ratio of 1.553 for adverse outcomes, but its 95% confidence interval was 0.799–3.017 and its P value was 0.194.

Multiple positivity had an adjusted odds ratio of 1.942, a 95% confidence interval of 0.798–4.726 and a P value of 0.144. Neither result was statistically significant, leaving the possibility that a modest association was missed because the study had limited statistical power. Medication intensity was associated with lower odds of an adverse outcome in the adjusted analysis.

Compared with no medication, receipt of one or two types of medicine was associated with an adjusted odds ratio of 0.177, a 95% confidence interval of 0.055–0.569 and a P value of 0.004. Patients receiving three or four types had an adjusted odds ratio of 0.102, a 95% confidence interval of 0.037–0.283 and a P value below 0.001. Stratified analyses found a similar association between escalated treatment and lower odds of the composite outcome in both antibody-positive and antibody-negative groups.

These are observational associations and do not demonstrate that the combination of medicines caused the improvement, nor do they establish the benefit of any particular drug or regimen. Several design limitations constrain what can be concluded. The cohort came from one center, was retrospective and included 247 women, and the investigators reported that statistical power was below the recommended 0.80 threshold.

Antibody testing was performed only once, on samples collected before conception or before seven weeks of gestation, without the 12-week confirmatory testing used in conventional antiphospholipid antibody evaluation. This raises the possibility of transient or false-positive results. The line immunoassay also offers multiplex testing but has less standardization than some other methods, and the absence of fetal chromosomal testing left a potential source of pregnancy loss unaccounted for.

In addition, a significant Hosmer–Lemeshow test indicated possible model-fit problems, and no healthy comparison group was available. Overall, the findings show that non-criteria antiphospholipid antibodies can be found in a substantial minority of women with recurrent pregnancy loss, but the study did not identify an independent association between their presence and adverse subsequent pregnancy outcomes. The result does not rule out such an association: treatment differences may have masked an underlying effect, single-time testing may have misclassified some patients, and the sample was too small to exclude modest associations.

The authors therefore call for larger prospective multicenter cohorts with repeated standardized antibody testing, fetal chromosomal analysis, healthy comparators and consistent treatment protocols. Until such evidence is available, the findings do not establish these antibodies as independent predictors of pregnancy complications.

Subject of Research: Association of non-criteria antiphospholipid antibodies with subsequent pregnancy outcomes in recurrent pregnancy loss

Article Title: Non-criteria antiphospholipid antibodies and pregnancy outcomes in recurrent pregnancy loss patients

Article References: Zhou, X., Luo, E., Cai, X., Chen, W., Xiao, L., Zhou, C., Wang, T., Xiong, M., Lai, F., & Deng, Y. (2026). Non-criteria antiphospholipid antibodies and pregnancy outcomes in recurrent pregnancy loss patients. Frontiers in Medicine, 13, Article 1961459. https://doi.org/10.3389/fmed.2026.1961459

Image Credits: AI Generated

DOI: 10.3389/fmed.2026.1961459

Keywords: recurrent pregnancy loss, antiphospholipid antibodies, adverse pregnancy outcomes, miscarriage, pregnancy loss, retrospective cohort, antibody testing, birth weight, Non-criteria, antiphospholipid, antibodies, pregnancy