The study team, led by researchers at the Reproductive Medicine Center of Sun Yat-Sen Memorial Hospital in Guangzhou, together with colleagues at Shenzhen Baoan Women’s and Children’s Hospital and several other institutions, analyzed data from 1,042 women who had experienced at least one miscarriage. Unlike most previous work, which concentrated almost exclusively on fasting measurements of insulin resistance, the investigators deliberately examined two distinct metabolic states: fasting insulin resistance, captured by the homeostasis model assessment of insulin resistance known as HOMA-IR, and postprandial insulin resistance, the dynamic metabolic condition that emerges after eating and that was assessed using the Stumvoll Insulin Sensitivity Index, or ISI-Stumvoll.

The technical distinction matters. HOMA-IR is calculated from fasting glucose and fasting insulin levels and reflects how efficiently the body’s tissues respond to insulin in the basal, overnight-fasted state. The Stumvoll index, by contrast, is derived from values obtained during an oral glucose tolerance test and estimates insulin sensitivity under the post-meal conditions that dominate everyday physiology. Postprandial insulin resistance has been flagged in the metabolic literature as an early marker of dysmetabolism, sometimes appearing before fasting measures become abnormal, yet its relationship to pregnancy outcomes had remained largely unexplored until now.

Using multivariable logistic regression models that adjusted for potential confounding factors, the researchers found that higher HOMA-IR values were associated with a significantly increased risk of miscarriage, with an odds ratio of 1.97 and a 95 percent confidence interval spanning 1.33 to 2.94, a result that reached statistical significance at P less than 0.001. In other words, women whose fasting insulin resistance was elevated faced nearly double the odds of losing the pregnancy compared with those whose fasting metabolism appeared healthier. Conversely, higher scores on the Stumvoll Insulin Sensitivity Index, which indicate greater insulin sensitivity, were associated with a reduced risk of miscarriage, with an odds ratio of 0.66 and a confidence interval of 0.45 to 0.98, significant at P equal to 0.038.

To probe the shape of these relationships more finely, the team employed restricted cubic spline analysis, a statistical technique that allows the association between a continuous exposure and an outcome to take on nonlinear forms rather than forcing a straight line. The results revealed a positive monotonic relationship between both HOMA-IR and the Stumvoll index and miscarriage risk, meaning risk rose steadily across the range of values without evidence of a threshold effect or a U-shaped curve. Formal tests for nonlinearity returned P values of 0.130 for HOMA-IR and 0.940 for the Stumvoll index, indicating no statistically significant departure from a linear trend for either marker.

Perhaps the most clinically consequential finding emerged when the researchers combined the two indicators. Women classified as insulin resistant by either HOMA-IR or the Stumvoll index—meaning they showed abnormal metabolism in the fasting state, the postprandial state, or both—had a significantly higher risk of miscarriage than women who were insulin sensitive on both measures. The combined classification yielded an odds ratio of 1.88 with a 95 percent confidence interval of 1.43 to 2.49 and a P value below 0.001. According to the authors, pairing the fasting and postprandial indicators was able to identify more individuals at elevated risk than either marker alone, a point with clear implications for how clinicians might screen women with recurrent or prior pregnancy loss.

The biological plausibility of a metabolic link to miscarriage is supported by a substantial body of prior research. Insulin resistance has been shown to exert direct toxic effects on human placental tissue, an effect that experimental work suggests can be mitigated by metformin. Animal studies have demonstrated that maternal insulin resistance causes oxidative stress and mitochondrial dysfunction in oocytes, potentially compromising egg quality from the very start of a pregnancy. Hyperinsulinemia, the compensatory surge of insulin that accompanies resistance, has been shown in mouse models to impair endometrial receptivity in early pregnancy, undermining the delicate dialogue between embryo and uterine lining. In women with unexplained recurrent pregnancy loss, insulin resistance has been associated with impaired decidualization—the transformation of the uterine lining that is essential for embryo implantation—along with altered expression of signaling molecules within the decidua itself.

Earlier clinical studies had already hinted at the connection. Research on patients undergoing assisted reproduction technology found that insulin resistance increased the risk of spontaneous abortion after treatment, and subsequent work in women with polycystic ovary syndrome undergoing in vitro fertilization linked insulin resistance to poorer ovarian sensitivity and adverse pregnancy outcomes. Systematic reviews and meta-analyses have identified high body mass index and insulin resistance as risk factors for spontaneous abortion in polycystic ovary syndrome patients receiving assisted reproductive treatment. What distinguished these earlier investigations, and what limited their reach, was their focus on infertility populations. The new study extends the association to women with miscarriage history in a broader clinical context, addressing what the authors describe as a critical gap in knowledge about insulin resistance and reproductive health outside fertility clinics.

The retrospective cohort design carries inherent limitations that the authors and outside observers will need to weigh. Because the data were collected from medical records at Sun Yat-Sen Memorial Hospital and Shenzhen Baoan Women’s and Children’s Hospital rather than through prospective follow-up, the analysis can establish association but not causation, and residual confounding by unmeasured factors—such as detailed dietary patterns, physical activity, or genetic predisposition—cannot be excluded. The study protocols were approved by the Ethics Committee for Clinical Research of Sun Yat-Sen Memorial Hospital, and the authors note that informed consent could not be obtained given the retrospective nature of the work. The underlying data are available from the corresponding author upon reasonable request, and the research was supported by the Guangdong Natural Science Foundation and the Yixian Clinical Research 5010 Project of Sun Yat-Sen Memorial Hospital.

Even with those caveats, the implications are difficult to ignore. Miscarriage is the most common complication of pregnancy, imposing substantial epidemiological, physical, psychological, and economic costs, and it carries significant consequences for mental health, including depression and anxiety following early pregnancy loss. If simple, widely available metabolic tests—fasting glucose and insulin for HOMA-IR, standard oral glucose tolerance values for the Stumvoll index—can flag women at heightened risk before conception, clinicians gain a potential window for intervention through lifestyle modification or insulin-sensitizing therapy. The authors conclude that both the fasting indicator HOMA-IR and the postprandial indicator ISI-Stumvoll were associated with miscarriage in women with a history of pregnancy loss, and that combining the two identifies more high-risk individuals. For the millions of women who endure the heartbreak of recurrent pregnancy loss, a routine blood test that reveals a modifiable risk factor would represent a meaningful step toward turning metabolic insight into reproductive hope.

Subject of Research: The association between fasting and postprandial insulin resistance patterns and miscarriage risk in women with a history of pregnancy loss.

Article Title: Association of Different Patterns of Insulin Resistance With Miscarriage in Population With Miscarriage History: a Multicenter Retrospective Cohort Study

Article References: Tang, J., Yu, Y., Zhang, J., Gan, Y., Yuan, S., Wang, Z., Ji, X., Mo, H., Zhang, Q., Shi, Y., & Chen, H. (2026). Association of Different Patterns of Insulin Resistance With Miscarriage in Population With Miscarriage History: a Multicenter Retrospective Cohort Study. Reproductive Sciences. https://doi.org/10.1007/s43032-026-02193-6

Image Credits: AI Generated

DOI: 10.1007/s43032-026-02193-6

Keywords: insulin resistance, miscarriage, HOMA-IR, Stumvoll Insulin Sensitivity Index, postprandial insulin resistance, recurrent pregnancy loss, reproductive medicine, glucose metabolism, endometrial receptivity, retrospective cohort study, Reproductive Sciences, polycystic ovary syndrome