Abstract

Background

The cardiometabolic index (CMI), a novel measure of obesity that integrates lipid profiles and indicators of abdominal adiposity, has emerged as a promising marker of metabolic health. However, its relationship with infertility remains largely unexplored. Using data from the National Health and Nutrition Examination Survey (NHANES) conducted between 2013 and 2018, this study investigated the potential association between CMI and infertility risk.

Methods

We utilized weighted multivariate logistic regression to examine the association between CMI and infertility and employed a restricted cubic spline model to explore potential non-linear relationships. Interaction tests and subgroups analyses were conducted to assess heterogeneity across different subgroups.

Results

The analysis included a nationally representative cohort of 1,142 women. After controlling for potential confounders, a positive association was identified between CMI and infertility risk (OR: 1.47; 95% CI: 1.05鈥�2.06; p鈥�=鈥�0.028). The restricted cubic spline model revealed a non-linear relationship (p鈥�=鈥�0.0109), with an inflection point at a CMI value of 0.341.

Conclusions

Our findings provides evidence of a positive association between CMI and infertility risk among U.S. adults. These results suggest that CMI could serve as a simple and effective surrogate marker for infertility risk assessment, offering valuable insights for reproductive health management.

Background

Infertility, defined as the inability to achieve pregnancy after one year of regular, unprotected sexual intercourse, affects approximately 15% of couples in their reproductive years worldwide [1], posing a significant public health challenge. According to the World Health Organization (WHO), the prevalence of infertility is increasing annually, with female infertility emerging as a critical concern that demands global attention and has spurred extensive research efforts [2]. The etiology of female infertility is complex and multifactorial, encompassing lifestyle factors, chromosome abnormalities, genetic mutations, ovulatory dysfunctions, endometriosis, tubal pathologiess, and idiopathic causes [3].

Recent studies have highlighted the pivotal role of lipid metabolism in infertility. Altered lipid profiles have been observed in individuals with reduced fecundity, characterized by elevated levels of total cholesterol, triglycerides (TG), and low-density lipoprotein cholesterol (LDL-C) levels, alongside decreased levels of high-density lipoprotein cholesterol (HDL-C). Women with abnormal lipoprotein levels were found to have a 19鈥�36% reduction in the likelihood of conception per cycle [4]. Both univariable and multivariable Mendelian randomization analyses have established a causal link between elevated LDL-C levels and an increased risk of female infertility [5]. Additionally, analysis of endometrial receptivity array (ERA) data from the Gene Expression Omnibus (GEO) database has revealed correlations between genes involved in lipid metabolism and endometrial receptivity in women with reproductive dysfunction [6].

Therefore, this study utilized data from the National Health and Nutrition Examination Survey (NHANES) to examine the relationship between CMI and infertility onset, aiming to identify clinical cutoff values for predicting infertility risk.

The Cardiometabolic Index (CMI) is a novel metric that integrates parameters related to obesity with quantitative assessments of blood lipid levels [7]. It has demonstrated potential as a prognostic tool for various health conditions, including depression, obstructive sleep apnea, microalbuminuria, stroke, and hypertension [8,9,10,11]. Given its diagnostic utility across multiple disorders and the established link between lipid metabolism and infertility, further investigation into the role of CMI in infertility diagnosis is warranted. Specifically, it is essential to determine clinical cutoff values for CMI to predict infertility risk. Currently, there is a lack of effective cutoffs for this purpose. Therefore, this study utilized data from the National Health and Nutrition Examination Survey (NHANES) to examine the relationship between infertility onset and CMI, aiming to identify clinical cutoff values for predicting infertility risk.

Materials and methods

Study population

This study analyzed data from the NHANES conducted between 2013 and 2018. The survey employed a national cross-sectional design with a multistage sampling methodology to assess the health status and lifestyle changes of the U.S. population. The selected cycles were specifically chosen to ensure consistency in the measurement of relevant variables. From an initial cohort of 29,400 participants, exclusions were made based on the following criteria: (1) male participants and women aged over 45 or under 20 (n鈥�= 25,545); (2) absence of CMI and fertility data (n鈥�= 2,498); and (3) incomplete data on covariates (n鈥�= 215), including race, age, marital status, education, poverty income ratio (PIR), diabetes, cardiovascular disease (CVD), hypertension, smoking, alcohol consumption, and reproductive health variables. After applying these criteria, the final sample consisted of 1,142 participants (Fig.听1).

Fig. 1

Flowchart of the sample selection from NHANES 2013鈥�2018

Exposure and outcome variables

The CMI, utilized as the exposure variable in this study, was calculated using a combination of anthropometric and biochemical parameters, specifically TG, waist circumference (WC), standing height (HT), and HDL-C. The formula for CMI is as follows: CMI =鈥塗G/HDL-C * [WC/HT], where TG and HDL-C are expressed in mmol/L, and WC and HT are measured in centimeters. Participants were categorized into three groups based on weighted quantiles [12]: Group Q1 (CMI 鈮も��0.23), Group Q2 (0.23 <鈥塁MI 鈮も��0.47), and Group Q3 (CMI >鈥�0.47). The weighted quantile method was chosen to ensure that the CMI groups accurately reflect the relative importance of each data point within the population.

The outcome variable, infertility, was determined using self-reported responses from the Reproductive Health Questionnaire (variable name: RHQ074) [13]. Participants were asked whether they had"attempted to conceive for a full year."Those who responded"yes"were classfied as infertile, while those who answered"no"were classified as fertile.

Covariates

This study included a comprehensive set of potential confounding variables to clarify the relationship between CMI and infertility. The covariates comprised demographic factors (race, age, marital status, education, and PIR), health conditions (diabetes, CVD, hypertension), lifestyle factors (smoking, alcohol consumption), and reproductive health variables. Racce was categorized as non-Hispanic White, non-Hispanic Black, other Hispanic, Mexican American, and others. Educational attainment was divided into three levels: below high school, high school completion, and post-secondary education. Diabetes status was determined based on self-reported medical diagnoses and categorized as either present or absent. Smoking habits were classified into three groups: current smokers (鈮� 100 cigarettes smoked in their lifetime and actively smoking), former smokers (鈮� 100 cigarettes smoked but no longer smoking), and never smokers (< 100 cigarettes smoked in their lifetime). Alcohol consumption was dichotomized based on whether participants consumed 12 or more alcoholic beverages annually. CVD was defined using self-reported data on five conditions: coronary heart disease, congestive heart failure, heart attack, stroke, and angina pectoris (variable names: MCQ160B, MCQ160 C, MCQ160D, MCQ160E, MCQ160 F) [14]. Hypertension was determined based on responses the question,"Have you ever been told you had high blood pressure?"(variable name: BPQ020), and categorized as"yes"or"no"[15]. Reproductive health variables included pregnancy history (yes/no), age at menarche (categorized as <鈥�10, 10鈥�15, and 15鈥�20 years), menstrual regularity in the past year (yes/no), history of pelvic inflammatory disease (yes/no), and oral contraceptive usage (yes/no).

Additionally, hematological parameters, including neutrophil, platelet, and lymphocyte counts (measured in 1000 cells/渭L), were analyzed. These parameters were employed to compute the systemic immune-inflammation index (SII), defined as the product of the platelet count and the neutrophil-to-lymphocyte ratio, as described in previous studies [16].

Statistical analysis

The statistical analysis framework adhered to the recommendations of the Centers for Disease Control and Prevention (CDC), and all computations were carried out through R software (version 4.2.3). The normality of continuous variables was assessed with the Shapiro鈥揥ilk test. For normally distributed data, t-tests were performed, and results were reported as mean 卤鈥塻tandard deviation. For non-normally distributed data, the Mann鈥揥hitney U test was applied, with outcomes presented as weighted medians and interquartile ranges (IQR). Categorical data were analyzed using the Chi-square test and summarized as frequencies and percentages. To evaluate multicollinearity among covariates, variance inflation factors (VIFs) were calculated. All covariates exhibited VIFs below 5, indicating no significant multicollinearity.

The association between infertility and CMI, treated as both a continuous and categorical variable, was examined using weighted logistic regression models. Interaction tests and subgroup analyses were conducted to ensure the robustness of the relationship across various strata, including education, hypertension, alcohol consumption, smoking status, pregnant history, and oral contraceptive use. Three progressive models were developed: a crude model with no adjustments, Model 1 adjusting for demographic variables (race, age, PIR, and education), and Model 2 further incorporating lifestyle factors, comorbidities, and reproductive health variables. When CMI was treated as a categorical variable, the P-value for the trend was calculated using the Wald test. To investigate potential non-linear relationships, restricted cubic spline (RCS) regression was employed. Statistical significance was defined as a P-value <鈥�0.05.

Results

Baseline characteristics

Table 1 summarizes the baseline characteristics of the 1,142 participants included in the analysis. Among these, 143 (12.5%) reported infertility, while 999 (87.5%) served as controls. The median age of the participants was 32 years (IQR: 25鈥�38 years). The racial composition of the cohort included 13% Non-Hispanic Black, 58% Non-Hispanic White, 7.3% Other Hispanic, 11% Mexican American, and 10% other races. Significant differences were observed between infertile and fertile participants in terms of age, age at menarche, prevalence of diabetes and hypertension, pregnancy history, menstrual cycle regularity, WT, and HDL-C levels. However, no significant differences were found in race, PIR, education, smoking status, alcohol consumption, CVD, pelvic inflammatory disease, oral contraceptive use, neutrophil count, platelet count, lymphocyte count, SII, HT, and TG levels. Notably, participants in the infertile group exhibited a significantly higher CMI (median: 0.42, IQR: 0.27鈥�0.68) compared to the control group (median: 0.33, IQR: 0.19鈥�0.58) (P鈥�< 0.001).

Table 1 Baseline characteristics of study participants

Association between CMI and infertility

As presented in Table听2, weighted logistic regression analysis revealed a significant positive association between CMI and infertility risk. This relationship held true whether CMI was analyzed as a continuous or categorical variable and remained robust across all statistical models. When CMI was treated as a continuous variable, each unit increase in CMI was associated with a significantly elevated likelihood of infertility. Specifically, the odds ratios (ORs) for infertility increased by 52%, 44%, and 47% in the crude model, Model 1, and Model 2, respectively. These findings suggest a dose-dependent relationship, where even modest increases in CMI may adversely impact reproductive outcomes. When CMI was categorized into tertiles, a statistically significant trend (P鈥�< 0.05) was observed across all models, with higher CMI tertiles corresponding to greater infertility risk. For instance, in Model 1, individuals in the highest CMI tertile exhibited a notably highly likelihood of infertility compared to those in the middle (OR: 2.54) and lowest (OR: 2.05) tertiles. Similarly, Model 2 demonstrated consistent results, with ORs of 2.76, 2.22, and 1.00 for the highest, middle, and lowest tertiles, respectively.

Table 2 Association between CMI and Odds rations (95% confidence intervals) of infertility, NHANES 2013鈥�2018

Non-linear relationship between CMI and infertility

To explore potential non-linear relationships between CMI and infertility, RCS regression was performed. After adjusting for relevant covariates, a significant non-linear association was identified (p鈥�= 0.0109) (Fig.听2). The analysis illustrated an upward trend in the odds of infertility with increasing CMI values and identified a critical infection point at 0.341. Participants with CMI >鈥�0.341 exhibited significantly higher odds of infertility compared to those with CMI <鈥�0.341. In the crude model, the OR for infertility in the higher CMI group was 1.87 (95% CI: 1.18鈥�2.97). In Model 1, the OR was 1.83 (95% CI: 1.12鈥�2.97), and in Model 2, the OR increased to 2.02 (95% CI: 1.18鈥�3.48) (Table听3).

Fig. 2

The RCS plot describes the association between CMI and the incidence of infertility. It was adjusted for all covariates in Table听2. OR odds ratio

Table 3 Correlation analysis between CMI and infertility grouped by cut-off value

Subgroup analyses

To further explore the relationship between CMI and infertility, subgroup analyses and interaction tests were performed, accounting for variables such as education level, smoking status, alcohol consumption, hypertension, pregnancy history, and oral contraceptive use. The subgroup analysis revealed varying associations between CMI levels and infertility across different groups (Fig. 3). A strong association was observed among alcohol consumers (P鈥�< 0.001), current smokers (P鈥�= 0.005), and individuals without hypertension (P鈥�= 0.001). Interaction analyses revealed that education level, smoking habits, CVD, pregnancy history, pelvic infections, menstrual cycle regularity, and oral contraceptive use did not significantly modify the association between CMI and infertility (P for interaction >鈥�0.05). Nevertheless, alcohol consumption and hypertension showed significant interactions with the CMI-infertility relationship (P for interaction <鈥�0.05) (Fig.听3).

Fig. 3

The association between CMI and infertility by different subgroups

Discussion

This study provides novel insights into the relationship between CMI and infertility using a representative sample of the U.S. adult population. Our analysis, which included 1,142 individuals, identified a positive, non-linear association between CMI and infertility, even after adjusting for multiple confounders. A significant inflection point was observed at a CMI value of 0.341, suggesting a potential threshold effect that warrants further investigation. When CMI was categorized into tertiles, logistic regression analysis reinforced this association, with higher CMI tertiles corresponding to increased odds of infertility. This relationship remained consistent across most demographic and health-related subgroups, except for smoking status and alcohol consumption, which appeared to influence the relationship.

The rising prevalence of obesity and overweight, especially among women of reproductive age, has become a pressing global health concern. Numerous studies have documented an inverse relationship between preconception maternal weight and live birth rates [17]. The underlying mechanisms are complex and multifactorial, involving disruptions to the neuroendocrine systems and the hypothalamic-pituitary-ovarian (HPO) axis. Adipose tissue, once considered inert, is now recognized as a dynamic endocrine organ with extensive physiological effects, influencing processes such as glucose homeostasis, immunoregulation, steroid production, reproduction, and hematopoiesis [18]. Adipokines such as leptin, adiponectin, and resistin interact with multiple molecular pathways implicated in insulin sensitivity, steroidogenesis, inflammation, and oocyte maturation. These interactions ultimately affect the HPO axis and ovulation [19]. Women with central obesity often exhibit insulin resistance and hyperinsulinemia, which negatively impact ovarian function [20]. Moreover, chronic inflammation and oxidative stress, both closely linked to obesity, have been shown to disrupt key reproductive processes, including estrous cyclicity, steroid hormone production, and ovulation. These conditions also impair the meiotic and cytoplasmic maturation of oocytes, reducing their developmental potential and the likelihood of successful fertilization and early embryonic development [21]. Furthermore, impaired stromal decidualization, a crucial process in placental development, has been observed in obese individuals and may contribute to an elevated risk of pregnancy complications like miscarriage, stillbirth, and preeclampsia [22]. Encouragingly, weight loss through lifestyle interventions has been shown to restore menstrual regularity, improving ovulatory function, and enhance the likelihood of conception [23].

The CMI has recently emerged as a novel metric for assessing obesity, offering insights into both visceral fat distribution and dysfunction [7]. As a composite measure incorporating waist-to-height ratio (WHtR) and TG/HDL-C, CMI offers a comprehensive evaluation of an individual's metabolic profile. The WHtR component reflects subcutaneous and visceral fat accumulation, with abdominal obesity increasingly recognized as a contributor to infertility [24]. Concurrently, the TG/HDL-C ratio serves as a surrogate marker for insulin resistance and metabolic syndrome, both of which have been linked to fertility challenges [25]. Our findings align with previous fstudies, demonstrating a significant independent relationship between elevated CMI and a heightened risk of infertility. This underscores the potential utility of CMI as a predictor of reproductive health outcomes. By identifying individuals at higher risk, CMI could serve as a valuable tool for guiding health promotion and prevention strategies aimed at improving reproductive outcomes.

Our study has strengthes, such as a large sample size and the inclusion of a cohort representive of the national population. Besides, rigorous statistical methods were employed to adjust for potential confounding variables, enhancing the reliability of our results. However, certain limitations should be considered. First, since this study is cross-sectional and does not include measurements at multiple time points, establishing a causal relationship is challenging. Longitudinal studies with larger cohorts are warranted to better elucidate the causal mechanisms linking infertility and CMI. Second, mediating factors such as hormonal profiles and lifestyle variables like diet and physical activity are critical for understanding the underlying mechanisms of the observed associations. Unfortunately, the NHANES dataset lacks information on hormonal profiles and provides only limited data on dietary and physical activity patterns. Future research should incorporate these variables to enable a more comprehensive analysis. Third, concerns about the small sample size in certain subgroups should be acknowledged. To address this issue, we calculated and reported the sample size for each subgroup in the forest plot. While the statistical power of some subgroups may be limited, the insights gained from these analyses contribute to our understanding of potential effect modifications, and larger studies are needed to validate these findings. Fourth, whether women鈥檚 infertility status is affected by male infertility is a matter of concern. However, since this study is based on the NHANES database, which does not provide specific information, we cannot further distinguish the influencing factor. Future research should focus on data on male reproductive health to explore the relationship more deeply. Furthermore, although we controlled for several confounding factors, the possibility of residual confounding cannot be entirely excluded. Lastly, as the participants were adults residing in the United States, further research is required to determine whether these findings are generalizable to other populations.

Conclusion

This study provides compelling evidence of an independent association between elevated CMI levels and an increased risk of infertility. These findings highlight the potential utility of CMI as a valuable tool for assessing fertility risk and emphasize the importance of addressing cardiometabolic health in reproductive medicine. To validate and expand on these findings, large-scale prospective studies involving diverse populations are warranted. Such research could elucidate the underlying mechanisms driving the CMI-infertility relationship and inform targeted interventions to optimize reproductive outcomes.