Adverse effects of dyslipidemia on embryo quality and reproductive outcomes in specific subgroups undergoing first IVF/ICSI cycles: A case-control study.

Female infertility, clinically defined as the inability of women to conceive after >1 year of regular sexual intercourse in the absence of contraception, [ 1 ] represents a profound global health burden affecting millions of couples. Assisted reproductive technology (ART), particularly in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI), has revolutionized the treatment landscape. However, despite continuous advancements in laboratory techniques and clinical protocols, the clinical pregnancy rate (CPR) per ART cycle remain suboptimal and exhibit significant interindividual variability. [ 2 ] This underscores a critical need to elucidate modifiable factors influencing oocyte competence, embryonic development, and ultimately, treatment success. Relevant evidence suggests that abnormal lipid metabolic parameters affect the reproductive performance of ART. Dyslipidemia (characterized by aberrations in serum lipid profiles including elevated total cholesterol [TC], triglycerides [TG], low-density lipoprotein cholesterol [LDL-C], and/or reduced high-density lipoprotein cholesterol [HDL-C] [ 3 ] ) is highly prevalent, affecting approximately 40.4% of the adult population in China alone, [ 3 ] and presents a major modifiable risk factor. The pathophysiological links between dyslipidemia and impaired reproductive outcomes are mechanistically plausible. Within the ovarian follicle, lipid-rich follicular fluid (i.e., excessive lipid accumulation) (a consequence of systemic dyslipidemia or localized dysregulation) can trigger a cascade of detrimental effects, including oxidative stress in the endoplasmic reticulum, and mitochondrial dysfunction. [ 4 ] These alterations can compromise oocyte maturation and quality, ultimately manifesting as impaired embryo development, reduced blastocyst formation, and lower implantation potential. [ 5 – 7 ] Furthermore, the detrimental impact of dyslipidemia established during pregnancy on complications like gestational diabetes mellitus, preeclampsia, intrahepatic cholestasis of pregnancy, and preterm birth is well-documented. [ 8 – 11 ] In contrast, the influence of preconception dyslipidemia on ART outcomes remains inadequately characterized and marked by contradictory findings. [ 7 , 12 , 13 ] A recent retrospective study of 1132 patients showed that dyslipidemia was associated with a lower cumulative live birth rate (CLBR) (odd ratio [OR] [95% confidence interval [CI]]: 0.702[0.533–0.881], P  = .044) following IVF or ICSI. [ 12 ] Conversely, another study reported no significant difference in CLBR between infertile patients with dyslipidemia and those with normal blood lipid levels. [ 13 ] These inconsistencies highlight significant limitations in the current literature: heterogeneous study populations and insufficient exploration of potential effect modifiers such as age, body mass index (BMI), and the specific pattern or severity of lipid abnormalities. Furthermore, embryo quality parameters, a direct reflection of oocyte competence and a strong predictor of pregnancy success, have received less attention than pregnancy endpoints. Polycystic ovary syndrome (PCOS) is often accompanied by metabolic disorders, such as insulin resistance, obesity, dyslipidemia, and thyroid dysfunction, [ 14 , 15 ] can independently and substantially influence ovarian function, oocyte quality, and ART outcomes. [ 16 – 18 ] Isolating the specific effect of dyslipidemia apart from this multifaceted syndromic background is challenging. Hence, we conducted the current study that excluded women with PCOS to minimize metabolic confounding and better isolate the association between abnormal serum lipid profiles with embryo quality and pregnancy outcomes in patients undergoing IVF/ICSI for the first time. In addition, we performed a subgroup analysis considering variables such as age (<35 vs ≥35 years), BMI (<24, 24–27.9 or ≥28 kg/m²), and the type and degree of elevated lipid levels (TC <5.2 [normal], 5.2–6.2 [marginally increased] or ≥6.2 [increased] mmol/L; TG 1.0 [normal] mmol/L; LDL-C <3.4 [normal], 3.4–4.1 [marginally increased] or ≥4.1 mmol/L [increased]). This study aims to identify specific subpopulations of women for whom dyslipidemia may pose a significant risk to ART success, thereby informing personalized risk assessment and potential pretreatment optimization strategies.

Conceptualization: Jiahuan Luo, Jiang Liu, Meng Rao, Shuhua Zhao, Li Tang. Data curation: Liqin Zuo, Shunqing Wang. Methodology: Dandan Zhang, Liqin Zuo, Shunqing Wang. Software: Jiahuan Luo, Dandan Zhang, Jinyuan Wang. Supervision: Shuhua Zhao, Li Tang. Validation: Jiang Liu, Meng Rao. Visualization: Jiahuan Luo, Jinyuan Wang, Liqin Zuo, Shunqing Wang. Writing – original draft: Jiahuan Luo. Writing – review & editing: Li Tang.

This retrospective cohort study was conducted at the First Affiliated Hospital of Kunming Medical University between April 2016 and March 2023. The study was approved by the hospital’s ethics committee (approval no. [2023] Ethical Review No. 163) and conducted in accordance with the principles of the Declaration of Helsinki. Given the retrospective nature of this study, the need for individual patient consent was waived. Couples undergoing their first IVF/ICSI treatment with complete lipid profile data (including TC, TG, HDL, and LDL) throughout the study period were eligible for inclusion in the current study. The inclusion criteria were as follows: female partners aged between 20 and 40 years old; retrieval of at least 1 oocyte per cycle; male partners within the normal blood lipids. The exclusion criteria included: female or male partners with chromosomal abnormalities or malignant tumors; female diagnosed with PCOS; presence of endometriosis or uterine anomalies in the female partner; history of 2 or more pregnancy losses; indications for preimplantation genetic testing for aneuploidy; and female partners with other endocrine or metabolic diseases. Prior to ovulation induction, the serum levels of TG, TC, LDL-C, and HDL-C were assessed in the same laboratory. As per the 2016 Chinese guidelines for the management of dyslipidemia in adults, [ 7 ] dyslipidemia is diagnosed if any of the following abnormal levels are present: TC ≥6.2 mmol/L, LDL-C ≥4.1 mmol/L, HDL-C ≤1.0 mmol/L, or TG ≥2.3 mmol/L. The participants who did not meet these criteria were classified as having normal blood lipid levels. Ultimately, 327 dyslipidemia and 756 normal blood lipid women were included for oocyte retrieval and embryo quality analysis, with 299 dyslipidemia and 696 normal blood lipid women evaluated for pregnancy outcomes. The flow chart is shown in Figure 1 . The flow chart for the included population. ICSI = intracytoplasmic sperm injection; IVF = in vitro fertilization; PCOS = polycystic ovary syndrome; PGT-A = preimplantation genetic testing for aneuploidy. Serum lipid concentrations were measured using a Cobas c701 autoanalyzer (Roche, Mannheim, Germany). TC, TG, HDL, and LDL were analyzed using the cholesterol oxidase, GPO-PAP, PEG-modified enzyme, and surfactant clearance methods, respectively. The intra-batch CVs for TC, TG, HDL, and LDL were ≤0.5%, ≤1.03%, ≤0.62%, and ≤0.33%, respectively, while the inter-assay CVs were ≤0.71%, ≤1.11%, ≤0.91%, and ≤0.47%, respectively. The primary outcomes included high-quality cleavage embryo rate, cumulative clinical pregnancy rate (CCPR), and CLBR. [ 19 , 20 ] Secondary outcomes were oocyte retrieval rate, fertilization rate, cleavage rate, embryo utilization rate, blastocyst formation rate, high-quality blastocyst rate, CPR, early pregnancy loss rate, and live birth rate (LBR) [ 19 ] in the first transfer cycle. Various rates were calculated based on different parameters to assess different stages of the assisted reproductive process. In cases where multiple live births were obtained in a single oocyte retrieval cycle, only the first live births were included in the analysis. [ 21 ] A complete-case analysis approach was adopted, including only couples with a full lipid profile; no imputation was performed for missing data. Data were analyzed using IBM SPSS Statistics for Windows (IBM Corp., Armonk), version 26.0. Given that all continuous variables exhibited skewed distributions, the results were expressed as median (P25–P75) and compared using the Mann–Whitney U test. Categorical variables are presented as frequencies (percentages) and compared using the Chi-square test. All tests were two-tailed, and statistical significance was set at P  < .05. Generalized linear models with an identity link function for continuous outcomes and a logistic link function for binary outcomes were used to assess the impact of abnormal lipid levels on embryo quality and pregnancy outcomes. [ 19 , 22 ] In the analysis of the oocyte retrieval rate, adjustments were made for female age, BMI, anti-Mullerian hormone concentration, basal follicle stimulating hormone, antral follicle count, infertility type, infertility diagnosis, infertility duration, ovarian stimulation regimens, and year of oocyte retrieval. In the analysis of laboratory outcomes, CCPR, and CLBR, further adjustments were made for variables such as male age, male BMI, and fertilization method (IVF or ICSI). Additional adjustments were made to the model for the type and stage of embryo transfer, endometrial thickness, and number of embryos transferred when analyzing the CPR, early pregnancy loss rate, and LBR of the first transfer cycle. To investigate whether the effects of lipid profiles on embryo quality and pregnancy outcomes varied according to patients’ age, BMI, and type and degree of lipid abnormality, subgroup analyses were conducted based on female age (< 35 or ≥35 years), BMI (<24, 24–27.9 or ≥28 kg/m²), and lipid subtypes (TC <5.2 [normal], 5.2–6.2 [marginally increased] or ≥6.2 [increased] mmol/L; TG 1.0 [normal] mmol/L; LDL-C <3.4 [normal], 3.4–4.1 [marginally increased] or ≥4.1 mmol/L [increased]). For multiple comparisons, we used the Bonferroni correction method. Owing to the retrospective design and fixed sample size, no a priori sample size calculation was conducted. As a result, subgroup analyses with limited sample sizes may have reduced statistical power to detect existing differences, increasing the risk of Type II errors. Therefore, results from these underpowered subgroup analyses should be interpreted with caution and regarded as exploratory.

The descriptive characteristics of the 1083 participants who met the inclusion criteria are presented in Table 1 . Compared to the normal lipid group, couples in the dyslipidemia group were significantly older (both female and male, P  < .01), and females had a higher BMI ( P  < .001). Moreover, the dyslipidemia group also required a higher total Gn dose during ovarian stimulation ( P  = .005) and had a different distribution in the number of embryos transferred in the first cycle ( P  = .024). Other demographic and clinical characteristics were comparable between the 2 groups. Characteristics of women with and without dyslipidemia. Bold values indicate results where the intergroup differences are statistically significant ( P < .05). AFC = antral follicle count, AMH = anti-Mullerian hormone, bFSH = basal follicle stimulating hormone, bLH = basal luteinizing hormone, BMI = body mass index, HDL-C = high-density lipoprotein cholesterol, ICSI = intracytoplasmic sperm injection, IVF = in vitro fertilization, LDL-C = low-density lipoprotein cholesterol, PPOS = progesterone-primed ovarian stimulation, TC = total cholesterol, TG = triglycerides. In the overall population, the oocyte retrieval rate was lower in the dyslipidemia group compared to the normal group ( P   =  .026). Generalized linear model confirmed consistent results (adjusted odd ratio [aOR]: 0.972, P  = .025) after adjusting for female age, BMI, anti-Mullerian hormone concentration, basal follicle stimulating hormone, antral follicle count, primary or secondary infertility, infertility diagnosis, infertility duration, ovarian stimulation protocols, and year of oocyte retrieval (Table 2 ). No significant associations were observed between dyslipidemia and any other measures of embryo quality or pregnancy outcomes (Table 2 ). Embryo quality and pregnancy outcomes in women with and without dyslipidemia. In the analysis of the oocyte retrieval rate, adjustments are made for female age, BMI, AMH concentration, basal follicle stimulating hormone, antral follicle count, primary or secondary infertility, infertility diagnosis, infertility duration, ovarian stimulation protocols, and year of oocyte retrieval. In the analysis of laboratory outcomes, CCPR, and CLBR, adjustments are made for female age, male age, female BMI, male BMI, AMH concentration, primary or secondary infertility, infertility diagnosis, infertility duration, ovarian stimulation protocols, year of oocyte retrieval, and fertilization method. Additional adjustments are made to the model for the type and stage of embryo transfer, endometrial thickness, and number of embryos transferred when analyzing the CPR, ePLR, and LBR of the first transfer cycle. Bold values indicate results where the intergroup differences are statistically significant ( P < .05). AMH = anti-Müllerian hormone, BMI = body mass index, CCPR = cumulative clinical pregnancy rate, CI = confidence interval, CLBR = cumulative live birth rate, CPR = clinical pregnancy rate, ePLR = early pregnancy loss rate, LBR = live birth rate, OR = odds ratio. The rates of each group are pooled for calculation. The rates are calculated individually for each cycle. Univariate generalized linear models. Multivariate generalized linear models. In women aged <35 years, dyslipidemia was negatively correlated with the CCPR (aOR: 0.661, P   =  .027) and CLBR (aOR: 0.667, P   =  .021). Among women aged ≥35 years, dyslipidemia was marginally associated with a trend towards a lower oocyte retrieval rate (aOR: 0.954, P  = .043). No other significant associations were found in this age group (Table 3 ). Embryo quality and pregnancy outcomes in women aged <35 years and ≥35 years with and without dyslipidemia. Generalized linear models are used for the data analysis. In the analysis of the oocyte retrieval rate, adjustments are made for female age, BMI, AMH concentration, basal follicle stimulating hormone, antral follicle count, primary or secondary infertility, infertility diagnosis, infertility duration, ovarian stimulation protocols, and year of oocyte retrieval. In the analysis of laboratory outcomes, CCPR, and CLBR, adjustments are made for female age, male age, female BMI, male BMI, AMH concentration, primary or secondary infertility, infertility diagnosis, infertility duration, ovarian stimulation protocols, year of oocyte retrieval, and fertilization method. Additional adjustments are made to the model for the type and stage of embryo transfer, endometrial thickness, and number of embryos transferred when analyzing the CPR, ePLR, and LBR of the first transfer cycle. Bold values indicate results where the intergroup differences are statistically significant ( P < .05). AMH = anti-Müllerian hormone, aOR = adjusted odds ratio, BMI = body mass index, CCPR = cumulative clinical pregnancy rate, CI = confidence interval, CLBR = cumulative live birth rate, CPR = clinical pregnancy rate, ePLR = early pregnancy loss rate, LBR = live birth rate. In the subgroup of BMI ≥28 kg/m 2 , dyslipidemia was associated with a lower cleavage rate (aOR: 0.901, P   =  .036) (Table S1, Supplemental Digital Content, https://links.lww.com/MD/R143 ). The subgroups of BMI < 24 kg/m² and 24 to 27.9 kg/m² showed no correlation with embryo quality or pregnancy outcomes (Table S1, Supplemental Digital Content, https://links.lww.com/MD/R143 ). There was a significantly lower oocyte retrieval rate in the group with marginally increased TG levels compared with the normal TG group (aOR: 0.950, P  = .012) after adjusting for confounding factors (Table S2, Supplemental Digital Content, https://links.lww.com/MD/R143 ). However, TG levels did not significantly correlate with other measures of embryo quality or pregnancy outcomes. In the TC, HDL-C, and LDL-C subgroups, no significant differences were observed in terms of embryo quality and pregnancy outcomes between the normal lipid group and the marginally increased or increased lipid groups (all P  > .05), as detailed in Table S2, Supplemental Digital Content, https://links.lww.com/MD/R143 . The application of the Bonferroni correction for multiple comparisons did not alter the interpretation of these findings (Table S3, Supplemental Digital Content, https://links.lww.com/MD/R143 ).

This retrospective cohort study investigated the association between dyslipidemia and outcomes of ART. The inherent limitations of the observational design preclude causal inference and must be considered when interpreting the following results. Our primary analysis revealed that after comprehensive adjustment for confounders, dyslipidemia was not significantly associated with cumulative pregnancy or live birth rates in the overall population. However, our exploratory subgroup analyses suggested that dyslipidemia exerted differential impacts on ART outcomes depending on female age, BMI, and lipid subtypes. A large retrospective study of 5030 infertile women found that the average age of women with dyslipidemia was higher than that of the non-dyslipidemia group (30.60 ± 4.27 vs 30.03 ± 4.18, P  < .001), and BMI was higher than that of non-dyslipidemia patients (22.89 ± 3.02 vs 21.39 ± 2.73, P  < .001). [ 7 ] These findings are consistent with those of the present study, implying that dyslipidemia is more prevalent among older and obese individuals. This study also observed that infertile patients with dyslipidemia required a significantly higher total gonadotropin dose during controlled ovarian hyperstimulation than those with normal lipid metabolism, which has been corroborated by other studies. [ 12 , 13 ] This could be explained by the proposed impairment of the hypothalamic–pituitary–ovarian axis by dyslipidemia in women, resulting in decreased sensitivity of the ovarian response to gonadotropins. [ 23 , 24 ] Therefore, more gonadotropins are required to promote follicular growth. Moreover, age, BMI, and the dosage of Gn are considered major factors affecting oocyte retrieval, which may also explain the lower oocyte retrieval rate observed among dyslipidemic patients in contrast to those with normal lipid metabolism in our study. Liu et al’s study also identified this phenomenon in patients with infertility. [ 12 ] Therefore, further in-depth studies are required to elucidate this specific mechanism. The existing literature presents a conflicting picture regarding dyslipidemia and embryo quality. A retrospective study showed that TG was negatively correlated with the number of oocytes, normal fertilized oocytes, cleaved embryos, and high-quality embryos. [ 5 ] Additionally, TC was inversely correlated with the number of normal fertilized oocytes and high-quality embryos, LDL was negatively correlated with the number of fertilized oocytes, and HDL was positively correlated with the number of oocytes, normal fertilized oocytes, and cleaved embryos, [ 5 ] while Yang et al [ 7 ] observed minimal impact of dyslipidemia on embryo quality, consistent with our overall analysis. Our exploratory analysis within the BMI ≥28 kg/m² subgroup found that dyslipidemia was a potential negative association with cleavage rate (aOR: 0.901, P  = .036). A prospective cohort study showed that patients with a BMI >30 kg/m 2 had a longer duration of cell cleavage from the stage of polar body appearance to 4-cell division than those with a BMI <25 kg/m 2 . [ 25 ] We therefore hypothesize that in the context of obesity, underlying dyslipidemia may exacerbate metabolic stress, potentially disrupting ATP production, which is a vital energy source for early embryonic developmental processes, [ 26 ] thereby impairing the cleavage process. It is crucial to note that this remains a speculative mechanism, as our study lacked follicular fluid or molecular data to confirm this pathway. The most significant conflict in the literature concerns pregnancy outcomes. Liu et al [ 12 ] reported a significant negative association between dyslipidemia and LBR in their cohort, whereas Yang et al [ 7 ] found no such association (a finding consistent with our overall results). This discrepancy was likely attributable to key methodological distinctions, most notably the study population: Liu et al [ 12 ] specifically focused on patients undergoing fresh embryo transfer, whereas our study and Yang et al [ 7 ] included patients in their first embryo transfer cycle. This fundamental difference in cohort selection could significantly influence the observed outcomes. Regarding cumulative pregnancy outcomes, our observation that dyslipidemia was not associated with these outcomes in the overall study population further align with the 2022 report by Jiang et al, [ 13 ] who similarly observed no significant differences in CCPR or CLBR after comprehensive adjustment for confounders. In contrast to another retrospective study by Liu et al, [ 12 ] it revealed a negative association between dyslipidemia and CLBR (aOR [95% CI]: 0.702 [0.533–0.881], P  = .044), methodological distinctions likely explained this discrepancy. Importantly, our exploratory analysis indicated that dyslipidemia was associated with lower CCPR and CLBR in women aged <35 years. Since previous studies did not stratify by age, it remained unclear whether there were differences in cumulative pregnancy outcomes among younger reproductive-aged patients. Patients were also stratified into BMI <24 kg/m 2 and BMI ≥24 kg/m 2 subgroups, and it was found that the CLBR was similar between the dyslipidemia group and the control group in 2 subgroups. [ 12 ] This is consistent with the findings of our BMI subgroup analysis, despite the differences in analytical methods and BMI cutoff values across the studies. Previous investigations present conflicting evidence regarding lipid profiles and ART outcomes. Cai et al [ 27 ] reported in a cohort of 2011 women that lower quartiles of TC, LDL-C, and TG correlated with clinical pregnancy, lower quartiles of LDL-C, TC, and TG levels and higher quartiles of HDL-C levels associated with live births; conversely, higher LDL-C and lower HDL-C quartiles associated with miscarriage. Additionally, Yang et al [ 7 ] found that infertile women with elevated TC levels (TC ≥5.20 mmol/l) were less likely to achieve a live birth (OR(95% CI):0.86 (0.75–0.98), P   =  .028). Notably, our exploratory stratified analysis of lipid subtypes and severity thresholds revealed no significant associations between LDL-C, TG, TC, or HDL-C levels and pregnancy outcomes, which implied that isolated lipid parameter elevations may not directly dictate pregnancy success in ART cycles. This inconsistency could be attributed to variations in the study population, sample size, and data analysis methods. This study had several limitations. First, its retrospective design did not allow the establishment of a causal relationship between dyslipidemia and IVF/ICSI outcomes. Second, due to the limited sample sizes in some subgroup analyses, the statistical power was reduced, increasing the risk of Type II errors. To address the increased risk of Type I errors arising from multiple comparisons, we employed the Bonferroni correction, acknowledging that its conservative nature may further increase the likelihood of Type II errors. Additionally, the exclusion of patients with PCOS, while necessary to control for metabolic confounding, limited the generalizability of our findings, as PCOS represents a substantial proportion of infertile women with dyslipidemia. Finally, the lack of follicular fluid lipidomics measurements limited mechanistic exploration. In conclusion, while dyslipidemia was not associated with ART outcomes in the overall population, our exploratory analyses suggest that it may selectively compromise ART success in specific subgroups. Specifically, dyslipidemia was negatively associated with CCPR and CLBR in younger women (<35 years), and cleavage rate in women with obesity (BMI ≥28 kg/m²). Marginally elevated TG levels further accentuate oocyte retrieval impairment. In women ≥35 years subgroups, we observed trends toward a lower oocyte retrieval rate. These findings suggest dyslipidemia might serve as a potential modifiable biomarker for personalized IVF counseling, warranting prospective validation to inform pre-stimulation lipid management strategies.

We would like to thank Editage ( www.editage.cn ) for English language editing.