Couples exposure to neonicotinoid insecticides and its influence on in vitro fertilization/intracytoplasmic sperm injection outcomes

Abstract

Neonicotinoid insecticides (NEOs), recognized endocrine-disrupting compounds, are extensively utilized in agricultural and livestock practices, exhibiting global environmental persistence and bioaccumulation in human populations and wildlife. Accumulating experimental and epidemiological evidence implicates NEO exposure in reproductive dysfunction. This study investigated paired NEO concentrations in serum, female follicular fluid, and male seminal plasma from 211 couples undergoing assisted reproductive technology (ART), evaluating associations with reproductive outcomes. NEO concentrations were quantified using liquid chromatography-tandem mass spectrometry (LC–MS/MS), and associations with ART parameters, considering partner exposure metrics, were examined utilizing covariate-adjusted generalized linear models. Results revealed N-demethyl-acetamiprid (N-dm-ACE) was commonly detected across all matrices. Female serum clothianidin (CLO) demonstrated a significant inverse correlation with fertilization potential and good-quality embryo yield, while elevated serum N-dm-ACE concentrations were linked to diminished biochemical pregnancy probability. Within follicular fluid, CLO presence impaired fertilization and cleavage rates, and higher N-dm-ACE levels compromised good-quality embryo developmental competence. Additionally, THX detection in seminal plasma was linked to reduced cleavage rates and lower biochemical and clinical pregnancy rates. Negative dose–response relationships were observed between N-dm-ACE exposure in female serum and both biochemical and clinical pregnancies, and between N-dm-ACE levels in follicular fluid and cleavage rate and biochemical pregnancy. These results demonstrate that partner NEO exposure detrimentally influences pivotal ART success metrics, highlighting the critical importance of dual parental biomonitoring in reproductive toxicology research for clinical risk assessment. Comprehensive longitudinal cohort studies are warranted to substantiate these associations and elucidate mechanistic pathways.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12940-025-01242-y.

Introduction

Neonicotinoid insecticides (NEOs) are a class of neurotoxic compounds that target nicotinic acetylcholine receptors (nAChRs) and were originally developed to replace organophosphate and carbamate insecticides due to their enhanced selectivity and lower acute toxicity to mammals. NEOs are absorbed and translocated throughout plant tissues, including pollen, nectar, guttation fluids, and edible components, thereby posing exposure risks to non-target organisms across various trophic levels [1, 2]. Their extensive applications span agricultural pest control, seed treatments, and veterinary uses, with over 3,000 NEO-based formulations registered in China alone as of 2023, constituting 16.2% of national pesticide registrations [3, 4]. Despite their agricultural utility, environmental persistence—evidenced by soil half-lives exceeding 1,000 days for imidacloprid (IMI) [5]—raises significant concerns regarding bioaccumulation and endocrine-disrupting effects in non-target species [6, 7]. Accumulating evidence suggests that NEOs and their metabolites can function as endocrine-disrupting chemicals (EDCs), interfering with hormone-regulated reproductive pathways [8]. In vivo studies in animals have demonstrated that NEO exposure impairs key reproductive processes, including gametogenesis, steroidogenesis, and embryo implantation [9–12]. However, epidemiological data in humans remain limited and often yield inconsistent findings [13, 14].

Assisted reproductive technologies (ART), particularly in vitro fertilization and intracytoplasmic sperm injection (IVF/ICSI), have become essential interventions for infertile couples, with global demand steadily increasing. IVF/ICSI facilitates direct observation of pollutant effects on oocyte maturation, fertilization, and early embryogenesis—processes otherwise inaccessible in natural conception [15, 16]. Historically, research into the impact of environmental contaminants on ART outcomes has predominantly focused on maternal exposure, given the female’s critical role in providing the uterine environment for embryonic and fetal development [17, 18]. Our prior investigation significantly extended the evidentiary foundation by quantifying NEOs residues within paired serum and follicular fluid specimens obtained from 436 subfertile women undergoing ART. Our analytical results demonstrated that multiple NEOs—including acetamiprid (ACE), imidacloprid (IMI), and thiamethoxam (THX) were ubiquitously quantifiable across both biological matrices, exhibiting significant correlations with compromised reproductive endpoints [13]. Specifically, individuals with quantifiable follicular concentrations of THX, IMI, or ACE demonstrated significantly reduced yields of metaphase II oocytes, 2PN zygotes, and good-quality embryos compared to non-exposed counterparts. A dose–response relationship was observed, with increasing cumulative NEO burden in follicular fluid correlating with lower retrieval of total and mature oocytes, diminished fertilization rates, and compromised embryo development. Moreover, elevated levels of the metabolite N-desmethyl-acetamiprid (N-dm-ACE) were linked to impaired oocyte maturation and reduced 2PN zygote formation. Serum THX quantification revealed negative associations with fertilization competence and early embryonic cleavage potential, while serum IMI was linked to poor blastocyst quality. Crucially, follicular fluid IMI levels demonstrated inverse relationships with pivotal clinical outcomes, encompassing biochemical conception confirmation, clinical gestation establishment, and live birth attainment [13]. In addition, we further discovered that exposure to NEOs in serum or follicular fluid was associated with disrupted sex hormones and diminished ovarian reserve, implicating the endocrine-disrupting effects of NEOs [19].

Emerging evidence, however, challenges the conventional female-centric view of reproductive toxicology, demonstrating that the paternal contribution extends far beyond DNA delivery. Grounded in the concept of “paternal origins of health and disease,” it is now understood that sperm act as critical vectors, transmitting epigenetic information and toxicant-induced damage that can profoundly influence fertilization, embryo viability, and offspring health [20, 21]. Indeed, the vulnerability of spermatogenesis to environmental insults is well-documented [22]. Epidemiological studies have explicitly linked paternal exposure to various toxicants with adverse ART outcomes [23, 24]. For instance, elevated seminal plasma concentrations of lead have been associated with reduced fertilization potential in IVF cycles [25], while exposure to EDCs like phthalates in male partners has been correlated with lower rates of clinical pregnancy and live birth [26]. Recent findings implicate seminal NEO metabolites in compromised semen quality. Wang and colleagues documented high detection frequencies for desmethyl-acetamiprid (DM-ACE), imidacloprid-olefin (IMI-olefin), and desmethyl-clothianidin (DM-CLO) within seminal plasma. Notably, IMI-olefin concentrations showed significant inverse dose–response relationships with critical semen quality indicators [14]. These clinical effects are mechanistically supported by toxicogenomic research demonstrating that paternal exposures can induce sperm DNA damage and transmit aberrant DNA methylation and non-coding RNA profiles to the oocyte upon fertilization [27, 28].

Despite this mounting evidence, a significant knowledge gap persists. The majority of studies continue to assess exposures in a sex-specific manner, often overlooking the male partner or failing to employ a couple-based design that can untangle the combined and potentially synergistic effects of co-exposure [13, 14]. This may stem from a historical research bias and the complexities of couple-level study designs. Consequently, critical knowledge gaps remain, including: (1) limited profiling of NEO levels across biologically relevant matrices in infertile couples; (2) inadequate evaluation of metabolite-specific toxicities [29]; and (3) insufficient assessment of mixture effects from real-world exposures.

To address these critical gaps, the present study comprehensively assesses NEO exposure in paired biological matrices from couples undergoing IVF/ICSI. Through multivariable modeling that integrates paternal and maternal NEO concentrations with ART outcomes, we will quantify these couple-level relationships. These findings will advance the mechanistic understanding of NEO-mediated fertility impairment and offer a scientific foundation for mitigating reproductive risks in clinically vulnerable populations.

Materials and methods

Study population

This investigation recruited a cohort of 211 ART couples initiating their first IVF/ICSI cycle at the Reproductive Medicine Center, Sixth Affiliated Hospital, Sun Yat-sen University (Guangdong, China). Participant enrollment occurred during a 36-month recruitment period spanning January 2017 through January 2020. Data originated from the national multicentre project "Establishment and Application of the Assisted Reproductive Population and Offspring Cohort in China" (2016YFC1000200), which maintains surveillance across > 20 provincial-level administrative divisions nationwide. This initiative has established an extensive multicenter birth cohort incorporating 60,000 family units and 190,000 individuals from medically assisted reproduction populations, alongside integrated cryopreserved biospecimens and multi-domain clinical data repositories. The inclusion criteria for participant recruitment were as follows: (1) Couples diagnosed with infertility after one year of regular unprotected intercourse without achieving pregnancy; (2) At least one partner meets the clinical indications for IVF/ICSI, and the couple is undergoing assisted reproductive therapy for the first time.; (3) The fertilization protocol used is standard IVF or ICSI; and (4) Both partners are registered residents of Guangzhou with no plans to relocate within the next two years. Exclusion Criteria: (1) Either partner exhibits chromosomal karyotype abnormalities (including balanced translocations), has a family history of genetic disorders, or has a history of congenital gonadal dysgenesis; (2) Either partner has a history of exposure to radiation, chemotherapy, immunosuppressive therapy, or reproductive toxicants within the past 6 months; (3) Either partner suffers from severe hepatic or renal dysfunction, or has an active infection; (4) The female partner has a history of ovarian removal or abnormal development of the Müllerian ducts, or any other significant reproductive tract abnormalities. A comprehensive questionnaire elicited baseline characteristics encompassing paternal and maternal age, body mass index (BMI), smoking status, and alcohol consumption patterns. This research protocol received formal approval from the Institutional Review Board at The Sixth Affiliated Hospital, Sun Yat-sen University (2024ZSLYFEC-006), with all participants providing documented informed consent.

IVF/ICSI procedure

Ovarian stimulation protocols were tailored to individual ovarian reserve profiles and response patterns, as previously described [13]. Transvaginal ultrasound-guided oocyte retrieval was conducted 36 h post-trigger administration, with concurrent fresh semen collection from male partners. Metaphase II (MII) oocytes underwent either conventional insemination (IVF) or intracytoplasmic sperm injection (ICSI), with fertilization confirmed by 2 pronuclei (2PN) observation at 24 h post-insemination. Embryo development was assessed morphologically at 16–20 h.

Fresh embryo transfers (FET) were scheduled on day 3–5 post-retrieval, with transfer number determined by maternal age, embryo grading, and prior reproductive history. Cycle cancellation criteria included premature progesterone elevation (≥ 1.5 ng/mL on trigger day), moderate-severe ovarian hyperstimulation syndrome (OHSS) risk, or inadequate endometrial preparation (< 7 mm trilaminar pattern), prompting a freeze-all strategy. Frozen-thawed embryo transfers (TET) employed either modified natural cycles (ovulation-triggered) or programmed artificial cycles (exogenous hormone-supported), selected based on cycle characteristics. Luteal phase support with vaginal progesterone ± estradiol was maintained until β-hCG confirmation 14 days post-transfer.

Outcome assessment

This investigation assessed six principal efficacy endpoints: fertilization success, embryonic cleavage progression, good-quality embryo yield, biochemical conception confirmation, clinical gestation establishment, and live birth attainment. Embryologists blinded to exposure status quantified retrieved metaphase II (MII) oocytes, two-pronuclei (2PN) zygotes, and morphologically superior embryos. Transfer-eligible cleavage-stage specimens demonstrated grade 1–2 morphology with ≥ 5 blastomeres, while good-quality counterparts exhibited 6–10 blastomeres within identical morphological parameters per established criteria [30]. Blastocyst assessment employed the Gardner system, with scores ≥ 3BB (integrating expansion degree, trophectoderm, and inner cell mass morphology) designating good-quality status [31].

The fertilization rate was calculated as the percentage of oocytes with 2PN divided by the total number of oocytes retrieved. The cleavage rate was defined as the proportion of 2PN zygotes that progressed to cleavage-stage embryos (≥ 2 cells) by Day 3 post-insemination. The good-quality embryo rate was calculated using the formula: (Number of good-quality embryos/Total number of cleaved embryos or blastocysts) × 100%. Biochemical pregnancy was defined by a positive β-hCG result in both urine and serum on day 14, with no further evidence of a gestational sac or fetal heartbeat. Clinical pregnancy was defined as an ultrasound-confirmed intrauterine pregnancy at 35 days post-embryo transfer. Live birth was defined as the delivery of one or more live neonates after 28 weeks of gestation.

Sample collection

Biological specimens were collected using standardized protocols. Venous blood from both partners was obtained via venipuncture into sterile EDTA-containing anticoagulant tubes during oocyte retrieval/semen collection, without fasting requirements. Follicular fluid aspiration was performed under ultrasound guidance with transvaginal fine needle puncture, prioritizing the retrieval of fluid from the first accessible large ovarian follicle, while avoiding potentially blood-contaminated dominant follicles to minimize hemoglobin interference. All follicular fluid samples were immediately inspected visually, with those showing erythrocyte contamination being excluded from further analysis. Seminal plasma samples were collected through masturbation in a private room near the semen analysis laboratory, using sterile plastic specimen cups. The collected samples provide a cross-sectional assessment of exposure at a single, clinically relevant timepoint during the IVF/ICSI procedure.

All biological specimens underwent immediate processing within 30 min post-collection. Whole blood samples were subjected to centrifugation at 2000 × g for 10 min (4 °C) to isolate serum. Follicular fluid samples were centrifuged at 2000 × g for 10 min at 4 °C to pellet cellular debris, after which the supernatant was aliquoted and stored at −80 °C until analysis. Freshly collected semen samples were allowed to liquefy at 37 °C for 30–60 min under sterile conditions, after which they underwent sequential centrifugation to isolate seminal plasma: first, samples were centrifuged at 300 × g for 10 min at room temperature to sediment spermatozoa, then the supernatant was carefully aspirated and transferred to a sterile microcentrifuge tube before being centrifuged again at 2000 × g for 10 min at 4 °C to eliminate residual cellular debris and particulate matter, and finally the resulting clarified seminal plasma supernatant was aliquoted into cryovials and stored at −80 °C until chemical analysis.

Analytical methods

Fifteen NEO and their metabolites were quantified across systemic circulation (serum), female reproductive (follicular fluid), and male reproductive (seminal plasma). Analytes encompassed parent compounds: ACE, THX, clothianidin (CLO), IMI, imidaclothiz (IMZ), dinotefuran (DIN), flonicamid (FLO), sulfoxaflor (SUF), nitenpyram (NIT), and thiacloprid (THI), alongside key metabolites: thiacloprid-amide (TA), N-dm-ACE, olefin-imidacloprid (Of-IMI), 6-chloronicotinic acid (6-CN), and N-desmethyl-thiamethoxam (N-DMT). Analytical workflows implemented validated salting-out assisted liquid–liquid extraction (SALLE) coupled with liquid chromatography-tandem mass spectrometry (LC–MS/MS) protocols, aligning with contemporary biomonitoring methodologies [13].

The SALLE extraction and clean-up process effectively removed interferences, ensuring the selective extraction of the target analytes from the biological matrices. Reversed-phase chromatography, utilizing methanol and water as mobile phases, enabled the efficient separation of the 15 NEOs. Method performance was rigorously assessed through the evaluation of calibration curve linearity, sensitivity, accuracy, and precision, as previously documented [32].

Statistical analysis

Descriptive analyses characterized demographic characteristics, clinical endpoints, and NEOs concentrations. Participant demographics and clinical features were expressed as mean ± standard deviation (SD) or proportional frequencies according to variable distribution. NEOs distributions across serum, follicular fluid, and seminal plasma matrices were displayed with geometric means (GM), minimum, medians and max concentrations. Specimens below the limit of detection (LOD) underwent imputation using LOD/√2. Couple NEO concentration correlations were evaluated via Spearman’s rank correlation.

Analytes exhibiting > 10% detection frequency underwent further analysis to ensure statistical robustness. For N-dm-ACE with > 80% detectability across matrices, participants were stratified into tertiles based on matrix-specific concentrations. Exposure–response trends across ascending tertiles (coded 1–3) were examined through linear regression modeling. Models incorporated covariate adjustments for age, BMI, smoking status, alcohol consumption, ovarian stimulation protocols, infertility etiology, type of fertilization, embryo quality, day of embryo transfer, number of embryos transferred and partner NEO levels. A generalized linear model (GLM) was used to assess the association between NEO exposure levels (as a continuous variable) and composite exposure quantity (as a categorical variable) with assisted reproductive outcomes. Association magnitudes were expressed as β coefficients (continuous exposures) or adjusted odds ratios (OR) with corresponding 95% confidence intervals.

To account for the multiple comparisons performed across various neonicotinoids and ART outcomes, we controlled the False Discovery Rate (FDR) using the Benjamini–Hochberg procedure. A nominal p-value < 0.05 was considered indicative of a potential association, while an FDR-adjusted q-value < 0.05 was used as the threshold for statistical significance robust to multiple comparisons. Statistical analyses were conducted using R version 4.3.2 (R Foundation for Statistical Computing, Austria).

Result

Demographic profile of the study cohort

This investigation enrolled 211 subfertile couples undergoing assisted reproductive technology (ART: IVF/ICSI). Key demographic and clinical parameters are comprehensively tabulated in Table 1. The female cohort averaged 32.23 ± 4.12 years, while male partners exhibited a mean age of 34.13 ± 5.43 years. Female BMI spanned 14.9–34.2 kg/m² (mean 22.12 ± 3.08 kg/m²), contrasting with male BMI averaging 24.97 ± 3.73 kg/m². Merely 7.6% of females and 2.8% of males met underweight diagnostic criteria (BMI < 18.5 kg/m²). Given established associations between smoking and alcohol consumption and impaired reproductive capacity, we documented that 31.3% of males reported tobacco use and 15.6% alcohol consumption, whereas no females endorsed either exposure.

Table 1.

Demographic and clinical characteristics of 211 couples

Individual characteristics	Females	Males
Maternal age (y)	32.23 ± 4.12	34.13 ± 5.43
≥ 35y	53 (25.1%)	86 (40.8%)
< 35y	158 (74.9%)	125 (59.2%)
BMI (kg/m2)	22.12 ± 3.08	24.97 ± 3.73
Underweight (BMI < 18.5)	16 (7.6%)	6 (2.8%)
Normal (18.5 ≤ BMI < 24)	138 (65.4%)	84 (39.8%)
Overweight (BMI ≥ 24)	57 (27.0%)	121 (57.4%)
Smoking status
Smoker	0 (0%)	66 (31.3%)
Non-smoker	100 (100%)	145 (68.7%)
Drinking status
Drinker	0 (0%)	33 (15.6%)
Non-drinker	100 (100%)	178 (84.4%)
Ever been pregnant	121 (57.3%)	NA
Ever delivered	43 (20.4%)	NA
Basal hormone profiles
FSH (IU/L)	7.62 ± 5.30	NA
LH (IU/L)	5.20 ± 2.84	NA
Estradiol (pg/mL)	48.34 ± 59.07	NA
Testosterone (ng/mL)	0.28 ± 0.77	NA
AMH (ng/ml)	2.95 ± 2.21	NA
AFC	11.89 ± 6.18	NA
Semen quality parameters
Sperm concentration (106/mL)	NA	54.37 ± 38.38
Total sperm count (106)	NA	181.40 ± 135.30
Total motility (%)	NA	55.54 ± 25.72
Progressive motility (%)	NA	47.89 ± 24.04
Sperm normal morphological rate (%)	NA	4.50 ± 3.48
DFI (%)	NA	17.14 ± 11.68
HDS (%)	NA	7.24 ± 4.99
Couple/cycle-specific characteristics
Infertility causes
Female factor alone	162 (76.8%)
Male factor alone	26 (12.3%)
Both	17 (8.1%)
Unexplained	6 (2.8%)
Oocyte insemination technique
IVF	161 (76.3%)
ICSI	50 (23.7%)
Treatment protocol
Long GnRH agonist protocol	115 (54.5%)
GnRH antagonist protocol	52 (24.6%)
Mild stimulation	6 (2.8%)
Other protocol	38 (18.1%)
Controlled ovarian hyperstimulation outcomes
Total number of oocytes retrieved	9.75 ± 5.01
Mature (MII) oocytes retrieved	6.27 ± 3.88
Normal (2PN) fertilized oocytes	5.70 ± 3.52
Total embryos	4.51 ± 3.01
High-quality embryos	3.74 ± 2.68
Embryo transfer day
Day 3	193 (91.5%)
Day 5	18 (8.5%)
Number of embryos transferred
1 embryo	77 (36.5%)
2 embryos	134 (63.5%)
Pregnancy outcomes
Biochemical pregnancy	119 (56.4%)
Clinical pregnancy	114 (54.0%)
Miscarriage	21 (18.4%)
Live birth	93 (44.1%)

Abbreviations: BMI body mass index, FSH follicular stimulating hormone, LH luteinizing hormone, AMH anti mullerian hormone, AFC antral follicle count, DFI DNA fragmentation index, HDS High DNA stainability, IVF in vitro fertilization, ICSI intracytoplasmic sperm injection

Table 1 additionally details female basal hormonal profiles, male seminal parameters, and ART outcome metrics. Among 211 completed ART cycles, IVF constituted 76.3% (n = 161) and ICSI 23.7% (n = 50). Infertility etiologies included: female-factor (76.8%), male-factor (12.3%), combined (8.1%), and unexplained infertility (2.8%). Oocyte retrieval yielded 1–32 oocytes per cycle (mean 9.75 ± 5.01). Embryo transfer procedures featured single-embryo transfers in 36.5% of cases, with cleavage-stage transfers predominating (91.5%) on day 3. Transfer-inclusive cycles yielded biochemical pregnancy, clinical pregnancy, and live birth rates of 56.4%, 54.0%, and 44.1% respectively (Table 1).

Distribution profiles of NEOs in biological matrices

Quantitative distributions and LODs for target NEOs across serum, follicular fluid, and seminal plasma compartments are detailed in Table 2. N-dm-ACE demonstrated predominant concentrations, with THX consistently ranking second highest among both sexes. Within female follicular fluid, analyte concentrations spanned 0.003 ng/mL (THI) to 0.239 ng/mL (N-dm-ACE), while serum measurements ranged from 0.017 ng/mL (FLO) to 0.226 ng/mL (N-dm-ACE). Male seminal plasma concentrations between 0.010 ng/mL (SUF) and 0.122 ng/mL (N-dm-ACE), contrasting with serum levels of 0.010 ng/mL (THI) to 0.269 ng/mL (N-dm-ACE). Analytes exceeding 10% detection frequency (N-dm-ACE, THX, CLO) fulfilled statistical modeling inclusion criteria..

Table 2.

Distribution of serum, follicular fluid and seminal plasma NEOs in couples undergoing IVF/ICSI

NEOs	Gender	Sample	Detection rate ≥ LOD (%)	GM ± SD	Min	Median	Max
N-dm-ACE	Male	Serum	100%	0.269 ± 2.355	0.029	0.237	3.080
		Seminal plasma	84.8%	0.122 ± 4.878	< LOD	0.171	10.560
	Female	Serum	100%	0.226 ± 2.619	0.021	0.203	3.054
		Follicular fluid	100%	0.239 ± 2.922	0.021	0.215	6.994
THX	Male	Serum	46.4%	0.036 ± 2.537	< LOD	< LOD	1.878
		Seminal plasma	49.8%	0.057 ± 2.173	< LOD	< LOD	3.298
	Female	Serum	19.0%	0.056 ± 1.731	< LOD	< LOD	2.517
		Follicular fluid	22.3%	0.055 ± 2.006	< LOD	< LOD	2.660
CLO	Male	Serum	25.6%	0.028 ± 2.280	< LOD	< LOD	1.250
		Seminal plasma	17.5%	0.064 ± 1.769	< LOD	< LOD	1.383
	Female	Serum	11.8%	0.060 ± 1.531	< LOD	< LOD	0.875
		Follicular fluid	38.4%	0.035 ± 2.435	< LOD	< LOD	3.376
ACE	Male	Serum	9.5%	0.020 ± 1.580	< LOD	< LOD	0.296
		Seminal plasma	40.0%	0.016 ± 1.964	< LOD	< LOD	0.342
	Female	Serum	15.2%	0.020 ± 1.932	< LOD	< LOD	1.403
		Follicular fluid	26.5%	0.005 ± 3.444	< LOD	< LOD	1.921
IMI	Male	Serum	5.5%	0.035 ± 1.510	< LOD	< LOD	0.440
		Seminal plasma	49.3%	0.048 ± 2.221	< LOD	< LOD	1.108
	Female	Serum	15.6%	0.047 ± 1.632	< LOD	< LOD	0.750
		Follicular fluid	45.0%	0.051 ± 2.401	< LOD	< LOD	1.630
SUF	Male	Serum	5.2%	0.024 ± 1.280	< LOD	< LOD	0.110
		Seminal plasma	4.7%	0.010 ± 1.334	< LOD	< LOD	0.104
	Female	Serum	3.8%	0.044 ± 1.236	< LOD	< LOD	0.169
		Follicular fluid	0.9%	0.022 ± 1.070	< LOD	< LOD	0.045
FLO	Male	Serum	2.8%	0.014 ± 1.196	< LOD	< LOD	0.094
		Seminal plasma	0	< LOD	< LOD	< LOD	< LOD
	Female	Serum	0.5%	0.017 ± 1.025	< LOD	< LOD	0.024
		Follicular fluid	1.9%	0.029 ± 1.445	< LOD	< LOD	3.517
6-CN	Male	Serum	1.9%	0.211 ± 1.199	< LOD	< LOD	2.243
		Seminal plasma	0	< LOD	< LOD	< LOD	< LOD
	Female	Serum	1.9%	0.187 ± 1.212	< LOD	< LOD	0.942
		Follicular fluid	0.5%	0.196 ± 1.112	< LOD	< LOD	0.909
THI	Male	Serum	0.9%	0.010 ± 1.050	< LOD	< LOD	0.018
		Seminal plasma	1.4%	0.011 ± 1.081	< LOD	< LOD	0.027
	Female	Serum	0	< LOD	< LOD	< LOD	< LOD
		Follicular fluid	6.2%	0.003 ± 1.808	< LOD	< LOD	0.406
DIN	Male	Serum	0	< LOD	< LOD	< LOD	< LOD
		Seminal plasma	20.4%	0.070 ± 1.803	< LOD	< LOD	1.185
	Female	Serum	0	< LOD	< LOD	< LOD	< LOD
		Follicular fluid	1.4%	0.163 ± 1.079	< LOD	< LOD	0.340
IMZ	Male	Serum	0.5%	0.046 ± 1.080	< LOD	< LOD	0.138
		Seminal plasma	4.7%	0.019 ± 1.268	< LOD	< LOD	0.208
	Female	Serum	0.5%	0.046 ± 1.71	< LOD	< LOD	0.125
		Follicular fluid	2.8%	0.027 ± 1.512	< LOD	< LOD	1.800
TA	Male	Serum	0.5%	0.018 ± 1.031	< LOD	< LOD	0.029
		Seminal plasma	9.0%	0.013 ± 1.224	< LOD	< LOD	0.047
	Female	Serum	0	< LOD	< LOD	< LOD	< LOD
		Follicular fluid	5.2%	0.013 ± 1.405	< LOD	< LOD	0.241
OF-IMI	Male	Serum	0	< LOD	< LOD	< LOD	< LOD
		Seminal plasma	3.3%	0.054 ± 1.371	< LOD	< LOD	2.384
	Female	Serum	8.1%	0.038 ± 1.318	< LOD	< LOD	0.204
		Follicular fluid	4.7%	0.072 ± 1.212	< LOD	< LOD	0.366
NIT	Male	Serum	0	< LOD	< LOD	< LOD	< LOD
		Seminal plasma	21.3%	0.043 ± 1.337	< LOD	< LOD	0.228
	Female	Serum	2.4%	0.017 ± 1.148	< LOD	< LOD	0.058
		Follicular fluid	1.9%	0.038 ± 1.435	< LOD	< LOD	6.248
N-DMT	Male	Serum	0	< LOD	< LOD	< LOD	< LOD
		Seminal plasma	1.9%	0.088 ± 1.337	< LOD	< LOD	0.759
	Female	Serum	0	< LOD	< LOD	< LOD	< LOD
		Follicular fluid	4.3%	0.043 ± 1.089	< LOD	< LOD	0.085

NEOs Neonicotinoids, IVF/ICSI in vitro fertilization/intracytoplasmic sperm injection, GM geometric mean, SD standard deviation, LOD limit of detection, Min minimum value, Max maximum value

Notably, 70.1% of females exhibited multiple NEOs co-exposure (> 2 compounds) within follicular environments, whereas only 47.9% demonstrated comparable multi-residue detection in serum. Among male serum specimens, 73 individuals (34.6%) revealed ≥ 1 detectable neonicotinoid, with maximum co-occurrence of four analytes observed in 13 participants (6.2%). Significantly, 80.1% of males manifested multi-NEO exposure (> 2 compounds) in seminal plasma.

Correlation of NEOs exposure levels in biological matrices

Cohabitation-induced exposure synchrony may arise from shared residential and behavioral patterns within couples. Our biomonitoring analysis identified statistically significant positive correlations for specific NEOs between partnered individuals. Serum concentrations demonstrated robust partner correlations for N-dm-ACE (r = 0.22), THX (r = 0.80), and CLO (r = 0.36), with all associations achieving statistical significance (p < 0.05) (Fig. 1A). Notably, a significant exposure correlation emerged between female follicular fluid and male seminal plasmaN-dm-ACE (r = 0.28, p < 0.05). Conversely, neither THX nor CLO exhibited significant correlations across these biological matrices (p > 0.05) (Fig. 1B).

Fig. 1.

Heatmap of spearman correlation coefficients of NEOs concentrations between males and females of the couples. A Correlation analysis of NEOs exposure levels in the serum of 211 couples. B Correlation analysis of NEOs exposure levels in male seminal plasma and female follicular fluid. Red represents positive correlation; blue represents negative correlation. FS, female serum; MS, male serum; FF, follicular fluid; SP, seminal plasma; * for p < 0.05

Female biological matrices NEOs concentrations and ART outcomes

In our analysis of female exposure, associations were evaluated by stratifying NEOs based on detection status (for THX and CLO) or median concentrations (for N-dm-ACE). At a nominal significance level (p < 0.05), several potential associations emerged. Specifically, serum N-dm-ACE concentrations above the median were linked to reduced rates of biochemical pregnancy, clinical pregnancy, and live birth (all p < 0.05) (Fig. 2). Similarly, high serum and follicular fluid N-dm-ACE concentrations were associated with a lower cleavage rate (all p < 0.05) (Fig. 2). We also observed that an increasing number of NEOs detected in serum was associated with diminished cleavage and good-quality embryo rates (all p < 0.05) (Table S1). However, none of these nominal associations remained statistically significant after applying a Benjamini–Hochberg correction to control the FDR (all q > 0.05) (Table S1). Furthermore, no significant associations were observed between the detection of THX or CLO in either serum or follicular fluid and any of the assessed ART outcomes (all p > 0.05) (Table S1).

Fig. 2.

Forest plot of associations between female serum N-dm-ACE and ART outcomes. The plot shows beta coefficients (β) for continuous outcomes (top) and odds ratios (OR) for binary outcomes (bottom) from fully adjusted generalized linear models. Models were adjusted for the covariates listed in the methods section

Male biological matrices NEOs concentrations and ART outcomes

For the male exposure analysis, neonicotinoids were stratified by detection status (detectable vs. non-detectable) for THX and CLO, and by the median for N-dm-ACE concentrations. At a nominal significance level, we identified several associations. The detection of THX in seminal plasma was associated with reduced embryo cleavage rates, as well as lower rates of biochemical pregnancy, clinical pregnancy, and live birth (all p < 0.05) (Fig. 3). Furthermore, serum N-dm-ACE concentrations above the median were linked to decreased probabilities of biochemical and clinical pregnancy (all p < 0.05). Similarly, elevated seminal plasma N-dm-ACE was associated with a diminished likelihood of live birth (p < 0.05). However, it is critical to note that none of these nominal associations withstood correction for multiple comparisons using the Benjamini–Hochberg procedure (all q > 0.05), as detailed in Table S2.

Fig. 3.

Forest plot of associations between seminal plasma THX and ART outcomes. The plot shows beta coefficients (β) for continuous outcomes (top) and odds ratios (OR) for binary outcomes (bottom) from fully adjusted generalized linear models. Models were adjusted for the covariates listed in the methods section

Couples' biological matrices NEOs concentrations and ART outcomes

To investigate the joint effects of NEO co-exposure at the couple level, we utilized fully adjusted generalized linear models. These models simultaneously incorporated both partners’ exposure data while controlling for a comprehensive set of covariates, including demographics, lifestyle factors, and clinical treatment parameters.

Within these models, we identified several nominal associations originating from each partner. For the female partner, exposure was primarily linked to early embryological outcomes. Specifically, detectable serum THX was associated with a lower rate of good-quality embryo formation (p < 0.05), while elevated N-dm-ACE concentrations in both serum and follicular fluid were linked to reduced cleavage rates (p < 0.05) (Fig. 4A-B). Furthermore, a higher count of detected NEOs in both serum and follicular fluid was associated with impaired embryo development and lower pregnancy success (p < 0.05) (Fig. 4A, C). For the male partner, exposure effects appeared to manifest at later stages of pregnancy. Even after accounting for female exposure, detectable THX in seminal plasma was associated with diminished clinical pregnancy success and live birth rates (Fig. 4D). Similarly, high concentrations of N-dm-ACE in seminal plasma were linked to a significant reduction in live births (p < 0.05) (Fig. 4D).

Fig. 4.

Forest plot of associations between couple NEOs exposure and ART outcomes. Forest plots show beta coefficients (β) for continuous embryological outcomes (A, B) and odds ratios (OR) for binary pregnancy outcomes (C, D) from fully adjusted models with p < 0.05. Associations are shown for exposures measured in female serum (A), follicular fluid (B, C), and seminal plasma (D). Models were adjusted for covariates listed in the Methods

Taken together, these compound-specific effects, which persisted after full covariate adjustment, suggest a pattern of bidirectional reproductive toxicity arising from couple-level NEO co-exposure (Table S3). However, it is important to note that none of these associations remained statistically significant after adjusting for multiple comparisons using the FDR (all q > 0.05).

Female biological matrices N-dm-ACE concentrations and ART outcomes

Due to the ubiquitous detection (100%) of N-dm-ACE in both follicular fluid and serum, we evaluated its exposure–response relationships using tertile-based stratification (low, medium, high). This trend analysis revealed several notable, albeit nominal, associations. In follicular fluid, increasing N-dm-ACE concentrations were associated with a significant dose-dependent reduction in both cleavage rates (p-trend = 0.015) and the likelihood of biochemical pregnancy (p-trend = 0.048) (Fig. 5A). A similar monotonic adverse trend was observed for serum N-dm-ACE across critical clinical endpoints, including lower rates of biochemical pregnancy and clinical pregnancy (all p-trend < 0.05) (Fig. 5A).

Fig. 5.

Dose–response relationships between individual N-dm-ACE exposure tertiles and pregnancy outcomes in ART. Odds ratios (ORs) for biochemical, clinical, and live birth outcomes according to tertiles (T1-T3) of N-dm-ACE concentrations in female (A) and male (B) biological matrices. The lowest tertile (T1) is the reference group. p-values for linear trend are shown, with an asterisk (*) denoting p < 0.05. All models were fully adjusted for covariates

However, while these dose–response trends were significant at a nominal level, none remained statistically significant after controlling for the FDR (all q > 0.05). Detailed results of this analysis are presented in Table S4.

Male biological matrices N-dm-ACE concentrations and ART outcomes

Due to high detection rates of N-dm-ACE in both male serum and seminal plasma (> 80%), we assessed dose–response relationships by stratifying concentrations into tertiles. A nominally significant monotonic trend was observed for serum N-dm-ACE, where increasing concentrations were associated with a reduced likelihood of achieving biochemical pregnancy and live birth (p-trend < 0.05) (Fig. 5B). However, this association was not robust to correction for multiple comparisons (q > 0.05). In contrast, no significant dose–response trends were found for either seminal or serum N-dm-ACE concentrations with early embryological outcomes, such as fertilization, cleavage, or high-quality embryo rates (all p-trend > 0.05) (Table S5).

Taken together, this pattern suggests that systemic male N-dm-ACE exposure may selectively impair post-implantation success rather than early embryogenesis. While this finding is exploratory, it provides a compelling hypothesis for future investigation.

Couples’ biological matrices N-dm-ACE concentrations and ART outcomes

To assess couple-level co-exposure effects of N-dm-ACE, we utilized fully adjusted generalized linear models that simultaneously included concentrations from both partners, while controlling for a comprehensive set of demographic and clinical covariates.

The analysis revealed that the female partner’s exposure was the primary driver of the observed nominal associations. Specifically, we found significant dose–response relationships where increasing serum N-dm-ACE concentrations were linked to lower probabilities of biochemical (p-trend = 0.032) and clinical pregnancy (p-trend = 0.021) (Fig. 6). Similarly, higher follicular fluid N-dm-ACE concentrations were associated with a monotonic decrease in embryo cleavage rates (p-trend = 0.015) (Table S6). While these trends for female exposure are notable, it is crucial to state that none of them remained statistically significant after correction for multiple comparisons (all q > 0.05).

Fig. 6.

Dose–response relationships between couple N-dm-ACE exposure tertiles and pregnancy outcomes in ART. Odds ratios (ORs) for biochemical, clinical, and live birth outcomes according to tertiles (T1-T3) of N-dm-ACE concentrations in female (A) and male (B) biological matrices. The lowest tertile (T1) is the reference group. p-values for linear trend are shown, with an asterisk (*) denoting p < 0.05. All models were fully adjusted for covariates

In contrast, no significant dose–response associations were found for the male partner’s N-dm-ACE exposure in either serum or seminal plasma across any reproductive endpoint (all p-trend > 0.05) (Table S6, Fig. 6).

Discussion

To our knowledge, this study represents one of the first comprehensive investigations into the associations of multiple NEOs with ART outcomes, uniquely integrating individual and joint couple-level exposures. Our analyses identified several suggestive, dose-dependent associations, particularly for the metabolite N-dm-ACE, linking its concentration to poorer outcomes across the reproductive continuum—from embryo development to live birth. However, a critical overarching finding is that none of these nominal associations withstood a rigorous correction for multiple comparisons. This suggests our study may have been underpowered to detect modest effects. We therefore position our findings as exploratory and hypothesis-generating, providing a compelling foundation for future, larger-scale research into the biological plausibility and public health implications of these consistent trends.

A key observation underpinning our results is the pervasive presence of N-dm-ACE, which was detected in 100% of serum and follicular fluid samples, and 84.8% of seminal plasma samples. This high prevalence contrasts sharply with other NEOs and can be attributed to its distinct physicochemical properties. As a major metabolite of ACE, N-dm-ACE possesses greater chemical stability and water solubility than its parent compound. While its polarity facilitates renal excretion after metabolism by CYP450 enzymes [29], its high-water solubility contributes to persistent contamination of water bodies, creating a continuous exposure route via drinking water. Furthermore, its chemical structure enhances ionization efficiency in LC–MS/MS analysis, leading to a lower limit of detection and making it more readily quantifiable at trace levels [32].

Interestingly, our data revealed distinct sex-specific exposure patterns, exemplified by the significantly higher detection of THX in male matrices. While the underlying mechanisms remain to be elucidated, we propose a testable hypothesis centered on sex-specific differences in NEO pharmacokinetics. On the female side, estrogen is known to upregulate key hepatic detoxification enzymes like CYP3A4, which could potentially accelerate systemic clearance of compounds like THX [33]. Conversely, males may face different challenges; androgens can suppress other enzymes like glutathione-S-transferase, potentially reducing systemic detoxification capacity [30]. Furthermore, male-specific physiology may create a uniquely vulnerable environment. For instance, toxicants could potentially penetrate the blood-testis barrier and accumulate within the seminiferous tubules, leading to prolonged, high-concentration exposure for developing sperm [34]. This localized exposure, possibly combined with limited metabolic capacity within the testes themselves, could induce damage that is not reflected by systemic serum levels alone [35]. However, we must stress that this is a speculative connection. Whether these specific pathways are clinically relevant for NEO metabolism in humans—and are substantial enough to explain our findings—remains unproven without dedicated pharmacokinetic studies. Furthermore, factors like occupational exposure patterns and physiological barrier selectivity could also contribute to any observed sex-based differences [36]. Our findings highlight the importance of the mixed exposure model, emphasizing the need for gender-specific metabolomics and barrier permeability mechanism research to provide a theoretical basis for accurate assessment of the reproductive risks posed by environmental pollutants.

The potential for paternal NEO exposure to adversely impact ART outcomes is strongly supported by toxicological evidence. A primary mechanism is the induction of oxidative stress (OS). Animal models demonstrate that NEOs can disrupt testicular antioxidant defenses, leading to an overproduction of reactive oxygen species (ROS) that damage sperm membranes and DNA [10, 37]. This aligns with findings that ACE exposure damages testicular interstitial cells and impairs testosterone synthesis, likely through ROS-mediated mitochondrial dysfunction [38]. Such damage can manifest as increased sperm DNA fragmentation (SDF), a well-established factor linked to poor embryo quality and pregnancy loss [11, 39].Beyond direct damage, there is growing concern that NEOs could induce epigenetic alterations in sperm, such as aberrant DNA methylation, which could disrupt embryonic gene expression and post-implantation development [27, 40]. These pathways provide a robust biological rationale for our observed associations between male NEO exposure and reduced pregnancy and live birth rates.

Similarly, several literature documents the detrimental effects of NEOs on the female reproductive system. Studies show that chronic exposure can disrupt the hypothalamic-pituitary–gonadal axis, leading to impaired follicular development, hormonal imbalances, and ovarian damage [11]. At the gamete level, in vitro models confirm that NEOs can directly impair oocyte maturation and reduce fertilization rates and blastocyst formation, consistent with our clinical observation linking a higher number of NEO exposures to poorer embryo quality [28, 41]. These findings are highly consistent with our clinical observations, where an increase in the exposure number of NEOs in the serum of infertile women is associated with a decrease in cleavage rate and good-quality embryo rate. Crucially, NEO toxicity may extend beyond fertilization to post-implantation events. Mechanistic studies suggest NEOs can disrupt placental hormone synthesis by abnormally activating aromatase (CYP19), leading to an imbalanced estrogen microenvironment critical for trophoblast invasion and decidualization [42, 43]. This specific mechanism offers a compelling explanation for our dose-dependent finding linking higher female serum N-dm-ACE with increased risks of biochemical and clinical pregnancy loss.

A key innovation of our study is the assessment of joint couple-level exposure, which reflects a more realistic real-world scenario. We observed significant positive correlations for N-dm-ACE, THX, and clothianidin levels between male and female partners, underscoring that couples often share common exposure sources through diet and environment. This finding suggests that focusing on a single partner may underestimate the total environmental burden on a couple’s reproductive potential. Our approach aligns with recent research, such as that by Al-Saleh et al., who found that associations between female phthalate exposure and pregnancy failure became more pronounced after adjusting for the male partner’s exposure, implying a cumulative or synergistic effect [26, 44]. The couple-based model is therefore essential for accurately assessing reproductive risks posed by ubiquitous environmental pollutants.

From a clinical perspective, our findings, while exploratory, may have important implications for ART practices. A significant portion of couples undergoing ART are diagnosed with idiopathic infertility, a frustrating label for both patients and clinicians. Our results raise the hypothesis that ubiquitous, low-level exposure to environmental contaminants like NEOs could be one of the many unmeasured factors contributing to this diagnostic challenge, potentially explaining a fraction of cases with otherwise normal clinical parameters but poor reproductive outcomes. Acknowledging this potential link has practical applications for patient counseling. While routine biomonitoring for these compounds is not yet feasible or warranted based on current evidence, clinicians can play a pivotal role in preconception care. This includes empowering patients with knowledge about potential exposure sources (e.g., diet, residential pesticide use) and recommending simple, precautionary exposure-reduction strategies. Such conversations can be integrated into existing lifestyle counseling alongside advice on diet, smoking, and alcohol consumption. This proactive approach is not only low-risk but may be particularly valuable for couples experiencing unexplained treatment failures, providing them with actionable steps they can take to optimize their chances of success. By considering this environmental dimension, clinicians may be better equipped to provide holistic support to their patients, potentially improving outcomes even in the absence of definitive mechanistic proof.

This study possesses several strengths, including its couple-based design, the assessment of multiple NEOs across three distinct biological matrices, and its focus on a clinically relevant ART population. However, we must also acknowledge its limitations. First, exposure was assessed from single spot samples, which may not capture the temporal variability of exposure to compounds with short biological half-lives. This could lead to non-differential misclassification, likely biasing our results towards the null and underestimating true effect sizes. Despite this limitation, our sampling occurred during a critically relevant biological window—the time of ovarian stimulation and gamete retrieval—which likely captures the exposure most directly impacting the outcomes of that specific ART cycle. Future longitudinal studies incorporating repeated measurements would be beneficial to better characterize long-term exposure patterns and their stability over time. Second, as previously noted, the lack of statistical significance after FDR correction indicates that our study was likely underpowered. The reported nominal associations should therefore be interpreted with caution. These limitations underscore the need for future validation in larger, multi-center cohort studies with longitudinal exposure assessment to confirm these important, hypothesis-generating findings.

Conclusion

In conclusion, our study identifies neonicotinoids, particularly the pervasive metabolite N-dm-ACE, as chemicals of significant concern for human reproductive health. Crucially, we demonstrate for the first time that co-exposure at the couple level is a significant risk factor, exerting a synergistic negative effect on ART outcomes. These findings highlight the importance of adopting a couple-level approach in environmental reproductive epidemiology and support the integration of precautionary exposure-reduction counseling into clinical practice to better serve couples striving for pregnancy.

Authors’ contributions

Zyu Liu and Hao Shi contributed equally and are considered co–first authors. Ziyu Liu: Investigation, Data curation, Interpretation of data, Writing – original draft. Hao Shi: Formal analysis, Methodology, Software. Nijie Li: Conceptualization, Validation, Sample collection. Xin Zhao: Administrative, technical, or material suppor. Zhenhan Xu: Sample testing, Data analysis. Guihua Liu: Funding acquisition, Supervision. Xiaoyan Liang: Critical review of the manuscript for important intellectual content. Xing Yang: Writing—review & editing, Funding acquisition.

Funding

The authors wish to acknowledge the Natural Science Foundation of China (No. 82571898), Natural Science Foundation of Guangdong Province (No. 2023A1515012940), and Key laboratory start-up project (Sixth Affiliated Hospital of Sun Yat-sen University) (No. 2023WST04) for their partial research support.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This research protocol received formal approval from the Institutional Review Board at The Sixth Affiliated Hospital, Sun Yat-sen University (approval number: 2024ZSLYFEC-006). The study was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to their enrollment in the study.

Consent for publication

Not required.

Competing interests

The authors declare no competing interests.