Association between relative fat mass and female infertility among reproductive-aged women from NHANES 2013-2020.

Our study included a total of 3,489 participants, of whom 468 (13.41%) were diagnosed with female infertility, consistent with previous research findings. The age range of the population was 18–45 years, with a mean age of 32.01 ± 0.20 years. The weighted analysis comparing the baseline characteristics between the infertility and fertility groups is presented in Table  1 . Compared to the fertility group, women in the infertility group were older (34.50 ± 0.50 years vs. 31.60 ± 0.19 years) and had higher BMI and WC. Additionally, they exhibited higher levels of UA and lower levels of high-density lipoprotein HDL. Significant differences were also observed between the groups in terms of marital status, smoking habits, DM, hypertension, history of PID, and pregnancy history. Apart from these, no significant differences were found for other potential confounding variables between the groups. Most importantly, the RFM was significantly higher in the infertility group compared to the fertility group (42.73 ± 0.42 vs. 40.41 ± 0.20, P  < 0.001). First, we included BMI, WC, and RFM separately in our multivariable regression models to compare their associations with infertility. In fully adjusted Model, the ORs were as follows, all of which were statistically significant: BMI: 1.020 (95% CI: 1.002, 1.039; p  = 0.031), WC: 1.014 (95% CI: 1.006, 1.022; p  = 0.002), and RFM: 1.039 (95% CI: 1.010, 1.068; p  = 0.009). The results indicated that each unit increase in RFM was more strongly associated with female infertility. Therefore, we further categorized RFM into quartiles. Using the first quartile (Q1) as the reference, the odds ratios for Q2, Q3, and Q4 showed increased risks of female infertility: Q2: OR = 1.66 (95% CI: 1.05, 2.64), Q3: OR = 1.79 (95% CI: 1.16, 2.74), Q4: OR = 2.23 (95% CI: 1.38, 3.60). Additionally, the trend test indicated statistical significance, further supporting the observed associations. All detailed regression analysis results, including those from other models showing statistical significance, are presented in Table  2 . The dose-response relationship in fully adjusted Model between increasing RFM and elevated risk of female infertility is illustrated in Fig.  2 , with a p-value for the linear relationship being less than 0.05. Table 2 Multivariable regression analysis results for BMI, WC, and RFM with female infertility, weighted. Model 1 Model 2 Model 3 OR (95%CI) P value OR (95%CI) P value OR (95%CI) P value BMI 1.034(1.017,1.051) < 0.001 1.032(1.013,1.051) 0.001 1.020(1.002,1.039) 0.031 WC 1.019(1.012,1.026) < 0.0001 1.017(1.009,1.026) < 0.0001 1.014(1.006,1.022) 0.002 RFM 1.058(1.034,1.083) < 0.0001 1.053(1.025,1.082) < 0.001 1.039(1.010,1.068) 0.009 RFMQ Q1 (< 36.27) Ref Ref Ref Ref Ref Ref Q2 (36.27–41.38) 2.104(1.364,3.245) 0.001 1.748(1.105,2.765) 0.018 1.66(1.05,2.64) 0.03 Q3 (41.38–46.25) 2.299(1.573,3.358) 46.25) 3.065(2.004,4.690) < 0.0001 2.720(1.697,4.359) < 0.0001 2.23(1.38,3.60) 0.002 P for trend < 0.0001 < 0.001 0.003 BMI: Body Mass Index, WC: Waist Circumference, RFM: Relative Fat Mass, UA: Uric Acid, TC: Total Cholesterol, HDL: High-Density Lipoprotein, PIR: Poverty-to-Income Ratio, DM: Diabetes Mellitus, PID: Pelvic Inflammatory Disease, RFMQ: Relative Fat Mass Quartiles, OR: Odds Ratio, 95% CI: 95% Confidence Interval. Model 1: Unadjusted; Model 2: Adjusted for age, race, education, marital status, and PIR; Model 3: Adjusted for age, race, education, marital status, PIR, UA, TC, HDL, smoking, alcohol consumption, moderate and vigorous activity, hypertension, DM, age at menarche, menstrual regularity, use of female hormones, PID, and pregnancy history. Multivariable regression analysis results for BMI, WC, and RFM with female infertility, weighted. BMI: Body Mass Index, WC: Waist Circumference, RFM: Relative Fat Mass, UA: Uric Acid, TC: Total Cholesterol, HDL: High-Density Lipoprotein, PIR: Poverty-to-Income Ratio, DM: Diabetes Mellitus, PID: Pelvic Inflammatory Disease, RFMQ: Relative Fat Mass Quartiles, OR: Odds Ratio, 95% CI: 95% Confidence Interval. Model 1: Unadjusted; Model 2: Adjusted for age, race, education, marital status, and PIR; Model 3: Adjusted for age, race, education, marital status, PIR, UA, TC, HDL, smoking, alcohol consumption, moderate and vigorous activity, hypertension, DM, age at menarche, menstrual regularity, use of female hormones, PID, and pregnancy history. Fig. 2 Dose-response relationship between RFM and female infertility risk from the RCS analysis. RCS regression was adjusted for age, race, education, marital status, and PIR. Model 3: Adjusted for age, race, education, marital status, PIR, UA, TC, HDL, smoking, alcohol consumption, moderate and vigorous activity, hypertension, DM, age at menarche, menstrual regularity, use of female hormones, PID, and pregnancy history. The red solid line represents ORs, and the red-shaded region represents the 95% CI. RFM: Relative Fat Mass, RCS: restricted cubic spline, PIR: Poverty-to-Income Ratio, UA: Uric Acid, TC: Total Cholesterol, HDL: High-Density Lipoprotein, DM: Diabetes Mellitus, PID: Pelvic Inflammatory Disease, OR: Odds Ratio, 95% CI: 95% Confidence Interval. Dose-response relationship between RFM and female infertility risk from the RCS analysis. RCS regression was adjusted for age, race, education, marital status, and PIR. Model 3: Adjusted for age, race, education, marital status, PIR, UA, TC, HDL, smoking, alcohol consumption, moderate and vigorous activity, hypertension, DM, age at menarche, menstrual regularity, use of female hormones, PID, and pregnancy history. The red solid line represents ORs, and the red-shaded region represents the 95% CI. RFM: Relative Fat Mass, RCS: restricted cubic spline, PIR: Poverty-to-Income Ratio, UA: Uric Acid, TC: Total Cholesterol, HDL: High-Density Lipoprotein, DM: Diabetes Mellitus, PID: Pelvic Inflammatory Disease, OR: Odds Ratio, 95% CI: 95% Confidence Interval. Subgroup analysis results, based on fully adjusted Model, included predefined stratification factors such as age, race, PIR, education level, BMI, smoking status, hypertension, diabetes, age at menarche, and history of PID. First, the subgroup analysis treating RFM as a continuous variable is illustrated in Fig.  3 . A significant interaction was observed between RFM as a continuous variable and DM ( p  = 0.01), suggesting that the association between RFM and infertility may vary by diabetes status. No significant interactions were detected for other subgroup variables. Table  3 demonstrated that higher RFM (Q3/Q4) is generally associated with increased infertility risk, with variations across subgroups. The association was stronger in younger women (< 35 years, OR = 2.98, 95% CI: 1.03–8.58, p  = 0.04), highly educated individuals (above high school, OR = 4.61, 95% CI: 1.81–11.76, p  = 0.002), and those with higher socioeconomic status (PIR ≥ 3.5, OR = 3.39, 95% CI: 0.83–13.83, p  = 0.07). The association between RFM and infertility varied by smoking status. Notably, former smokers showed a stronger association (Q2: OR = 5.31, 95% CI: 2.87–8.60, Q4: OR = 3.00, 95% CI: 1.72–8.06), while no clear pattern was observed among current or never smokers. Among individuals with DM, higher RFM was associated with an elevated risk of infertility, although the confidence intervals were wider, indicating potential variability in the estimates. Regarding PID, the association between RFM and infertility remained significant in those without a history of PID (Q4 vs. Q1, OR = 2.48, 95% CI: 1.20–5.10), while the effect was attenuated among individuals with prior PID, possibly due to the dominant role of infection-related infertility in this group. These findings highlight the importance of considering individual metabolic, inflammatory, and reproductive factors when evaluating the impact of RFM on infertility. Fig. 3 Subgroup analysis results of RFM as a continuous variable with female infertility risk. Analyses were adjusted for age, race, education, marital status, PIR, UA, TC, HDL, smoking, alcohol consumption, moderate and vigorous activity, hypertension, DM, age at menarche, menstrual regularity, use of female hormones, PID, and pregnancy history, except for the grouping variable itself. RFM: Relative Fat Mass, PIR: Poverty-to-Income Ratio, UA: Uric Acid, TC: Total Cholesterol, HDL: High-Density Lipoprotein, DM: Diabetes Mellitus, PID: Pelvic Inflammatory Disease, OR: Odds Ratio, 95% CI: 95% Confidence Interval. Subgroup analysis results of RFM as a continuous variable with female infertility risk. Analyses were adjusted for age, race, education, marital status, PIR, UA, TC, HDL, smoking, alcohol consumption, moderate and vigorous activity, hypertension, DM, age at menarche, menstrual regularity, use of female hormones, PID, and pregnancy history, except for the grouping variable itself. RFM: Relative Fat Mass, PIR: Poverty-to-Income Ratio, UA: Uric Acid, TC: Total Cholesterol, HDL: High-Density Lipoprotein, DM: Diabetes Mellitus, PID: Pelvic Inflammatory Disease, OR: Odds Ratio, 95% CI: 95% Confidence Interval. Table 3 Subgroup analysis of RFM in quartiles with female infertility risk in fully adjusted model 3, weighted. Characteristics Q1 ( 46.25) P for trend P for interaction Age 0.07 = 35y Ref 1.29(0.70,2.39) 1.27(0.56,2.85) 1.50(0.54,4.19) 0.48 Race 0.06 Non-Hispanic White Ref 1.98(1.05,3.73) 2.15(0.95,4.88) 2.42(0.92,6.36) 0.08 Non-Hispanic Black Ref 0.65(0.25,1.66) 1.22(0.42,3.51) 2.03(0.51,8.02) 0.23 Mexican American Ref 0.72(0.25, 2.07) 0.90(0.21, 3.85) 2.16(0.39,3.07) 0.23 Other Hispanic Ref 3.55(0.67, 8.70) 1.53(0.10, 2.26) 3.36(0.13,4.53) 0.52 Other races Ref 1.46(0.65, 3.27) 4.66(1.16, 18.75) 1.67(0.13, 2.88) 0.11 PIR 0.24 PIR ≤ 1.3 Ref 0.79(0.36,1.73) 1.08(0.45,2.60) 1.29(0.43,3.88) 0.55 1.3 < PIR < 3.5 Ref 3.00(1.17, 7.70) 2.17(0.69, 6.77) 2.02(0.43, 9.47) 0.46 PIR ≥ 3.5 Ref 1.94(0.96, 3.91) 2.27(0.88, 5.83) 3.39(0.83,13.83) 0.07 Educational level 0.07 Below high school Ref 1.40(0.41, 4.78) 0.28(0.05, 1.52) 0.56(0.11, 2.98) 0.16 High school Ref 1.34(0.54, 3.29) 1.40(0.41, 4.75) 0.68(0.13, 3.39) 0.71 Above high school Ref 1.86(1.03, 3.34) 2.50(1.33, 4.70) 4.61(1.81,11.76) 0.002 Smoking 0.11 Never Ref 1.16(0.59,2.28) 1.52(0.73,3.15) 1.45(0.47,4.47) 0.41 Former Ref 5.31(2.87, 8.60) 2.73(1.00, 5.60) 3.00(1.72,8.06) < 0.001 Now Ref 1.33(0.61, 2.92) 1.04(0.43, 2.53) 1.27(0.36, 4.41) 0.86 History of diabetes 0.10 No Ref 1.62(0.98,2.69) 1.65(0.93,2.94) 1.93(0.81,4.58) 0.12 Borderline Ref 2.47 (2.23, 5.9) 1.11 (1.08,8.62) 1.93 (1.80,3.44) 0.05 Yes Ref 3.44(3.10,6.35) 3.92(2.66, 5.53) 2.78(2.64,5.48) 0.83 History of hypertension 0.98 No Ref 1.57(0.93,2.65) 1.52(0.78,2.96) 1.78(0.71,4.50) 0.23 Yes Ref 1.74(0.45, 6.68) 2.66(0.64,10.97) 4.50(0.96,21.05) 0.04 Age of menarche, % 0.08 Younger than 10y Ref 0.12(0.00,0.27) 4.09(1.95,8.05) 3.43(0.87,7.85) 0.01 10y- 12y Ref 1.95(0.84,4.53) 1.53(0.68,3.45) 2.22(0.73,6.77) 0.26 13y- 15y Ref 1.70(0.84,3.44) 1.80(0.74,4.36) 2.25(0.66,7.63) 0.17 16y and older Ref 4.29(0.44,6.02) 0.91(0.00,1.22) 6.42(0.36,9.40) 0.43 Pelvic infection, % 0.70 No Ref 1.68(1.04,2.70) 1.85(1.05,3.25) 2.48(1.20,5.10) 0.02 Yes Ref 0.54(0.06, 4.97) 0.25(0.01, 6.10) 0.03(0.00, 1.97) 0.27 RFMQ: Relative Fat Mass Quartiles, PIR: Poverty-to-Income Ratio, UA: Uric Acid, TC: Total Cholesterol, HDL: High-Density Lipoprotein, DM: Diabetes Mellitus, PID: Pelvic Inflammatory Disease, OR: Odds Ratio, 95% CI: 95% Confidence Interval. Subgroup analyses are adjusted for age, race, education, marital status, PIR, UA, TC, HDL, smoking, alcohol consumption, moderate and vigorous activity, hypertension, DM, age at menarche, menstrual regularity, use of female hormones, PID, and pregnancy history, except for the grouping variable itself. Subgroup analysis of RFM in quartiles with female infertility risk in fully adjusted model 3, weighted. RFMQ: Relative Fat Mass Quartiles, PIR: Poverty-to-Income Ratio, UA: Uric Acid, TC: Total Cholesterol, HDL: High-Density Lipoprotein, DM: Diabetes Mellitus, PID: Pelvic Inflammatory Disease, OR: Odds Ratio, 95% CI: 95% Confidence Interval. Subgroup analyses are adjusted for age, race, education, marital status, PIR, UA, TC, HDL, smoking, alcohol consumption, moderate and vigorous activity, hypertension, DM, age at menarche, menstrual regularity, use of female hormones, PID, and pregnancy history, except for the grouping variable itself.

The NHANES is a perpetual program strictly and meticulously overseen by the National Center for Health Statistics (NCHS), a division of the Centers for Disease Control and Prevention (CDC). It uses a complex, multistage probability sampling method to collect nutrition and health-related information on the U.S. noninstitutionalized population and makes this data available to researchers worldwide to study various factors affecting health. Since 1999, NHANES has been conducted on a 2-year cycle, each cycle sampling approximately 10,000 individuals to gather demographic information, nutritional information, physical examination, laboratory tests, and medical history. All study designs are reviewed and approved by the NCHS Research Ethics Review Board, and all participants provide informed consent, acknowledging the study’s purpose and potential risks before participation. Based on the study’s objectives, the NHANES 2013–2020 cycles were extracted, comprising a total of 44,960 participants. The final eligible sample was selected using the following exclusion criteria 1 : Excluded all males ( n = 22173) 2 ; Excluded females not within the reproductive age range of 18–45 years ( n = 16285) 3 ; Excluded participants who had undergone bilateral oophorectomy and hysterectomy ( n = 214) 4 ; Excluded participants who did not provide infertility history ( n = 978) 5 ; Excluded participants who did not complete measurements for BMI, WC, and height ( n = 210) 6 ; Excluded participants with missing data on potential covariates ( n  = 1611). Ultimately, a sample of 3489 female participants, comprising 468 with infertility and 3021 without infertility, was selected for the final analysis, with detailed exclusion criteria and the selection process illustrated in Fig.  1 . Fig. 1 Flowchart showing participant selection process and exclusion criteria. Flowchart showing participant selection process and exclusion criteria. The exposure variable, RFM, is calculated using the formula for women: RFM = 76-(20×height/WC) 18 . In this formula, height and WC are measured by trained health technicians at the mobile examination centers (MEC). Detailed measurement methods can be accessed on the NHANES website. Weight is measured in kilograms (kg), while height and WC are measured in centimeters (cm). All measurements are precise to one decimal place (0.1). Infertility in this study was assessed using reproductive health data from the NHANES database. Female participants were evaluated based on their responses to two questions: RHQ074 and RHQ076. RHQ074 asked whether the participant had attempted to conceive for at least one year without success, while RHQ076 inquired whether the participant had ever consulted a doctor or other healthcare provider due to their inability to conceive. If participants answered “yes” to either of the above questions, they were classified as infertile. The reference group consisted of women who did not self-report infertility. Given the nature of NHANES data collection, this group may include individuals who have not been clinically diagnosed with infertility but may still experience difficulties conceiving. This classification is consistent with previous NHANES-based infertility studies that rely on self-reported infertility status 22 , 23 . Our analyses adjusted for a range of covariates identified from prior research 24 – 26 , aiming to accurately elucidate this relationship. The selected demographic characteristics include age (< 35y and ≥ 35y), race (non-Hispanic white, non-Hispanic black, Mexican American, other Hispanic, and other races), educational level (below high school, high school, and above high school), marital status (never married, living alone, and living with partner), and family poverty-to-income ratio (PIR) (≤ 1.3, 1.3–3.5, and ≥ 3.5). Health indicators considered include the presence of moderate and vigorous physical activity. Alcohol consumption was defined as either present or absent based on whether participants consumed at least 12 alcoholic drinks in the past year. Smoking status was classified as never ( 100 cigarettes/lifetime, but not currently smoking), or current (> 100 cigarettes/lifetime and still smoking). Additionally, two important comorbidities, hypertension and diabetes, were incorporated into the analysis. Hypertension was defined as either a previous diagnosis, current use of antihypertensive medication, or multiple blood pressure measurements at the MEC indicating levels ≥ 140/90 mmHg. Diabetes mellitus (DM) was categorized as present, borderline, or absent. The criteria for the present category included participants with a previous diagnosis, current use of antidiabetic medication or insulin, a fasting glucose level > 126 mg/dl, or a plasma glucose level ≥ 200 mg/dl at 2 h post-oral glucose tolerance test (OGTT). Impaired fasting glucose and impaired glucose tolerance were classified as borderline. Participants who did not meet these criteria were considered to have no DM. Several covariates related to female reproductive health should also be considered in the analysis, including age at menarche, menstrual regularity, use of female hormone medication, history of pelvic inflammatory disease (PID), and pregnancy history. These variables were assessed through a reproductive health questionnaire, and detailed classification criteria can be found in Table  1 . Additionally, several hematological indicators that may affect female infertility were also considered in our study, including uric acid (UA), total cholesterol (TC), and high-density lipoprotein (HDL). Table 1 Baseline characteristics of the study population by infertility status, weighted. Characteristics Total participants Diagnosis of infertility P value No Yes Number 3489 3021 468 Age, years 32.01 ± 0.20 31.60 ± 0.19 34.50 ± 0.50 < 0.0001 BMI, kg/m 2 29.47 ± 0.23 29.09 ± 0.25 31.79 ± 0.64 < 0.001 WC, cm 95.60 ± 0.51 94.57 ± 0.55 101.90 ± 1.35 < 0.0001 Height, cm 162.73 ± 0.18 162.56 ± 0.20 163.79 ± 0.39 0.01 RFM 40.74 ± 0.19 40.41 ± 0.20 42.73 ± 0.42 < 0.0001 UA, mg/dL 4.55 ± 0.03 4.52 ± 0.03 4.76 ± 0.08 0.004 TC, mg/dL 178.88 ± 0.99 178.55 ± 1.06 180.88 ± 2.02 0.27 HDL, mg/dL 57.05 ± 0.38 57.41 ± 0.40 54.88 ± 1.03 0.03 Educational level, % 0.41 Below high school 8.98 9.00 8.88 High school 19.05 18.59 21.85 Above high school 71.97 72.41 69.28 Marital status, % < 0.0001 Never married 32.23 34.95 15.49 Living alone 8.79 8.54 10.32 Living with partner 58.99 56.51 74.19 Age group, % < 0.001 = 35y 39.60 37.65 51.52 BMI group, % < 0.001 = 30 kg/m 2 39.92 37.79 52.97 Race, % 0.26 Non-Hispanic White 57.74 57.03 62.10 Non-Hispanic Black 12.73 12.87 11.83 Mexican American 11.57 11.63 11.15 Other Hispanic 7.38 7.65 5.67 Other races 10.60 10.81 9.25 PIR, % 0.19 PIR ≤ 1.3 30.57 31.26 26.31 1.3 < PIR < 3.5 32.32 32.09 33.76 PIR ≥ 3.5 37.11 36.65 39.93 Alcohol intake, % 0.94 No 14.62 14.59 14.76 Yes 85.38 85.41 85.24 Smoking, % 0.03 Never 68.84 69.95 62.00 Former 13.03 12.49 16.36 Now 18.13 17.56 21.64 Vigorous activity 0.57 No 62.21 61.96 63.75 Yes 37.79 38.04 36.25 Moderate activity 0.54 No 46.74 46.43 48.64 Yes 53.26 53.57 51.36 History of diabetes, % 0.02 No 89.24 89.58 87.11 Borderline 4.90 5.05 4.00 Yes 5.86 5.36 8.88 History of hypertension, % < 0.001 No 87.13 88.27 80.14 Yes 12.87 11.73 19.86 Age of menarche, % 0.15 Younger than 10 4.15 3.84 6.01 10–12 47.03 46.89 47.90 13–15 42.45 42.73 40.71 16 and older 6.38 6.54 5.38 Regular menstrual periods, % 0.09 No 6.50 6.03 9.43 Yes 93.50 93.97 90.57 Female hormones taken, % 0.05 No 96.21 96.72 93.08 Yes 3.79 3.28 6.92 Pelvic infection, % < 0.001 No 95.99 96.63 92.04 Yes 4.01 3.37 7.96 Ever pregnant, % < 0.0001 No 34.63 37.53 16.81 Yes 65.37 62.47 83.19 BMI: Body Mass Index, WC: Waist Circumference, RFM: Relative Fat Mass, UA: Uric Acid, TC: Total Cholesterol, HDL: High-Density Lipoprotein, PIR: Poverty-to-Income Ratio. Continuous variables are presented as weighted means with standard errors and compared between groups using weighted linear regression. Categorical variables are presented as weighted percentages and compared between groups using weighted chi-square tests. Baseline characteristics of the study population by infertility status, weighted. BMI: Body Mass Index, WC: Waist Circumference, RFM: Relative Fat Mass, UA: Uric Acid, TC: Total Cholesterol, HDL: High-Density Lipoprotein, PIR: Poverty-to-Income Ratio. Continuous variables are presented as weighted means with standard errors and compared between groups using weighted linear regression. Categorical variables are presented as weighted percentages and compared between groups using weighted chi-square tests. In all statistical analyses, we strictly adhered to the CDC guidelines for NHANES statistical analyses. This necessitated the application of appropriate sample weights provided by NHANES for each participant to account for the complex sampling design. MEC weights were chosen for our analysis, based on our study objectives. Descriptive statistics were employed to present the baseline characteristics of the study population, with continuous variables expressed as weighted means and standard errors, and categorical variables represented as weighted percentages. To compare baseline characteristics between the infertility and fertility groups, continuous variables were analyzed using survey-weighted linear regression, while categorical variables were evaluated using survey-weighted Chi-square tests. Survey-weighted multivariable logistic regression analyses were utilized to explore the relationships between BMI, WC, and RFM with infertility, incorporating three models: Model 1 (unadjusted Model), Model 2 (minimally adjusted Model), and Model 3 (fully adjusted Model). In these analyses, Model 1 refers to the unadjusted basic model, which includes only the primary exposure variables without adjusting for any other factors. Model 2, known as the minimally adjusted model, accounts for key demographic characteristics, including age, race, education, marital status, and PIR. The final Model 3, the fully adjusted model, incorporates all potential confounding variables, including the key demographic characteristics from Model 2, as well as UA, TC, HDL, smoking, alcohol consumption, moderate and vigorous activity, hypertension, DM, and female reproductive health factors. Additionally, RFM was converted into quartiles (Q1 < 36.27, 36.27 ≤ Q2 < 41.38, 41.38 ≤ Q3 < 46.25, and Q4 ≥ 46.25) and the above regression analyses were repeated, with the first quartile serving as the reference group. All regression analysis results are presented as odds ratios (OR) with 95% confidence intervals (CI), where an OR of 1 serves as the threshold indicating the direction of the relationship between the exposure and outcome variables. Moreover, restricted cubic splines (RCS) were employed to further visualize whether a linear relationship exists between RFM and female infertility. After the above analyses were completed, subgroup analyses were conducted to evaluate whether the association between RFM and infertility differs across key demographic and clinical factors, including age, race, PIR, education level, BMI, smoking status, DM, hypertension, age at menarche, and history of PID. These analyses were performed to assess potential effect modification rather than to validate findings from the total sample. These factors were selected based on biological plausibility, prior NHANES-based infertility studies 27 , and established evidence suggesting their role in modulating metabolic and reproductive health. We intentionally limited subgroup analyses to these key variables to minimize the risk of multiple comparisons and ensure a hypothesis-driven approach. Linear regression analysis was used for trend test. Similar to the regression analysis, the subgroup analyses were repeated with RFM categorized into quartiles. Interaction tests were employed to complement the subgroup analyses. During all statistical analyses, a two-sided p-value of less than 0.05 was considered statistically significant. For statistical analyses, we harnessed the capabilities of EmpowerStats ( www.empowerstats.com ; X&Y Solutions, Inc., Boston MA) and R version 4.0.5 ( http://www.R-project.org , The R Foundation).

Our findings demonstrate a significant association between RFM and female infertility in a population-based sample, supporting the hypothesis that excess adiposity may play a role in reproductive health. However, given the cross-sectional design, these results should be interpreted as hypothesis-generating rather than confirmatory. Future longitudinal and mechanistic studies are needed to explore the causal role of adiposity, as well as the clinical utility of RFM in infertility risk assessment.

In this cross-sectional study, we used nationally representative NHANES data and found a positive linear relationship between RFM and the increased risk of infertility in women. Moreover, after adjusting for all covariates, RFM showed a stronger association with female infertility compared to BMI and WC; each unit increase in RFM was more closely related to infertility in women. Even when RFM was divided into categorical variables, this association remained stable. The results of the subgroup analysis further confirmed our findings. Although the observed differences in mean RFM values between infertile and non-infertile women were statistically significant, their absolute magnitude was relatively modest. However, even small increases in body fat percentage can have meaningful effects on hormonal balance, insulin resistance, and systemic inflammation, all of which are implicated in infertility. The odds ratios per unit increase in RFM, while not large, suggest a dose-dependent association where individuals with higher RFM (Q3/Q4) face an elevated risk of infertility. This underscores the potential relevance of adiposity management in reproductive health interventions, particularly in populations at risk for metabolic dysfunction. This study suggested a significant association between RFM and female infertility in a population-based sample. To our knowledge, this is the first study to investigate the association between RFM and female infertility. Previous studies have shown that obesity has negative effects on reproductive health, such as decreased fertility and infertility. Compared to women with normal weight, obese women are more likely to experience early pregnancy miscarriage and congenital defects 28 , 29 . For example, Rich-Edwards et al. found that women with obesity had a 1.8-fold higher risk of infertility compared to those with normal weight 30 , while Ramlau-Hansen et al. observed a nonlinear association between BMI and infertility risk, with stronger effects at higher BMI thresholds 31 . Obesity can affect the endocrine system, thereby impacting ovulation, women with a BMI over 27 kg/m 2 are more likely to experience anovulatory infertility 7 , 32 . A randomized controlled trial with a 5-year follow-up of 577 women with obesity and infertility found that a 6-month lifestyle intervention led to significant weight loss, which was associated with improved sexual function, including higher intercourse frequency, better vaginal lubrication, and increased overall sexual function scores in the intervention group 33 . Similarly, another study showed that weight loss during the periconceptional period in obese and infertile women could increase conception rates, reduce the risk of gestational hypertension, and lower the risk of preterm birth 34 . Additionally, obesity can damage the endometrial environment and tolerance by delaying the window of implantation, leading to metabolic abnormalities, which in turn cause infertility and poor ART outcomes in women 35 . More recent studies have attempted to use WC and waist-to-hip ratio (WHR) as alternative obesity indices, demonstrating that central adiposity may exert stronger effects on reproductive health than general obesity 36 – 38 . For instance, Ke et al. reported that each 5 cm increase in WC was associated with a 15% increase in infertility risk, suggesting that abdominal adiposity plays a crucial role 26 . However, these measures are still influenced by factors such as muscle mass and body frame size, limiting their accuracy. Our study is among the first to utilize RFM, a novel adiposity index that has been shown to better reflect total body fat percentage 18 . By demonstrating a significant association between higher RFM and increased infertility risk, our findings support prior research while offering a more precise measure of adiposity. Additionally, our subgroup analyses highlight potential effect modification by metabolic and reproductive factors, advancing current knowledge on how adiposity-related infertility risk varies across populations. Further research is needed to compare RFM with traditional obesity indices in prospective cohorts to better understand its clinical utility in reproductive health assessments. The potential mechanisms by which obesity affects female infertility are closely related to fat accumulation. Adipokines, primarily derived from adipose tissue, are associated with a range of metabolism-related diseases 39 . Excessive fat can also cause sympathetic nervous system hyperactivity, leading to abnormal secretion of adipokines (such as adiponectin and leptin). This abnormality results in insulin resistance and chronic inflammation, both of which are recognized as independent risk factors for infertility 40 . Additionally, excessive adipose tissue can disrupt the normal function of the hypothalamic-pituitary-gonadal (HPG) axis, which is part of the endocrine system responsible for controlling hormone levels and regulating reproduction. In women, excessive adipose tissue increases the peripheral aromatization of androgens to estrogens, affecting gonadotropin secretion 41 . Obese women also exhibit reduced levels of growth hormone (GH), sex hormone-binding globulin (SHBG), and insulin-like growth factor-binding protein (IGFBP), along with elevated leptin levels. Consequently, the neuro-regulation of the hypothalamic-pituitary-ovarian (HPO) axis may become severely impaired, leading to infertility in women 42 . Therefore, indicators to assess fat distribution may be more useful in evaluating the relationship between obesity and female infertility. BMI remains the most widely used tool for diagnosing obesity 43 . However, BMI cannot distinguish between lean body mass and fat mass, nor can it assess subcutaneous and visceral fat deposition 44 . Increased muscle mass in young or muscular individuals can lead to a higher BMI, even if they are not actually overweight 45 . Due to these limitations, some studies have begun exploring the use of WC as an alternative to BMI for assessing obesity 36 . Nevertheless, WC also cannot differentiate between subcutaneous and visceral fat 46 . Therefore, it is necessary to continue developing new metrics for assessing body fat distribution. RFM is a new body fat measurement index that can more accurately estimate overall body fat percentage compared to traditional measurement methods like BMI and WHR 18 . In a large multiethnic cohort in the United States, RFM has shown stronger correlation with fat mass measured through DEXA (Dual-Energy X-ray Absorptiometry) compared to BMI 18 . Prior studies have already examined DXA-measured body fat percentage in relation to female infertility, confirming that higher adiposity is associated with increased infertility risk 47 . Although NHANES includes DXA-derived body fat percentage as a gold standard measure of adiposity, our study aimed to identify a more accessible and clinically feasible metric for assessing infertility risk. While DXA provides precise body fat estimation, it requires specialized equipment and is not widely available in routine clinical practice. In contrast, RFM is a simple, cost-effective alternative that can be easily applied in large-scale population studies. Currently, RFM has been shown to be associated with various diseases, such as DM, hypertension and coronary artery disease 20 , 48 , 49 . Given the association between fat accumulation and female infertility, RFM, as a new index for assessing body fat distribution, may provide a more detailed reflection of the relationship between obesity and infertility. Our study showed that RFM has a closer association with female infertility compared to BMI and WC. In the future, RFM may serve as a reliable predictive indicator for infertility in obese women, although further prospective studies are warranted. Although RFM, BMI, and WC are all measures of adiposity, they capture different aspects of body composition, and their unit changes are not directly comparable. Previous studies have demonstrated strong correlations between RFM and BMI as well as between RFM and WC, suggesting shared yet distinct contributions to metabolic risk 18 . However, the exact numerical equivalency between unit changes in these measures remains unclear. Given that the observed odds ratios per unit increase in RFM, BMI, and WC are similar (approximately 1.03–1.05), future studies should explore how these indices translate into clinically meaningful differences in body composition and infertility risk. There are several limitations worth considering in this study. Firstly, this study employs a cross-sectional research design, which cannot establish a causal relationship between RFM and female infertility. Secondly, despite the inclusion of numerous covariates, there may still be uncontrollable confounding factors, such as a family history of infertility and sexual intercourse frequency, due to data limitations. Thirdly, relying on self-reported infertility data may introduce recall bias and misclassification bias. Specifically, classifying all women who answered “No” to RHQ074 or RHQ076 as “not self-report infertility” may lead to non-differential misclassification bias, potentially attenuating the observed associations. This classification does not account for women who may have undiagnosed infertility or subfertility but did not self-report these conditions. Additionally, NHANES lacks detailed reproductive history data, preventing the construction of a “high-risk” fertility group that incorporates key factors such as sexual activity, contraceptive use, and relationship status. Unlike datasets such as the National Survey of Family Growth (NSFG), NHANES does not provide sufficient granularity to precisely classify reproductive potential, which may introduce heterogeneity into the comparison group. However, our study methodology aligns with prior NHANES-based infertility research, ensuring consistency in approach and comparability of findings. Moreover, RFM does not differentiate between visceral and subcutaneous fat, both of which have distinct metabolic effects on reproductive health. Lastly, it cannot be determined whether female infertility is the cause of failed pregnancies, the cause of male infertility, or both, as NHANES does not provide information in this regard.

Infertility refers to the inability of a couple to achieve pregnancy after engaging in regular, unprotected sexual activity for 12 months or longer 1 . Presently, it has emerged as a critical global health concern, impacting approximately 15% of the world’s population 2 . An astonishing fact is that infertility rates in some developing countries can reach up to 30% 3 . In these instances, nearly half of the infertility is specifically attributable to female-related issues 4 . Therefore, behind these rising rates lies not only the decline in global birth rates and population numbers but also the psychological burden and marital instability that infertility imposes on couples 5 , 6 . Female infertility is a multifactorial reproductive disorder, involving ovulatory dysfunction, tubal obstruction, cervical factors, endometriosis, uterine pathologies or unexplained infertility, often triggering by genetic, environmental, and lifestyle factors 7 , 8 . Compared to genetic factors, environmental exposures and lifestyle choices are among the most modifiable factors and are the most commonly received and adopted medical advice by these patients. Therefore, early identification of these modifiable risk factors and timely intervention are essential to alleviate the socio-economic burden of infertility and mitigate its adverse effects on couples. Obesity has become a major global public health concern, particularly severe among women, with projections showing it will affect over 21% of women worldwide by 2025 9 , 10 . This has elevated the discussion of obesity and female infertility to a prominent level, as obesity can impair female reproductive health through various mechanisms, including irregular menstruation and endometrial pathology 11 , 12 . Consequently, identifying reliable metrics to accurately assess the detrimental effects of obesity on female infertility has become highly sought after. Evidently, researchers have continuously made efforts, progressing from the classic Body Mass Index (BMI) to waist circumference (WC) and waist-to-height ratio (WHtR), among others 13 – 15 . Despite the simplicity of these metrics, their shortcomings are quite evident, primarily because they cannot accurately distinguish abdominal obesity 16 . Therefore, researchers have gradually shifted their focus to specialized assessments like Dual-Energy X-ray Absorptiometry (DXA), sacrificing simplicity and convenience for greater accuracy 17 . When considering applicability, these findings are difficult to translate from the laboratory to real-world clinical settings, highlighting the urgency of continuing to seek objective and accurate obesity assessment metrics. Recently, relative fat mass (RFM) has emerged as a new gender-specific anthropometric indicator to accurately assess obesity and body fat percentage 18 . When compared to DXA standards, its error rate in classifying obesity is significantly reduced 18 . Therefore, RFM has garnered significant attention and widespread recognition within the academic community for its superior ability to assess the various adverse health impacts of obesity 19 – 21 . However, research on the relationship between RFM and female infertility remains sparse, highlighting a gap that needs further exploration. Thus, our study utilizes data from the National Health and Nutrition Examination Survey (NHANES) to analyze the relationship between RFM and female infertility. Our objective is to evaluate the association between RFM and infertility to improve the understanding of its potential role as a risk factor.