Associations between dietary inflammatory index with menopause and hysterectomy: findings from a national health and nutrition examination survey.

The characteristics of the participants are shown in Table  1 . Among the 9,469 participants aged 20–65 years with complete records of menopausal status and the E-DII, 5,773 women were premenopausal, 2,178 were naturally menopausal, and 1,518 underwent hysterectomy. The median age at the time of the interview and the E-DII score were 43 and 0.60, respectively. After being divided based on quartiles of the E-DII score, participants with higher E-DIIs were younger, smoked, were less likely to have a high school education, were married/cohabiting, engaged in enough leisure-time physical activity, and tended to have greater family income and greater burdens of chronic diseases (diabetes and hypertension). In addition, those patients experience a higher rate of hysterectomy and an earlier age of menopause and hysterectomy. However, there was no significant difference between the four quartiles in terms of parity or age at menarche. Table 1 Baseline characteristics of women 20–65 years of age included in this study according to quartiles of E-DIIa Variable Total ( n  = 9469) Q1 ( n  = 2367) Q2 ( n  = 2367) Q3 ( n  = 2367) Q4 ( n  = 2368) P -value E-DII 0.60 (-0.25 ,1.61) -0.70 (-1.13, -0.40) 0.29 (0.07 ,0.48) 1.15 (0.91 ,1.43) 2.73 (2.14 ,3.76) < 0.001 DII 1.13 (-0.51 ,2.45) -1.59 (-2.51, -0.87) 0.57 (0.16 ,0.93) 1.96 (1.58 ,2.39) 3.28 (2.77 ,3.77) < 0.001 Age at interview (years) 43.00 (31.00 ,54.00) 45.00 (33.00 ,55.00) 43.00 (30.00 ,53.00) 43.00 (30.00 ,53.00) 43.00 (30.00 ,53.00) 0.004 Ethnicity < 0.001  Mexican American 1592 (8.88) 429 (8.61) 422 (9.88) 355 (8.18) 386 (8.89)  Other Hispanic 1124 (5.97) 270 (5.14) 269 (5.23) 280 (6.52) 305 (7.22)  Non-Hispanic White 3677 (65.09) 976 (68.82) 909 (65.24) 955 (64.97) 837 (60.48)  Non-Hispanic Black 2097 (12.41) 365 (7.71) 529 (12.72) 577 (13.86) 626 (16.25)  Other Race 979 (7.65) 327 (9.72) 238 (6.93) 200 (6.47) 214 (7.16) Education < 0.001  Lower than high school 2004 (13.99) 376 (9.22) 471 (13.66) 525 (15.49) 632 (18.60)  High school 1970 (20.40) 380 (14.46) 476 (19.54) 551 (23.84) 563 (24.95)  More than high school 5490 (65.60) 1611 (76.32) 1417 (66.77) 1290 (60.66) 1172 (56.44)  Not recorded 5 (0.01) 0 3 (0.03) 1 (0.01) 1 (0.01) Marital status < 0.001  Married/Cohabited 5413 (62.01) 1473 (67.81) 1394 (63.98) 1294 (59.95) 1252 (54.91)  Widowed/Divorced/Separated 1988 (18.41) 430 (15.41) 458 (17.43) 526 (18.86) 574 (22.69)  Never married 2063 (19.55) 462 (16.71) 515 (18.58) 545 (21.17) 541 (22.38)  Not recorded 5(0.03) 2 (0.07) 0 2 (0.02) 1 (0.02) Family PIR < 0.001   3 3183 (46.17) 1014 (56.92) 840 (47.67) 741 (42.11) 588 (35.67)  Not recorded 724 (5.76) 191 (5.94) 179 (5.53) 166 (5.29) 188 (6.32) Body mass index (kg/m 2 ) < 0.001  Underweight (< 18.5) 184 (1.95) 46 (2.04) 40 (1.68) 54 (2.26) 44 (1.80)  Normal weight (18.5 to < 25) 2655 (31.51) 810 (38.49) 666 (31.21) 587 (27.26) 592 (27.83)  Overweight (25 to  30) 3966 (37.89) 873 (31.40) 1007 (39.51) 1037 (41.00) 1049 (40.74) Smoking 3410 (38.70) 720 (33.46) 822 (35.42) 949 (43.95) 919 (43.12) < 0.001 Physical activity < 0.001  Insufficiently active group 5999 (57.14) 1249 (44.88) 1518 (58.26) 1587 (63.04) 1645 (64.63)  Active group 3470 (42.86) 1118 (55.12) 849 (41.74) 780 (36.96) 723 (35.37) Diabetes 904 (7.08) 167 (5.11) 207 (7.12) 226(7.37) 304 (9.17) 0.002 Hypertension 2713 (25.35) 574 (20.96) 680 (26.75) 685 (26.56) 774 (27.90) 0.001 Hyperlipidemia 3047 (31.70) 743 (31.18) 714 (30.50) 764 (32.24) 826 (33.06) 0.642 Age of Menarche (years) 0.522  < 9 373 (3.36) 85 (3.33) 86 (2.87) 94 (3.50) 108 (3.80)  10 to 12 4199 (43.24) 1038 (43.03) 1102 (44.68) 1048 (43.08) 1011 (42.09)  13 to 15 4213 (46.06) 1090 (46.14) 1021 (45.82) 1065 (46.98) 1037 (45.26)  > 15 684 (7.33) 154 (7.50) 158 (6.63) 160 (6.44) 212 (8.85) Parity (no. of live births) 0.755  0 2262 (27.75) 618 (28.89) 571 (27.05) 576 (27.46) 497 (27.42)  1 1637 (17.83) 414 (17.90) 442 (18.19) 403 (18.57) 378 (16.55)  > 1 5570 (54.42) 1335 (53.21) 1354 (54.75) 1388 (53.97) 1493 (56.04) Menopausal status 0.029  Premenopause 5773 (61.10) 1467 (60.30) 1511 (62.66) 1458 (61.83) 1337 (59.55)  Natural menopause 2178 (23.42) 577 (26.16) 507 (22.50) 528 (22.20) 566 (22.37)  Hysterectomy 1518 (15.48) 323 (13.53) 349 (14.84) 381 (15.97) 465 (18.08) Early natural menopause 497 (7.76) 110 (6.45) 117 (7.45) 119 (7.62) 151 (9.81) 0.072 Age of menopause (years) b 46.00 (39.00 ,50.00) 48.00 (41.00 ,52.00) 45.00 (38.00 ,51.00) 46.00 (39.00 ,50.00) 44.00 (35.00 ,50.00) < 0.001 Age of hysterectomy (years) 40.00 (34.00 ,45.00) 41.00 (36.00 ,47.00) 39.00 (32.00 ,45.00) 40.00 (34.00 ,45.00) 38.00 (32.00 ,44.00) 0.001 DII, dietary inflammatory index; E-DII, energy-adjusted dietary inflammatory index; PIR, poverty-to-income ratio; a Values for categorical and continuous variables with skewed distribution are expressed as n (%) and median (interquartile ranges), respectively; b Values of age included those from both natural menopause and hysterectomy induced menopause; Baseline characteristics of women 20–65 years of age included in this study according to quartiles of E-DIIa DII, dietary inflammatory index; E-DII, energy-adjusted dietary inflammatory index; PIR, poverty-to-income ratio; a Values for categorical and continuous variables with skewed distribution are expressed as n (%) and median (interquartile ranges), respectively; b Values of age included those from both natural menopause and hysterectomy induced menopause; In addition, as shown in Fig.  2 and Supplementary Table 1 , the Spearman analysis results indicated that the E-DII was positively correlated with body mass index and parity but negatively correlated with physical activity, age at menopause and hysterectomy. The correlation coefficients were 0.08, 0.06, -0.14, -0.12 and − 0.11, respectively. Physical activity was positively correlated with both the age of menopause and the number of hysterectomies, and parity was positively correlated with only the age of hysterectomy. The correlation coefficients were 0.05, 0.1 and 0.07, respectively. Most importantly, the age at hysterectomy was strongly correlated with the age at menopause, with a 0.95 correlation coefficient. Fig. 2 The heatmap of the correlation between E-DII and baseline covariables using the Spearman correlation analysis among 9,469 participants. E-DII, energy-adjusted dietary inflammatory index The heatmap of the correlation between E-DII and baseline covariables using the Spearman correlation analysis among 9,469 participants. E-DII, energy-adjusted dietary inflammatory index The associations of the E-DII with natural menopause and hysterectomy are shown in Table  2 . In all three models, there seemed to be no positive results to support the association between the E-DII and natural menopause (all p  > 0.05). Conversely, we found that the E-DII was positively correlated with hysterectomy in all three models. After fully adjusting for age at interview, ethnicity, education, marital status, family PIR, body mass index, smoking, physical activity, diabetes, hypertension, hyperlipidaemia, age of menarche and parity, the OR was 1.487 (95% CI: 1.069, 2.069; p  = 0.023) in quartile 3 versus quartile 1 and 1.729 (95% CI: 1.198, 2.497; p  = 0.005) for quartile 4 versus quartile 1. Furthermore, when we performed a subgroup analysis in two subgroups divided by 45-year-old age, we observed a positive association between the E-DII score and early natural menopause (fully adjusted OR in quartile 4 versus quartile 1: 1.552, 95% CI: 1.043, 2.309; p  = 0.035) but not later natural menopause (fully adjusted OR in quartile 4 versus quartile 1: 1.328, 95% CI: 0.767, 2.301; p  = 0.316), as shown in Table  3 . Table 2 Associations of E-DII with natural menopause and hysterectomy Variable Odds ratio (95% CI), P -value Model I a Model II b Model III c Natural menopause Quartile 1 Reference Reference Reference Quartile 2 0.828 (0.645, 1.062) 0.141 1.175 (0.866, 1.594) 0.304 1.190 (0.871, 1.625) 0.268 Quartile 3 0.828 (0.662, 1.034) 0.100 1.160 (0.886, 1.519) 0.284 1.118 (0.852, 1.468) 0.413 Quartile 4 0.866 (0.695, 1.078) 0.200 1.236 (0.887, 1.721) 0.214 1.227 (0.875, 1.722) 0.229 P for trend 0.180 0.220 0.268 Hysterectomy Quartile 1 Reference Reference Reference Quartile 2 1.055 (0.824, 1.352) 0.672 1.369 (0.925, 2.025) 0.114 1.223 (0.853, 1.753) 0.278 Quartile 3 1.151 (0.900, 1.472) 0.267 1.623 (1.166 , 2.258) 0.005 1.487 (1.069 , 2.069) 0.023 Quartile 4 1.353 (1.087 , 1.684) 0.008 1.875 (1.295 , 2.715) 0.001 1.729 (1.198 , 2.497) 0.005 P for trend 0.006 < 0.001 0.002 E-DII, energy-adjusted dietary inflammatory index. CI, confidence interval a Crude model; b Adjusted for age at interview, ethnicity, education, marital status and family PIR; c Adjusted for age at interview, ethnicity, education, marital status, family PIR, body mass index, smoking, physical activity, diabetes, hypertension, hyperlipidemia, age of menarche and parity Associations of E-DII with natural menopause and hysterectomy E-DII, energy-adjusted dietary inflammatory index. CI, confidence interval a Crude model; b Adjusted for age at interview, ethnicity, education, marital status and family PIR; c Adjusted for age at interview, ethnicity, education, marital status, family PIR, body mass index, smoking, physical activity, diabetes, hypertension, hyperlipidemia, age of menarche and parity Table 3 Associations between E-DII and natural menopause in two subgroups divided by 45-year-old age Variable Odds ratio (95% CI), P -value Model I a Model II b Model III c Early natural menopause Quartile 1 Reference Reference Reference Quartile 2 1.168 (0.865, 1.576) 0.314 1.365 (0.968, 1.924) 0.081 1.335 (0.920, 1.938) 0.134 Quartile 3 1.195 (0.808, 1.768) 0.375 1.372 (0.928, 2.027) 0.118 1.258 (0.864, 1.831) 0.236 Quartile 4 1.577 (1.098 , 2.266) 0.016 1.587 (1.066 , 2.362) 0.026 1.552 (1.043 , 2.309) 0.035 P for trend 0.037 0.044 0.063 Later natural menopause Quartile 1 Reference Reference Reference Quartile 2 0.767 (0.579, 1.016) 0.068 1.353 (0.787, 2.329) 0.278 1.422 (0.810, 2.496) 0.226 Quartile 3 0.756 (0.595 , 0.906) 0.025 1.402 (0.915, 2.146) 0.125 1.398 (0.888, 2.202) 0.154 Quartile 4 0.723 (0.575 , 0.909) 0.007 1.362 (0.794, 2.336) 0.266 1.328 (0.767, 2.301) 0.316 P for trend 0.006 0.203 0.264 E-DII, energy-adjusted dietary inflammatory index. CI, confidence interval a Crude model; b Adjusted for age at interview, ethnicity, education, marital status and family PIR; c Adjusted for age at interview, ethnicity, education, marital status, family PIR, body mass index, smoking, physical activity, diabetes, hypertension, hyperlipidemia, age of menarche and parity Associations between E-DII and natural menopause in two subgroups divided by 45-year-old age E-DII, energy-adjusted dietary inflammatory index. CI, confidence interval a Crude model; b Adjusted for age at interview, ethnicity, education, marital status and family PIR; c Adjusted for age at interview, ethnicity, education, marital status, family PIR, body mass index, smoking, physical activity, diabetes, hypertension, hyperlipidemia, age of menarche and parity

The NHANES is an ongoing national cross-sectional survey that is being conducted to collect data on the health and nutritional status of the U.S. population. The study protocols were ratified by the Centers for Disease Control and Prevention, and all participants provided informed consent. Details regarding the interview, examination, and sample collection protocols are available at www.cdc.gov/nchs/nhanes . Our main outcomes of interest were the E-DII score and menopausal status. First, the total number of participants in the 2007–2016 NHANES cycle was 50,588. After excluding subjects with unknown menopausal status ( n  = 35,029), age  65 years ( n  = 5,652), and unusually low or high total energy intake ( 5000 kcal/day) ( n  = 438), a total of 9,469 subjects were enrolled in the final analysis (Fig.  1 ). Fig. 1 Study flowchart. Flow chart showing the process of participants selection. Of 50,558 participants from 2007 to 2016 of National Health and Nutrition Examination Survey (NHANES), 9,469 remained in the final analysis Study flowchart. Flow chart showing the process of participants selection. Of 50,558 participants from 2007 to 2016 of National Health and Nutrition Examination Survey (NHANES), 9,469 remained in the final analysis In accordance with the NHANES protocol, dietary intake was documented via a 24-h dietary history interview at the Mobile Examination Center and validated by the Nutrition Methodology Working Group. In the present study, 27 of the 45 food parameters were available to calculate DII, including energy, carbohydrates, fats, protein, fibre, alcohol, β-carotene, cholesterol, n-3 fatty acid, n-6 fatty acid, polyunsaturated fatty acid, monounsaturated fatty acid, saturated fat, thiamine, magnesium, zinc, selenium, iron, vitamin A, vitamin B6, vitamin B12, vitamin C, vitamin D, vitamin E, riboflavin, folic acid, caffeine, and niacin. The DII calculation formula and the whole calculation process have been reported in numerous studies previously [ 15 , 17 ]. In brief, each value of the 27 parameters above was standardized after the global daily mean intake was subtracted and divided by its standard deviation, both of which are available in N. Shivappa’s paper [ 15 ]. To control for the effect of total energy intake, we used the E-DII, in which we calculated all the food parameters per 1,000 kcal of consumption. In addition, we planned to exclude participants with unusually low or high total energy intake ( 5000 kcal/day). The E-DII score was subsequently used to categorize variables into quartiles from the total sample for analysis. Menopausal status was determined via a two-question reproductive health questionnaire. Women were asked “Have you had at least one menstrual period in the past 12 months? (Please do not include bleeding caused by medical conditions, hormone therapy, or surgeries).” Women who answered “no” were subsequently asked “What is the reason that you have not had a period in the past 12 months?” Participants were considered premenopausal if the answer to the first question was “Yes” or if the answer to the first question was “No” and the answer to the second question was “Pregnancy”, “Breast feeding”, “Medical conditions/treatments”, and/or “Other”. Otherwise, women were classified as postmenopausal if they answered “no” to the first question and indicated the reason to be natural menopause or hysterectomy. In addition, for women who reported having gone through menopause or hysterectomy, age at occurrence was recorded. A number of covariates, such as age of interview, ethnicity, education, marital status, family poverty-to-income ratio (PIR), body mass index, smoking, physical activity, diabetes, hypertension, hyperlipidaemia, age of menarche and parity, were included as possible confounders in the association between the E-DII and menopause. Current and former smokers were included on the basis of the NHANES Questionnaire ‘Smoking-Cigarette Use’ section. A family PIR  3 means that household income is more than triple the poverty threshold. Participants with body mass index < 18.5, 18.5 to < 25, 25 to  30 were considered underweight, normal weight, overweight, and obese, respectively, according to the World Health Organization standards. Physical activity was categorized into insufficiently active and active groups (≥ 150 min/week of moderate or/and ≥ 75 min/week of vigorous leisure-time physical activity) [ 18 ]. Diabetes, hypertension and hyperlipidaemia were determined by those who were ever told they had them by a doctor. All analyses incorporated sample weights, strata, and primary sampling units to produce accurate national estimates. The sample baseline characteristics of all included participants were divided into quartiles according to the E-DII score, with the nonnormally continuous variables reported as the median (interquartile) and the categorical variables reported as numbers with percentages. The differences among the four groups were compared via the Kruskal‒Wallis test (nonnormally continuous variables) and χ2 test (categorical variables). We subsequently constructed three models via single and multivariate generalized linear models with different adjustments to calculate odds ratios (ORs) and corresponding 95% confidence intervals (CIs), aiming to test the associations of the E-DII with natural menopause and hysterectomy. Model I was not adjusted for any confounders, and Model II was adjusted for age at interview, ethnicity, education, marital status, and family PIR. Model III was fully adjusted for age at interview, ethnicity, education, marital status, family PIR, body mass index, smoking, physical activity, diabetes, hypertension, hyperlipidaemia, age of menarche and parity. To further evaluate the effect of the E-DII on early natural menopause (defined as a woman aged < 45 years) [ 7 ], subgroup analyses were performed. In addition, Spearman correlation analysis was performed to explore the bivariate associations between the E-DII and age at menopause and hysterectomy.

Menopause is a clinical diagnosis that is defined as 12 months after the final menstrual period; it may be spontaneous (natural menopause) or secondary menopause, including removal of the bilateral ovaries or uterus. Additionally, various genetic, environmental and lifestyle factors, such as chemical exposure, parity, the use of oral contraceptives, current smoking habits and dietary components, are related to the onset of menopause [ 1 – 3 ]. Clearly, the occurrence of natural menopause is inevitable for women in their lifetime. However, for such women at menopause, the decline in oestrogen can lead to changes throughout the body, including bone loss, increased abdominal fat and adverse cardiovascular risks. Previous studies have indicated that early menopause and hysterectomy are associated with increased cardiovascular and all-cause mortality [ 4 , 5 ]. Generally, the average age at menopause is 49 to 52 years [ 6 , 7 ]. Early menopause often occurs between the ages of 40 and 45 [ 7 ]. Additionally, the prevalence of hysterectomy worldwide may be approximately 20% in middle-aged women according to incomplete statistics [ 8 – 10 ]. The incidence of hysterectomy increases with age and is highest among women aged above 70 years (estimated at over 45%) [ 10 ]. Although not all women experience significant symptoms, early menopause and hysterectomy have been proven to be related to frailty, a complex pathophysiological phenomenon that contributes to the risk of several adverse health outcomes [ 11 ]. Some evidence suggests that inflammation mediates the onset of menopause among U.S. adults [ 12 ]. Recently, it was reported that a higher intake of green and yellow vegetables was associated with an earlier age of menopause, whereas higher intakes of some dairy products, such as low-fat milk, skim milk, and low alcohol, were associated with a later menopause onset [ 3 ]. In other words, diet may be a modifiable factor closely related to inflammation and menopause. To clarify the role of diet and inflammation in human health, the dietary inflammatory index (DII) was developed to assess inflammation in the diet [ 13 , 14 ]. Forty-five food parameters were included to calculate the DII and were associated with 6 inflammatory biomarkers: interleukin (IL)-1β, IL-4, IL-6, IL-10, TNF-α, and C-reactive protein [ 15 ]. An elevated DII score suggested a diet with proinflammatory effects, whereas a lower negative DII score indicated a diet that was anti-inflammatory. Moreover, the energy-adjusted DII (E-DII) has been shown to improve the prediction of many diseases in comparison with the DII [ 16 ]. Nevertheless, there are no reports regarding the associations of the DII or E-DII with menopause and hysterectomy. Therefore, to fill this gap, the aim of this cross-sectional study was to investigate the associations of the E-DII with menopause and hysterectomy using data from the National Health and Nutrition Examination Survey (NHANES). Additionally, we explored the correlations between the E-DII and the ages of menopause and hysterectomy patients.

In conclusion, this study revealed that an elevated E-DII was associated with an increased prevalence of early menopause and hysterectomy. A higher E-DII was inversely correlated with the age at both menopause and hysterectomy. In brief, we recommend that a proinflammatory diet be established as a modifiable factor for women’s health.

The main findings of the present study are as follows: [ 1 ] an elevated E-DII was associated with early natural menopause; [ 2 ] an elevated E-DII was associated with hysterectomy; and [ 3 ] the E-DII was negatively correlated with the age of both menopause and hysterectomy. The first reliable epidemiological estimates for the timing of natural menopause revealed a median age of 48–52 years [ 19 ]. Another more recent study reported a mean age of menopause of 49.1 (48.8, 49.4) years in the U.S. population [ 20 ]. In our study, the median age of menopause was 46 years, which we initially estimated with a combination of the ages of natural and hysterectomy-induced menopause. We subsequently re-estimated this value alone and reported that the median ages of patients who experienced natural menopause and hysterectomy-induced menopause were 49 (45, 52) and 40 (33, 45), respectively. The age range of hysterectomy-induced menopause is broadly similar to the age range of hysterectomy onset. This means that taking measures to prevent hysterectomy is undoubtedly practical and crucial for women. Notably, on the one hand, participants with increasing E-DIIs were subjected to menopause and hysterectomy earlier, which was generally consistent with the Spearman analysis results. On the other hand, the associations of a higher E-DII with early natural menopause and hysterectomy remained significant after fully adjusting for age at interview, ethnicity, education, marital status, family PIR, body mass index, smoking, physical activity, diabetes, hypertension, hyperlipidaemia, age of menarche, and parity. This is the first study with a large sample size to assess the associations of the E-DII with menopause and hysterectomy. As mentioned above, early menopause and hysterectomy are related to a variety of adverse health impacts. For example, compared with women who had reached menopause at 50–51 years, the risk of incident coronary heart disease and stroke in those with premature menopause (age < 40 years) and early menopause (age < 45 years) was 1.55 and 1.30 times greater, respectively [ 21 ]. In fact, existing evidence has substantiated the utility of the DII as a tool to characterize the inflammatory potential of diet and to predict cardiovascular disease incidence and mortality [ 22 ]. Thus, if the E-DII is a predictor of earlier menopause, maintaining a lower E-DII may be more beneficial for preventing numerous serious health outcomes (e.g., coronary heart disease, stroke and even mortality). Interestingly, we only observed a positive association between the E-DII and early natural menopause, which was not the case for later natural menopause. One possible reason is that the occurrence of natural menopause after 45 years of age or even older seems to be a common physiological phenomenon regardless of the E-DII. Additionally, previous studies revealed that high intake of some dairy products and low intake of alcohol were associated with later menopause onset [ 3 , 23 ]. In the DII calculation, the nutrients in dairy products are treated as anti-inflammatory components, and alcohol is considered a proinflammatory component. Therefore, our finding that a dietary lifestyle with anti-inflammatory potential (i.e., a lower negative E-DII score) was not significantly related to early menopause incidence was consistent with previous studies by Grisotto G et al. and Purdue-Smithe AC et al. [ 3 , 23 ]. Since variation in total nutritional intake is relatively large among individuals, the E-DII is an integrated index used to evaluate dietary inflammatory potential, which may synthetically reflect consumption lifestyle and provide more appropriate instructions to patients at high risk of menopause. Another finding was that the E-DII might be predominantly associated with hysterectomy, contributing to the most significant ORs and 95% CIs. The prevalent causes of hysterectomy are uterine fibroids, endometriosis, endometrial carcinoma, etc., in which inflammatory pathways play vital roles [ 24 ]. Although there is no information on the exact causes of hysterectomies in the raw NHANES data, dietary nutrients such as vitamin A, carotenoids and yogurt, which are potential protective factors in uterine fibroid growth and development, have been established [ 25 ]. Recently, DII was shown to be negatively related to enterolignans, an essential marker for microbiota diversity, which implies that an anti-inflammatory diet could regulate uterine fibroid pathogenesis through the improvement of systematic inflammation and the gut flora [ 26 ] and, eventually, possibly decrease the hysterectomy rate. Furthermore, the Spearman analysis results demonstrated that the E-DII was negatively correlated with physical activity, whereas physical activity was positively correlated with the age at menopause and hysterectomy. Traditionally, engaging in enough leisure-time physical activity has a high anti-inflammatory effect, which is consistent with the finding that the E-DII is negatively correlated with physical activity. Hence, on the one hand, physical activity and a lower E-DII as mutual factors might yield greater benefits in preventing early menopause and hysterectomy. On the other hand, more recreational physical activity should be encouraged for women, as their daily physical activity from the household and work gradually decreases with age [ 27 , 28 ]. Notably, there were positive correlations among the E-DII, parity, age at menopause and hysterectomy. Indeed, pregnancy and fertility are always concomitant with the consumption of multiple foods and excessive nutrients, which might lead to elevated E-DIIs to some extent. On the other hand, women who are prone to pregnancy and fertility are relatively active in oestrogen, which might maintain resistance to dietary inflammation and decrease early menopause and hysterectomy onset. Nonetheless, the interactions among dietary anti-inflammatory factors, parity, and the detailed mechanisms underlying the benefits of preventing menopause and hysterectomy merit further investigation. The data from a large sample of the U.S. representative population from the NHANES helped us estimate the associations of the E-DII with menopause and hysterectomy after adjusting for potential confounding by a series of variables. However, several limitations were inevitable in our study. First, as in previous cross-sectional studies, the nature of the data collected could not allow us to deduce causal relationships since dietary intake, menopausal status, age at menopause and hysterectomy were collected during the same period. Second, the DII from the 24-h dietary intake recall and 27 available DIIs out of 45 parameters might be a limiting in its ability to assess nutritional intake fluctuations and reflect the whole dietary inflammatory potential, respectively. However, in the cohort of 500,000 participants in the UK Biobank, there were only 29 eligible nutrient parameters among the 45 parameters [ 29 , 30 ], and previous research suggested that the associations between the DII and prespecified outcomes reported in the NHANES were generally consistent with those reported in the UK Biobank [ 30 ]. Thus, we suggest that our findings obtained via the NHANES data have clinical merits and significance. Finally, some residual or unknown confounding factors might exist and interfere with our results. Overall, these findings should be interpreted with caution in clinical practice.

Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1