Is there an association between dietary antioxidant capacity, phytochemical index, inflammatory index, and ongoing pregnancy in in vitro fertilization patients?: A cross-sectional study

Abstract

This study aimed to describe the association between the total antioxidant capacity, phytochemical index (PI), and dietary inflammatory index (DII) of the diet and ongoing pregnancy rate in infertile couples undergoing in vitro fertilization treatment. The study has a prospective cross-sectional design. Couples who met the inclusion criteria and underwent embryo transfer within the first 3 months after the interview were included. Total dietary antioxidant capacity, PI, and DII were calculated using a 1-year food frequency questionnaire. In addition, serum antioxidant and oxidant capacity, oxidative stress index, and C-reactive protein were assessed in venous blood. A total of 189 couples were included. Among these couples, the ongoing pregnancy rate was 32.3% (61/189). A significant difference was found between the DII scores of participants in the groups with and without ongoing pregnancy (P = .040 for males; P = .036 for females). A multivariate logistic regression analysis, adjusted for potential confounders, revealed no significant association between 1-year food frequency questionnaire derived dietary antioxidant capacity, DII, and ongoing pregnancy (P = .185 for females, P = .491 for males; P = .411 for females, P = .258 for males respectively). However, the PI was negatively associated with ongoing pregnancy (odds ratios: 0.169, 95% confidence intervals: 0.040–0.725, P = .027). The findings of the present study indicated an inverse association between the PI and ongoing pregnancy, underscoring the necessity for further research to investigate the impact of diet, particularly the PI and dietary inflammation, on reproductive outcomes.

1. Introduction

Infertility, observed in 10% to 15% of reproductive-age couples globally and in 1 in 4 couples in developing countries, is diagnosed when a couple is unable to conceive after 12 months of regular, intercourse without barrier or hormonal protection.^[1,2] Technological developments create new opportunities for infertility treatment, and despite increasing infertility rates. In vitro fertilization (IVF) is becoming increasingly widespread, and its growing use contributes to a successful plateau in fertility rates.^[3,4]

In the context of IVF treatment, the success rates are influenced by a combination of both immutable and modifiable factors. While factors such as ovarian reserve and age are considered immutable and thus fixed, there has been a recent surge of interest in the role of modifiable factors, including diet.^[1] Emerging evidence indicates that the nutritional status of both parents during the periconceptional period exerts a significant influence on early fetal development and may also have long-term implications for the health of the offspring.^[5,6]

Despite significant advancements in understanding the impact of nutrition on fertility, the existing literature remains inconclusive.^[7] Considering the literature, studies examining the relationship between diet and fertility have reported that only healthy lifestyle recommendations were examined.^[8] In addition to being conducted on small samples, the studies reported that possible confounding factors were not taken into account in participant selection. For example, participants may have been included who smoked or drank alcohol, or who had inflammatory conditions such as endometriosis.^[9,10] Furthermore, foods are not consumed in isolation; it is likely that there are synergistic effects between foods and food groups consumed together. The literature’s focus on specific foods and food groups rather than the overall effects of diet may lead to the potential synergistic effects of diet being overlooked.^[11] The correlation between dietary components and infertility both in vitro and vivo remains unclear.^[1] The primary gap identified is the lack of evaluation on the combined impacts of male and female diets.^[7]

The World Health Organization (WHO) emphasizes the critical importance of researching infertility treatment methods and investigating factors related to fertility, as these efforts are essential for enhancing reproductive health.^[2,11] Given the significant emotional and financial strain, it is crucial to establish cost-effective lifestyle therapies for infertile couples.^[12] Most individuals who undergo IVF treatment are amenable to lifestyle modifications, making it essential to explore adjustments that can enhance the probability of IVF success.

In the literature, the adverse effects of oxidative stress and inflammation on reproductive health are well known. It has been demonstrated that oxidative stress delays embryo development, reduces embryo quality, and decreases implantation success in IVF patients by triggering an inflammatory response. In a study involving 107 women investigating the effect of oxidative stress on ART, follicular fluid samples were collected from the women prior to ART treatment, and antioxidant and oxidant capacity as well as certain inflammatory markers were evaluated. The study found that elevated levels of pro-inflammatory molecules negatively impacted implantation success.^[13] In a systematic review that included studies examining the effects of sperm oxidative stress levels on ART in nonhuman mammals (primarily mice, pigs, and cattle), it was found that sperm oxidative stress exposure significantly reduced fertilization rates.^[14] Various studies have evaluated the use of foods with antioxidant and anti-inflammatory properties, both alone and in combination, to mitigate these effects.^[15,16] However, most of these studies focus on individual components; the overall dietary pattern and synergistic effects between nutrients especially combined and interactive effects of multiple nutrients and bioactive compounds (e.g., antioxidants, polyphenols, fiber, unsaturated fatty acids) are not sufficiently considered. This gap highlights the need for studies that more comprehensively examine the complex interactions of nutrition on reproductive health.

This study was conducted based on the hypothesis that higher total antioxidant capacity (TAC) and phytochemical index (PI), along with a lower dietary inflammatory index (DII), would be associated with increased ongoing pregnancy rates among infertile couples undergoing IVF treatment. Within the scope of this hypothesis the aim of this study was to investigate the relationship between dietary TAC, PI, and DII with ongoing pregnancy outcomes in infertile couples undergoing IVF treatment.

2. Materials and methods

2.1. Study design and population

Couples undergoing IVF treatment at Hacettepe and Ankara Universities IVF centers in Ankara, Turkey, between September 2021 and December 2023, and meeting the following inclusion and exclusion criteria were invited to participate in the study. Informed consent form was obtained from patients who met our criteria and agreed to participate. The study protocol was conducted in compliance with Declaration of Helsinki and was authorized by Hacettepe University Clinical Research Ethics Committee with the decision number 2021/12-19 at the meeting numbered 2021/12. Sample size estimation was based on the primary analysis planned using a binomial logistic regression test using ongoing pregnancy as the primary outcome. A priori power analysis was performed using G*Power version 3.1.9.7.^[17] The analysis assumed a small effect size (Cohen’s f² = 0.02), a significance level (α) of .05, and a statistical power (1 – β) of 0.80. Inclusion criteria were female age <41 years, male age <55 years, and body mass index (BMI) between 18.5 and 35.0 kg/m². Individuals with chronic and inflammatory diseases requiring continuous medication, those who have changed their dietary habits in the last 6 months, individuals using dietary supplements (excluding folic acid), those who smoke or consume alcohol (as these habits are known to affect oxidative stress, antioxidant capacity, and inflammatory status. Their inclusion could have confounded the evaluation of the relationship between dietary indices and ongoing pregnancy outcomes), female partners diagnosed with infertility due to anatomical causes, endometriosis, or tubal factors, and male partners diagnosed with azoospermia, and couples who have experienced <2 failed IVF cycles were not included in the study. Eligible couples who met the inclusion criteria and voluntarily agreed to participate were informed about the study procedures, and written informed consent was obtained from them. A data collection form including sociodemographic characteristics and health and nutrition information was applied to the couples by the researcher, the dietary intake of the participants was evaluated with 1-year food frequency questionnaire (FFQ) and anthropometric measurements were taken at the time of the first visit to the IVF clinic. The antioxidant capacity, PI and inflammatory index of the diet were then evaluated using 1-year FFQ in couples. In addition, total antioxidant, TOC and C-reactive protein (CRP) levels were analyzed in serum. Following enrollment and consent, the participating couples underwent standard IVF procedures in accordance with clinical protocols. Three months after the procedure, the couples’ medical records were reviewed to determine whether the treatment process had been successfully completed, and data regarding the ongoing pregnancy status were recorded. Couples who were unable to complete the IVF process up to the embryo transfer successfully for various reasons were excluded from the final analysis (Fig. 1).

Figure 1.

Illustrates that among the 218 couples initially enrolled in the study, 29 were excluded for various reasons. Spontaneous pregnancies occurred in 6 couples, while embryo development failure led to cycle cancelations in 10 cases. Six couples discontinued treatment due to financial constraints, and 5 couples transitioned from the planned IVF protocol to intra-uterine insemination (IUI) treatment. Additionally, data from 1 couple was unavailable, and another couple postponed their treatment for preimplantation genetic diagnosis. IUI = intra uterine insemination, IVF = in vitro fertilization.

2.2. Assessment of dietary intake

Dietary intake was evaluated by using a 1-year FFQ via face to face by the researchers. The foods consumed by the participants in the previous year were questioned with frequency and quantities from every meal to 1/3 months. The amount of food consumed was determined using a food catalogue with photographs.^[18] Although there is no standardized FFQ form in our country, the form that is most suitable for our culture and widely used has been preferred. As the questionnaire had not been formally validated, total error and coefficient of variation values are not available. The assessment was conducted once during the study period to capture habitual dietary patterns. To strengthen our dietary assessment, we also collected 24-hour dietary recalls and compared them with FFQ estimates, which showed consistency and thereby provide support for the reliability of our findings. Energy, macro- and micronutrient intake levels were calculated with the computer-aided nutrition program “Nutrition Information System (V. 8.1).”^[19] The total dietary antioxidant capacity, PI, inflammatory index of the participants was assessed using data obtained from a 1-year FFQ. Participants reporting daily energy intakes below 500 kcal or above 3500 kcal were excluded from the analysis, in line with commonly applied cutoff thresholds to identify implausible dietary reports.^[20]

2.3. Assessment of dietary antioxidant capacity

The values determined for 100 g of foods using the ferric reducing antioxidant potential, oxygen radical absorption capacity, total radical-trapping antioxidant parameters, trolox equivalent antioxidant capacity databases were defined to the Nutrition Information System (V. 8.1) and calculated.^[19] The database containing the antioxidant capacity of 3100 foods published by Carlsen et al^[21] and the database containing the antioxidant capacity of 156 foods published by Pellegrini et al^[22] and the database published by Haytowitz et al^[23] were used in the ferric reducing antioxidant potential, total radical-trapping antioxidant parameters, and trolox equivalent antioxidant capacity analyses. When a specific food was not present in the databases, a similar one with comparable characteristics was substituted.

2.4. Assessment of phytochemical index

Phytochemical index (PI) was determined as the ratio of energy obtained from phytochemical-rich foods to total energy, as developed by McCarty (PI = [daily kcal from phytochemical-rich foods/total kcal] × 100).^[24] When calculating PI, fruits and vegetables with high phytochemical content (except potatoes due to their high starch content), legumes, whole grains, oilseeds, olives and olive oil were included. Fruit and vegetable juices with high phytochemical content were also included. Data obtained from the participants’ 1-year FFQ were transferred to the formulation and calculations were made.

2.5. Assessment of dietary inflammatory index

The DII was calculated using the database developed by Shivappa et al,^[25] in which foods are categorized according to their inflammatory and anti-inflammatory properties. A total of 41 ingredients were used, excluding alcohol, saffron and rosemary as they were not consumed by any participants. Firstly, in line with Nutrition Information System (V. 8.1), energy, carbohydrate, protein, fat, saturated fat, polyunsaturated fatty acids, monounsaturated fatty acids, cholesterol, trans fatty acid, omega 6, omega-3, selenium, fiber, vitamin A, retinol, vitamin D, vitamin E, vitamins of B₁, B₂, B₃, B₆, B₁₂, C, magnesium, iron, zinc, caffeine, flavone, flavonol, flavanone, anthocyanidin, isoflavonoid intake values and onion, garlic, pepper, green and black tea, ginger, turmeric, and thyme consumption values were obtained. The values obtained were subtracted from the global consumption amount and divided by the standard deviation specific to the relevant value to find the z-score. Then, the percentile and center were converted to percentile and multiplied by the effect score related to the relevant value. After the same process was performed for all values, all were added together to obtain the participant’s dietary inflammatory score. A higher DII score was considered to indicate that the diet had proinflammatory properties, while a lower score was considered to indicate that the diet had anti-inflammatory properties.^[25] The maximally pro-inflammatory diet score is +7.98, and the maximally anti-inflammatory DII score is −8.87.^[25,26] The effective range in most studies is approximately − 5.5 to +5.5.^[27]

2.6. Anthropometrical measurements

Within the scope of the study, the weight and height of the participants, waist, hip and neck circumferences were measured; BMI and waist/hip ratio were calculated from these values.^[28] For neck circumference measurements, values of 34 cm or more for women and 37 cm or more for men were considered at risk. For waist circumference, measurements of 94 cm or more for men were classified as at risk, and 102 cm or more as high risk; for women, 80 cm or more was classified as at risk, and 88 cm or more as high risk. The waist-to-hip ratio was evaluated according to WHO 2011 criteria, with values below 0.90 for men and below 0.85 for women considered normal.^[29] All anthropometric measurements were carried out by the same trained investigator (NKH), and each measurement was repeated twice in rotational order on the same participant to enhance accuracy and minimize measurement error.

2.7. Blood collection and assessment of serum biomarkers

Venous blood was obtained from all participants in the morning following an overnight fasting of at least 12 hours and centrifuged at 3000×g for 10 minutes. To reduce variability, participants were instructed to refrain from vigorous physical activity for at least 24 hours before sample collection and to maintain usual hydration. Serum was separated and collected in Eppendorf’s and frozen at −80°C for biochemical evaluations. Samples were thawed at +4^oC the night before analysis. Serum TAC was analyzed by a commercially available Kit (Relassay, Turkey) in accordance with the instructions of the manufacturer, based on the colorimetric assay of oxidation reduction at the wavelength of 658 nm. The assay results are presented as Trolox Equivalent, an analog of vitamin E (mmol Trolox Eq/L).^[30] Total oxidant capacity (TOC) was analyzed by a commercially available Kit (Relassay, Turkey). The kit was calibrated with hydrogen peroxide and results were expressed as micromole hydrogen peroxide equivalent/Liter (μmol H₂O₂ Eq/L).^[31] The ratio of TOC to TAC was indicated as the oxidative stress index. CRP level was obtained using immunoturbidimetric method. Anti-CRP antibodies reacted with the antigen in the sample to form an antigen/antibody complex. After agglutination, this turbidity level was measured: human hs-CRP ELISA kit, sensitivity: 0.053 ng/mL, detection range from 0.1 to 40 ng/mL.^[32] In addition, the intra-assay and inter-assay coefficients of variation of the kits used were reported by the manufacturer as 2.7% and 2.5%, respectively.

2.8. Clinical management and assessment of outcomes

The IVF procedure was performed according to European Society of Human Reproduction and Embryology, American Society for Reproductive Medicine guidelines in both centers.^[33,34] Participants were assigned to 1 of 3 stimulation protocols based on clinical indications: luteal-phase gonadotropin-releasing hormone (GnRH) agonist (long) protocol, follicular-phase GnRH antagonist protocol or follicular-phase GnRH agonist/flare (short) protocol. Throughout gonadotropin stimulation, patients were regularly evaluated by clinical staff for estradiol (E2) concentrations, number and size of developing follicles by transvaginal ultrasonography, and endometrial thickness for endometrial receptivity. When at least 3 follicles reached a diameter of ≥17 mm, final oocyte maturation was triggered by administering human chorionic gonadotropin (hCG) approximately 35 to 36 hours prior to oocyte retrieval.^[33]

Oocyte retrieval was performed under sedation. Additionally, male partners provided semen samples following 2 to 7 days of sexual abstinence. Semen analysis and preparation was conducted according to WHO criteria for fertilization stage.^[35] Following these procedures fertilization, in which motile spermatozoa were co-incubated with mature (MII) oocytes in culture media for approximately 16 to 18 hours executed. Patients undergoing cryo-thaw cycles were subjected to endometrial preparation protocols. Embryo development was monitored daily, and embryos were graded based on cell number, symmetry, fragmentation, and other morphological criteria.^[36] The embryo transfer procedure was conducted 3 to 5 days following the oocyte retrieval. Hormone of hCG levels were monitored to confirmation for pregnancy.

Clinical pregnancy is confirmed when an ultrasound detects cardiac activity and when at least 1 gestational sac is visible. Because it shows the number of pregnancies that have the potential to result in live births. The ongoing pregnancy defines as confirmation of fetal viability at 12 weeks of gestation through ultrasound (e.g., detection of fetal heartbeat). Its rate is a crucial outcome metric used to evaluate the effectiveness of assisted reproductive techniques (ART) cycles.

2.9. Statistical analyses

The adherence of numerical variables to a normal distribution was assessed using the Kolmogorov–Smirnov and Shapiro–Wilk tests. Continuous variables with a normal distribution were summarized using mean and standard deviation, whereas those not normally distributed were described using median and interquartile range (IQR). Frequency distributions were utilized to describe categorical variables. Comparisons of differences between groups for independent variables with a normal distribution were conducted using the t-test for independent samples. The Mann–Whitney U test was employed for variables exhibiting non-normal distribution. Values in groups with and without ongoing pregnancies were analyzed by comparing women and men within each group. The association between categorical variables and groups was assessed using the chi-square test. Logistic regression analysis was executed to determine the factors affecting ongoing pregnancy. The dependent variable was the presence or absence of ongoing pregnancy, and the independent variables based on dietary intakes were analyzed by dividing them into quartiles (Q₁–Q₄). Analyses were conducted by creating 3 separate models: model 1 included crude analyses, model 2 was adjusted for age, physical activity and education status, and model 3 was created as a model adjusted for, age, physical activity and education status, total energy intake and BMI. The results were reported with odds ratios (OR) and 95% confidence intervals (CI), and P values were calculated to evaluate the linear relationship of variables with ongoing pregnancy. The Hosmer–Lemeshow goodness-of-fit test was used to assess the model fit for the logistic regression analysis. Statistical significance was determined at the .05 level for independent variables in the constructed model. The statistical analyses were conducted by IBM^© SPSS Statistics software V 27 (IBM Corp., Armonk).^[37]

3. Results

3.1. Description of participants

A total of 218 couples was recruited for our study, among which 189 couples were included in the analysis after considering the exclusion criteria (Fig. 1). The average age was 30.9 ± 5.19 years for females, and 33.7 ± 5.07 years for males (Table 1). No significant difference in age was observed between males and females when compared based on ongoing pregnancy status. The median duration of infertility was 37 months. With no significant differences between groups (P = .471), the main causes of infertility were male factor infertility (17.5%), polycystic ovary syndrome (16.4%), anovulation (7.9%), reduced ovarian reserve (26.5%), and unexplained infertility (31.7%). Of the 73 pregnancies that achieved clinical pregnancy, 12 resulted in abortions, while 61 (32.3%) progressed to ongoing pregnancies.

Table 1.

General characteristics of infertile couples undergoing in vitro fertilization (IVF) treatment according to ongoing pregnancy status.

Variable	Ongoing pregnancy NO n = 128 couples				Ongoing pregnancy YES n = 61 couples				Total n = 189 couples				P
	Male	Female			Male	Female			Male	Female			P m ale	P f emale
Age (yr)	34 ± 5.26	31.2 ± 5.32			33.2 ± 4.63	30.2 ± 4.9			33.7 ± 5.07	30.9 ± 5.19			.229†	.281†
Duration of infertility (mo)	38.5 (42)				37 (28.5)				37 (36)				.684‡
Cause of infertility (n, %)
Diminished ovarian reserve	38		29.7		12		19.7		50		26.5		.471§
Unexplained infertility	37		28.9		23		37.7		60		31.7
Polycystic ovary syndrome	19		14.8		12		19.7		31		16.4
Anovulation	10		7.8		5		7.8		15		7.9
Male factor	24		18.8		9		14.8		33		17.5
Number of retrieved oocytes (median, IQR)	7 (7.0)				8 (10.0)				8 (10.0)				.009 ‡
Number of mature (metaphase-2) oocytes (median, IQR)	5.0 (7.0)				9.0 (11.0)				7.0 (9.0)				.003 ‡
Embryo transfer stage (n, %)
Cleavage	53			41.4	10			16.4	63			33.3	<.001§
Blastocyst	75			58.6	51			83.6	126			66.6
Transfer type (n, %)
Fresh	69			53.9	19			31.1	88			46.6	.005§
Frozen-Thawed	59			46.1	42			68.9	101			53.4
Blastocyst morphology (n, %)*
Excellent	28			37.3	33			64.7	61			48.4	.025§
Good	9			12.0	3			5.9	12			9.5
Fair	20			26.7	9			17.6	29			23.0
Poor	18			24.0	6			11.8	24			19.0
Clinical pregnancy status (n, %)
Yes	12			9.4	61			100.0	73			38.6	<.001 §
No	116			90.6	–			–	116			61.4

IQR = interquartile range, IVF = in vitro fertilization.

^*The blastocyst morphology grading was done as follows: Excellent (3AA, 4AA, 5AA), Good (3,4,5,6 AB or BA), Fair (3,4,5,6 BB or AC or CA), and Poor (3,4,5,6 BC or CC). In cases where multiple embryos were transferred, the grading was based on the embryo with the best morphology.

^†Paired samples t test.

^‡Mann–Whitney U test.

^§χ² test, bolding indicates statistical significance P < .05.

3.2. Nutritional status and anthropometry

Median energy and nutrient intakes of participating couples over the past year according to ongoing pregnancy are presented in Table S1, Supplemental Digital Content, https://links.lww.com/MD/Q198. All intakes except vitamin C and potassium were observed to be similar between the couples with and without ongoing pregnancy (P > .05). Males in couples with ongoing pregnancy were found to intake more potassium and vitamin C in their diets throughout the previous 12 months. The median vitamin C intake among males in couples with ongoing pregnancies was 137.8 (90.60) mg, whereas the median intake among males in couples without pregnancies was 103.1 (87.12) mg (P = .038). Similar to vitamin C, potassium intake was found to be higher in males in the pregnant groups. The median potassium intake among males in couples with ongoing pregnancies was 3289.4 (1305.27) mg, whereas the median intake among males in couples without pregnancies was 2784.8 (1430.96) mg (P = .041).

According to anthropometric measurements of the participants except neck circumference, all of them were similar between groups. In the groups where ongoing pregnancy occurred, the number of participants with an elevated neck circumference was found to be lower in both males and females (P = .025; .007, respectively; Table S2, Supplemental Digital Content, https://links.lww.com/MD/Q198).

3.3. Dietary indexes and serum parameters

Dietary TAC, PI, and DII values obtained from 1-year food frequency of participating couples are presented in Table 2. Dietary antioxidant capacity values, and PI values were not different according to the ongoing pregnancy status. Merely the DII score exhibited a significant difference in both males and females within couples experiencing ongoing pregnancy. The mean value was 3.37 ± 3.22 for nonpregnant females; but 2.19 ± 3.13 for pregnant females (P = .040). Likewise, it was 2.71 ± 2.98 for husbands of nonpregnant females and 1.58 ± 2.99 for husbands of pregnant females (P = .036).

Table 2.

Nutritional characteristics of infertile couples undergoing in vitro fertilization (IVF) treatment according to 1-year food frequency.

	Ongoing pregnancy NO n = 128 couples		Ongoing pregnancy YES n = 61 couples		Total n = 189 couples		P
	Male	Female	Male	Female	Male	Female	P ‡	P §
ORAC (µmol TE)	14,242.8 (11,402.21)	12,739.0 (11,439.03)	17,820.7 (9611.95)	16,370.9 (9566.06)	15,468.3 (11,005.10)	14,191.4 (10,433.0)	.257*	.186*
TEAC (mmol TE)	3.8 (2.84)	3.1 (29.63)	4.3 (2.18)	3.7 (19.81)	3.9 (2.51)	3.4 (2.63)	.683*	.249*
TRAP (mmol TE)	3.2 (2.58)	2.6 (2.73)	3.3 (2.32)	2.9 (20.82)	3.3 (2.55)	2.8 (2.42)	.669*	.148*
FRAP (mmol)	11.7 (7.4)	9.1 (8.84)	12.2 (6.95)	10.9 (59.16)	11.9 (6.93)	10.0 (7.71)	.707*	.202*
PI	26.2 ± 11.43	30.5 ± 12.15	28.1 ± 14.39	32 ± 16.18	26.8 ± 12.42	31.0 ± 13.54	.400†	.533†
DII	2.71 ± 2.98	3.37 ± 3.22	1.58 ± 2.99	2.19 ± 3.13	2.35 ± 3.02	2.98 ± 3.23	.040 †	.036 †

Bolding indicates statistical significance P < .05.

DII = Dietary inflammatory index, IVF = in vitro fertilization, PI = Phytochemical index.

^*Mann–Whitney U test.

^†Paired samples t test.

^‡Comparison of males.

^§Comparison of females.

Cohen’s d effect sizes were calculated to quantify the magnitude of the differences between groups beyond statistical significance. The PI showed very small effect sizes in both females (d = 0.11) and males (d = 0.15). In contrast, the DII demonstrated small-to-moderate effect sizes in both females (d = –0.37) and males (d = –0.38), indicating that lower DII scores were associated with ongoing pregnancy. These results suggest that the inflammatory potential of the diet may be more influential for pregnancy outcomes than the PI or antioxidant capacity.

Table 3 shows serum antioxidant and oxidant capacity values according to ongoing pregnancy status. TAC and TOC were similar between groups. The median and IQR values for total serum antioxidant capacity were 1.7 (0.50) mmol/L in males in the groups where ongoing pregnancy did not occur and 1.8 (0.45) mmol/L in males in the groups where pregnancy did occur (P = .103). In females, the serum TAC value was 2.02 (0.40) in the groups where pregnancy did not occur and 1.97 (0.27) mmol/L in females who did (P = .222). In terms of oxidant capacity, median values were found to be the same in the groups where pregnancy occurred and where it did not. However, CRP levels were observed significantly between groups of females. The median and IQR values of CRP in pregnant women was 0.71 (1.08) mg/L while it was 1.33 (2.31) mg/L in nonpregnant women. (P = .017).

Table 3.

Serum parameters of infertile couples undergoing in vitro fertilization (IVF) treatment according to ongoing pregnancy status.

	Ongoing pregnancy NO n = 128 couples		Ongoing pregnancy YES n = 61 couples		Total n = 189 couples		P †	P ‡
	Male	Female	Male	Female	Male	Female
	Median (IQR)	Median (IQR)	Median (IQR)	Median (IQR)	Median (IQR)	Median (IQR)
Serum antioxidant capacity (mmol/L)	1.7 (0.50)	2.02 (0.40)	1.8 (0.45)	1.97 (0.27)	1.7 (0.51)	1.9 (0.37)	.103*	.222*
Serum oxidant capacity (mmol/L)	5.8 (3.82)	6.0 (3.39)	5.8 (3.32)	6.0 (2.04)	5.8 (3.72)	6.0 (2.91)	.945*	.758*
Oxidative stress index	0.32 (0.30)	0.29 (0.15)	0.33 (0.26)	0.31 (0.13)	0.32 (0.28)	0.30 (0.14)	.525*	.278*
CRP (mg/L)	0.81 (1.63)	1.33 (2.31)	1.05 (1.75)	0.71 (1.08)	0.87 (1.64)	1.0 (1.78)	.685*	.017 *

CRP = C-reactive protein, IQR = interquartile range, IVF = in vitro fertilization.

^*Mann–Whitney U test.

^†Comparison of males.

^‡Comparison of females, bolding indicates statistical significance P < .05.

3.4. Predictors of ongoing pregnancy: logistic regression analysis

A binary logistic regression analysis was utilized to ascertain the impact of dietary and serum parameters on ongoing pregnancy, with adjustments made for potential confounding variables. The logistic regression model showed good fit according to the Hosmer–Lemeshow test (χ² = 13.518, P = .095). In females, higher levels of the PI (Q₃ and Q₄) compared to the reference group (Q₁) were found to significantly reduce the likelihood of ongoing pregnancy. In model-3, which adjusted for all possible confounding factors, women in the Q₃ group had 82% lower odds of ongoing pregnancy compared to the reference (OR = 0.179, 95% CI: 0.043–0.749). In the highest quartile, Q₄, the probability of pregnancy was 93% lower (OR = 0.064, 95% CI: 0.011–0.388; Table 4).

Table 4.

Logistic regression analysis for factors affecting ongoing pregnancy.

Variables	Model-1 (crude)		Model-2*		Model-3†
DII	Female	Male	Female	Male	Female	Male
Q1	Ref	Ref	Ref	Ref	Ref	Ref
Q2	2393 (0.611–9372)	3473 (0.559–21.591)	2261 (0.560–9124)	3048 (0.473–19.658)	2478 (0.547–11.232)	4401 (0.557–34.758)
Q3	1243 (0.252–6046)	2660 (0.339–20.856)	1077 (0.199–5833)	2616 (0310–22.111)	1341 (0.220–8168)	7222 (0.547–95.388)
Q4	2904 (0.543–15.526)	3891 (0.504–30.057)	2474 (0.432–14.185)	4590 (0.555–37.923)	3674 (0.492–27.440)	14.592 (1029–206.859)
P	.385	.537	.442	.515	.411	.258
ORAC
Q1	Ref	Ref	Ref	Ref	Ref	Ref
Q2	0.643 (0.113–3674)	0.273 (0.034–2222)	0.614 (0.104–3604)	0.281 (0.033-2359)	0.589 (0.101–3427)	0.203 (0.022–1897)
Q3	7709 (0.416–142.895)	0.422 (0.026–6847)	8742 (0.448–170.440)	0.416 (0.021–8298)	9704 (0.490–192.316)	0.283 (0.012–6427)
Q4	5036 (0.199–127.623)	0.416 (0.015–11.146)	6129 (0.225–166.978)	0.457 (0.015–14.268)	7193 (0.250–206.805)	0.574 (0.017–19.910)
P	.221	.653	.219	.681	.185	.491
PI
Q1	Ref	Ref	Ref	Ref	Ref	Ref
Q2	0.510 (0.154–1688)	1017 (0.247–4181)	0.497 (0.139–1779)	0.967 (0.226–4145)	0.429 (0.117–1570)	0.696 (0.146–3319)
Q3	0.225 (0.061–0.826)	0.288 (0.059–1395)	0.219 (0.056–0.861)	0.285 (0.055–1476)	0.179 (0.043-0.749)	0.138 (0.021–0.907)
Q4	0.095 (0.019–0.485)	0.919 (0.193–4362)	0.090 (0.016–0.493)	0.540 (0.079–3712)	0.064 (0.011–0.388)	0.244 (0.028–2131)
P	.020	.312	.027	.384	.016	.168
CRP	0.786 (0.584–1058)	0.975 (0.841–1130)	0.790 (0.586–1066)	0.974 (0.834–1137)	0.821 (0.588–1147)	0.973 (0.803–1179)
P	.113	.735	.124	.737	.248	.782
TAC	0.261 (0.042–1646)	6750 (1304–34.948)	0.263 (0.041–1691)	5512 (1004–30.261)	0.253 (0.036–1774)	8009 (1224–52.389)
P	.153	.023	.159	.049	.167	.030
TOC	0.973 (0.871–1086)	0.958 (0.837–1096)	0.976 (0.872–1093)	0.969 (0.840–1118)	0.974 (0.865–1097)	0.955 (0.815–1118)
P	.621	.532	.676	.668	.663	.565

Bolding indicates statistical significance P <.05.

CRP = C-reactive protein, DII = dietary inflammatory index, ORAC = oxygen radical absorbance capacity, PI = phytochemical index, TAC = total serum antioxidant capacity, TOC = total serum oxidant capacity.

^*Model-2, adjusted for age, physical activity and education status,

^†Model-3, adjusted for age, physical activity and education status, total energy intake and BMI.

When serum antioxidant capacity and dietary antioxidant capacity levels were examined, the serum antioxidant capacity in males was observed to influence the pregnancy status of their partners. In model 3, elevated serum TAC in males augmented the likelihood of ongoing pregnancy in partners by 8 times (OR = 8.009, 95% CI: 1.224–52.389, P = .030). In females, no significant correlation was found between serum TAC and the likelihood of ongoing pregnancy. However, dietary antioxidant capacity was also found to have no effect on ongoing pregnancy in either males or females. Despite significant differences observed in binary analyses regarding inflammation markers, logistic regression analysis indicated that neither the DII nor CRP levels significantly influenced ongoing pregnancy outcomes (e.g., female Q₄: OR = 3.674, 95% CI: 0.492–27.440; male Q₄: OR = 14.592, 95% CI: 1.029–206.859, P for females = .411, P for males = .258).

4. Discussion

This study investigated the association between the TAC, PI, and DII of the diet and ongoing pregnancy rate in infertile couples undergoing IVF treatment. We found that high phytochemical intake was inversely associated with ongoing pregnancy (Table 4). This finding suggests that phytochemical intake may have an unexpected adverse effect on reproductive health and reveals that nutrition is not only a supportive factor in infertility treatment but also a factor that needs to be carefully regulated.

Considering the antioxidant properties of phytochemicals, it is plausible that different mechanisms may underlie this unexpected finding, which was contrary to the initial hypothesis. One possible explanation is that excessive intake of certain phytochemicals, such as phytoestrogens, can disrupt reproductive hormone balance by modulating estrogen receptor activity, which can lead to adverse effects on ovulation processes or endometrial receptivity.^[38] In experimental animal studies, high levels of phytoestrogen exposure have been shown to inhibit follicle development, reduce ovulation frequency, and disrupt uterine or oviduct function.^[39] In human studies, however, the results are conflicting; while some studies show no adverse effects, other findings point to the potential endocrine-disrupting effects of phytoestrogens.^[40,41] The second is that high levels of antioxidant intake expressed as antioxidant paradox, may have suppressed physiological processes by prooxidant effect.^[42] The third is that phytates, tannins, and other compounds found in plants that intensely inhibit absorption increase with increased phytochemical intake and affect the absorption levels of beneficial compounds. The fourth possible mechanism is thought to be a reverse causality mechanism. Individuals consciously preferred healthy dietary behaviors due to infertility and they preferred plant-based foods, which resulted in a higher index value.

We found that in the crude model, male serum antioxidant capacity values were significantly associated with ongoing pregnancy outcomes. The association remained consistent after multivariate adjustment (Table 4). However, serum antioxidant capacity alone may not be a sufficient indicator for the evaluation of an individual’s antioxidant capacity. The fact that dietary antioxidant capacity results were not found to be significant and semen antioxidant capacity values were not measured also suggests that this effect may not be significant. From a physiological perspective, it is thought that oxidative stress and inflammation are currently accepted as the main causes of infertility.^[43–45] In fact, the literature has introduced a new concept called inflammatory infertility.^[46,47] In recent years, the effects of antioxidant-rich diets and antioxidant supplements on male fertility have attracted considerable interest. For example, there are 2 separate Cochrane reviews that have evaluated the effects of antioxidants on males and females.^[15,16] A total of 76 randomized controlled trials on the use of antioxidants in female infertility compared the effects of oral antioxidant supplementation with placebo control, standard treatment protocol and different antioxidants. Although the data obtained were of low and very low quality due to inadequacies in the methods of the included studies and risks of bias, it was reported that antioxidant supplementation increased the clinical pregnancy rate compared to standard treatment protocol or placebo control (OR = 1.65 95% CI 1.43–1.89; P < .001, I² = 63%). In other words, while the probability of clinical pregnancy of an infertile woman receiving standard treatment protocol was 19%, the probability of clinical pregnancy of a woman using antioxidant supplementation was found to be 25% to 30%. However, due to the high heterogeneity and very low quality of the included studies, the effect of antioxidant supplementation on live birth rates is not certain (OR = 1.81, 95% CI 1.36–2.43; P < .001, I² = 29%).^[15] In a Cochrane review including a total of 10,303 men and 90 studies on the use of antioxidants in male infertility, it was found that live birth rates were 16% in couples who did not use antioxidants and 17% to 27% in those who did (OR = 1.43, 95% CI 1.07–1.91 P = .02, I² = 44%).^[16] Notwithstanding this rational foundation, certain investigations have indicated contrary findings. A randomized clinical investigation revealed that antioxidant supplementation had no effect on male infertility.^[48] Although the fact that the effect of supplementation and not dietary intake was examined suggests the generalizability of our study, the randomization design of the study and the large sample size increase the reliability of the results. This can be explained by the absence of any oxidative stress effect that antioxidants could potentially repair.^[49] Considering the spermiogram values of the study sample, it is thought that this reasoning may also be valid for this study. It was observed that the sperm counts and motility rates of the participant men were higher than the WHO criteria. According to WHO criteria, sperm count should be at least 39 million and motility rate should be at least 40%.^[50] In the study sample, the mean sperm count was 58 million and the motility rate was 41%. Studies on the effects of antioxidants on reproductive outcomes have also identified various challenges. First, the uncertainty surrounding combinations and duration of treatment, another being unclear about which treatment should be applied to which patient groups, and finally, the uncertainty surrounding long-term effects.^[51]

Our results show that inflammatory index values were lower in both females and males in the ongoing pregnancy group, and in females this finding was further supported by lower CRP levels. However, according to the result of logistic regression analysis in which all variable effects were analyzed together, inflammatory index was not found to be associated in the realization of ongoing pregnancy. The literature on this topic is equivocal, likely reflecting differences in study populations and methodologies. In a cross-sectional study conducted on 144 women in Iran, it was reported that DII did not affect pregnancy outcomes in IVF cycles.^[3] In another study in which DII scores of 3496 women conducted from the data of the NHANES study were analyzed, it was reported that high scores obtained from the index were significantly and positively associated with an increased risk of infertility.^[52] A systematic review of 10 studies aimed to evaluate the effects of serum CRP levels on pregnancy outcomes in women undergoing ART examined CRP levels at different stages of the cycle. It found that low CRP levels at the onset significantly increased the likelihood of clinical pregnancy. However, pregnancy rates were reported to be higher in women with higher CRP levels at implantation.^[53] It is thought that the pro-inflammatory diet of the majority of the sample may have been effective in the obtained result.

To our knowledge, there is currently no published research that has explored the effects of dietary antioxidants, phytochemicals and inflammatory indexes on ongoing pregnancy rates in a couple-based approach. Current approaches to assessing dietary intake, such as FFQs, short-term dietary surveys, or single indices such as DII or phytochemical indices, each have methodological limitations and may not fully reflect habitual dietary patterns. To address this issue, our study applied several commonly used dietary indices that complement each other’s weaknesses and provide more reliable estimates. The strongest aspect of the study is the use of 1-year FFQ, which enables the determination of the routine eating habits of the participants. Although it is a known fact that oogenesis and spermatogenesis take approximately 3 months, considering the long data collection process of the study, it is thought that 1-year FFQ is the safest method in order not to be affected by seasonal variations.^[54] Furthermore, in contrast to typical infertility research, the inclusion of couples rather than a single gender is believed to yield a more precise representation of the factors influencing pregnancy success. At the same time, it is thought that the fact that not a single index but many indices that will cover the deficiencies of each other were examined together and that these were supported by the results of antioxidant and oxidant capacity and CRP in serum increased the reliability of the data of the study.

Nonetheless, some limitations must be explained. First of all, owing to the cross-sectional design of the study, causal inferences between variables cannot be made. Second, although the 1-year FFQ is considered the most reliable recording method, it may be subject to recall bias. In order to minimize recall bias, researchers conducted face-to-face interviews and used a catalog of food photographs to help determine portion sizes. Moreover, considering the willingness of this patient group to change their diet, the fact that the food consumption record was taken only once can also be considered a limitation. Third, due to the nature of the PI, the fact that foods with no energy value such as green tea and black tea are not included in the calculation should also be considered. Fourth, the fact that the sample consists of relatively young adults who do not smoke or drink alcohol also increases homogeneity, making it difficult to generalize the results to different groups. Finally, although the effect of processing and storage conditions was tried to be considered in the calculations of dietary antioxidant capacity, it is thought that the high variability of phenolic substance contents may have affected the results.

On the other hand, high variability was observed among the variables; this situation may reflect the heterogeneous nature of dietary reporting, interindividual differences in nutrient metabolism, and biological responses. Such variability has also been observed in previous studies examining diet-related indices and fertility outcomes.^[7,55,56] Despite this variability, Cohen’s d effect size calculation was performed to provide additional information about the magnitude of the differences.

5. Conclusions

In conclusion, our findings suggest that although there was no significant association related to antioxidant capacity and inflammatory responses, PI of females was found to be associated ongoing pregnancy outcomes. This study is believed to contribute novel insights to the literature and establish a significant foundation for future research. Further prospective studies with larger sample sizes, controlled dietary interventions, and additional biomarkers are needed to clarify these associations and their underlying mechanisms.

Acknowledgments

The authors sincerely thank to the hospital administration for their support in facilitating the conduct of this study. We are also deeply thankful to all the couples who generously volunteered to participate, thereby making this research possible.

Author contributions

Conceptualization: Nagihan Kircali-Haznedar, Yavuz Emre Şükür, Pelin Bilgiç.

Data curation: Nagihan Kircali-Haznedar.

Investigation: Nagihan Kircali-Haznedar.

Methodology: Nagihan Kircali-Haznedar, Sezcan Mümüşoğlu, Pelin Bilgiç.

Project administration: Pelin Bilgiç.

Supervision: Sezcan Mümüşoğlu, Pelin Bilgiç.

Writing – original draft: Nagihan Kircali-Haznedar, Pelin Bilgiç, Sezcan Mümüşoğlu.

Writing – review & editing: Sezcan Mümüşoğlu, Yavuz Emre Şükür, Pelin Bilgiç.

Supplementary Material

medi-104-e44901-s001.pdf ^{(170.3KB, pdf)}