Abstract
Background
Little is known regarding joint associations of maternal diet quality and adiposity in relation to milk concentrations of inflammatory markers.
Objectives
We evaluated whether maternal diet quality and prepregnancy body mass index (BMI) predict C-reactive protein (CRP) concentration in human milk.
Methods
We used Cincinnati data from Global Exploration of Human Milk study, a prospective cohort of healthy mother–infant dyads (n = 100). We computed a dietary inflammatory index (DII) from diet recalls obtained 4–13 wk postpartum, where higher score indicating a more proinflammatory diet. CRP concentrations in week 4 milk samples were measured after validating a high-sensitivity enzyme-linked immunosorbent assay (ELISA) kit. Participants were categorized into low and high DII groups at the median and BMI groups defined as normal (<25) or elevated (≥25). The associations of categorical DII score and BMI with milk CRP concentrations were tested using linear regression. Then, we compared milk CRP concentrations across 4 combinations of DII × BMI.
Results
Median (Q1, Q3) values were as follows—DII: 0.13 (−1.54, 1.27); BMI: 25.4 (22.8, 30.4); and milk CRP: 101.7 (54.8, 172.3) ng/mL. Across the 4 DII × BMI groups, milk CRP was significantly different (P = 0.009) but there was no significant difference in maternal age, education, or ethnicity. Having BMI ≥ 25 was associated with higher milk CRP concentrations than having BMI <25 (β ± SE: 96.19 ± 27.23 ng/mL; P = 0.0006), although the main effect of DII was not statistically significant. The DII × BMI interaction was significant: BMI predicted milk CRP concentrations only when DII was low. Women with low DII and normal BMI had lower CRP concentrations [n = 30; 64.3 (38.2, 121.4) ng/mL] than those in the other 3 groups combined [n = 70; 124.1 (71.2, 181) ng/mL; P = 0.019].
Conclusions
Diet modifies the relation between maternal BMI and milk CRP. The study of milk cytokines and other bioactive compounds should include measures of diet quality, when feasible, in addition to maternal adiposity.
Keywords: human milk bioactives, exclusive breastfeeding, C-reactive protein, inflammation, BMI, maternal dietary inflammatory index, inflammatory compounds
Introduction
Human milk contains both anti-inflammatory and proinflammatory components, many of which may have biological impact on the breastfeeding infant, such as the regulation of growth, metabolism, immune development, and the microbiome [1,2]. Given that early infancy is a sensitive developmental period [[3], [4], [5]], it is critical to understand the link between modifiable maternal factors and concentrations of milk inflammatory components. Maternal diet and prepregnancy BMI (in kg/m2) are known to alter some nutrients in milk [[6], [7], [8], [9]], but little is known regarding overall maternal diet quality and BMI in relation to milk concentrations of bioactive compounds.
Both poor diet quality and high BMI contribute to a generalized low-grade inflammatory state [10]. Epidemiological studies consistently report an association between obesity and elevated serum concentrations of proinflammatory components [11,12]. Moreover, serum concentrations of inflammatory markers are positively associated with consuming a Western-type dietary pattern, than in those with greater adherence to healthy dietary patterns such as the Mediterranean diet [[13], [14], [15], [16]]. Elevated inflammatory cytokines in the blood may lead to increased uptake in milk [17], which is consistent with the findings that greater prepregnancy BMI is associated with greater C-reactive protein (CRP) concentrations in human milk [18,19]. However, our understanding of the effects of maternal diet on human milk inflammatory compounds and its joint association with BMI is limited.
The dietary inflammation index (DII) is an evidence-based algorithm to compute the overall inflammatory potential of a diet independent of regional or cultural dietary habits [20]. A lower DII score indicates a more anti-inflammatory diet, whereas a higher DII score represents a more proinflammatory diet. DII score is positively associated with higher serum concentrations of inflammatory markers such as CRP, IL-6, and TNF-α in general adult populations [[21], [22], [23], [24]] and pregnant women [[25], [26], [27]]. Current knowledge on the influence of dietary inflammation potential during lactation on milk inflammatory components is very limited. Moreover, the existing literature lacks representation of lactating women from the Midwestern United States, where dietary patterns of general adults tend to be high in energy-dense foods, added sugar, and alcohol and poor in fiber, vegetables, fruits, whole grains [28], and fish [29]. Addressing this knowledge gap is especially timely given that United States Department of Agriculture and Department of Health and Human Services recently issued the first detailed dietary guidelines specifically for pregnancy and lactation in the United States [30]. During the extensive review process that led to these guidelines, an advisory committee noted the critical need for more research on the diets of lactating women and its effects on human milk quality to inform future guidelines [31].
Leveraging existing data and biospecimens from a Midwestern cohort of lactating women, we hypothesized that participants with normal prepregnancy BMI and an anti-inflammatory dietary pattern during lactation would exhibit lower human milk CRP concentrations than participants with an overweight or obese BMI and/or inflammatory dietary pattern.
Methods
Study participants
We used deidentified data and biospecimens from the Cincinnati site of the Global Exploration of Human Milk (GEHM) study (2007–2008) [9,32], which was a multicountry, longitudinal cohort of mother–infant pairs. Participants were enrolled within the first 2 wk postpartum through community-based recruitment in Cincinnati metro area and followed ≤52 wk postpartum. The eligibility of mother–infant dyads was assessed at 2 time points: week 2 and week 4 postbirth.
Initial selection criteria for women at week 2 postpartum were aged between 18 and 49 y, having intention to breastfeed for ≥75% of feeds for ≥3 mo, and residing within the Cincinnati metro area. Eligibility criteria for infants at week 2 were being singleton term and healthy, with a birthweight of ≥2500 g. Mothers who did not achieve ≥75% of feeds as breast milk by week 4 were excluded from further follow-up and replaced until a sample size of 120 eligible dyads was reached. The institutional review board of Cincinnati Children’s Hospital Medical Center approved the GEHM study, and all mothers provided informed consent in the original study [32]. At week 4 of the Cincinnati GEHM study, 120 dyads were deemed eligible to continue with follow-up measurements.
Primary exposures: prepregnancy BMI and maternal DII score
Maternal sociodemographic and clinical data, including maternal age, race, educational level, height, and self-reported prepregnancy weight, and infant birth weight and gestational age were collected via interview at the baseline GEHM visit. Maternal height was measured at the baseline visit at 2 wk postpartum using a Road Rod Portable Stadiometer (Hopkins Medical Products) to the nearest 0.1 cm. BMI was calculated using maternal recall of prepregnancy body weight and measured maternal height. We imputed BMI values for the 2 participants who had missing height values. To do this, we used the stepAIC() function in R for stepwise variable selection of sociodemographic factors and available anthropometric measurements to predict maternal BMI in the full sample [33]. The final linear regression model included prepregnancy weight, gestational weight gain, and total DII score (R2 = 0.93) in the interpolation of maternal prepregnancy BMI for these 2 participants.
Maternal dietary data were collected via 24-h dietary recalls administered on 3 random days between weeks 4 and 13 postpartum during in-person visits or phone interviews, and these 3 random days served as a proxy of postpartum dietary patterns, including dietary patterns in the period leading up to milk sample collection. Trained interviewers used the USDA 5-step automated multiple-pass method to improve completeness and portion size accuracy of dietary recall [9]. Nutrient analysis was performed using Nutrition Data Systems for Research (NDSR; Nutrition Coordinating Center, University of Minnesota, MN, United States) software, 2013 version. The mean intake of 3-days for each dietary component was calculated to obtain estimated usual daily intake.
The DII score for each mother was computed using a combination of the raw nutrient and food group intake data generated in NDSR. The 45 possible dietary parameters of the DII algorithm and our steps in calculating a DII score for each participant are summarized in Figure 1 and detailed in the Supplemental Information; see Shivappa et al. [20] for a detailed explanation of the DII scoring algorithm and how it was derived. The DII score typically ranges from −8.87 (anti-inflammatory) to +7.98 (proinflammatory). There is no strict requirement to use all 45 possible dietary parameters. In our analysis, we included 40 dietary parameters and excluded 5 (alcohol, eugenol, turmeric, saffron, and isoflavones) due to lack of data on these components (Figure 1). Of the 40 included dietary parameters, 33 were individual nutrients and 7 foods/ingredients (onion, garlic, ginger, pepper, green/black tea, thyme/oregano, and rosemary), all available in NDSR 2013 for assessment of intake (Supplemental Information).
FIGURE 1.
Dietary inflammatory index (DII) parameters, calculation, and interpretation. An overview of the DII algorithm. DII has 45 parameters, and the steps to compute DII score is summarized in this figure. Having a lower DII score indicates having a more anti-inflammatory diet, and having a higher DII score indicates having a more proinflammatory diet. In our analysis, 40 different parameters were used in the calculation of the DII score.
Primary outcome variable: quantification of CRP concentrations in human milk
At each in-person visit, mothers provided a milk sample from a complete breast expression (i.e., the entire contents of 1 breast, totaling ∼60–80 mL) using a hospital-grade electric breast pump (Medela Symphony; Medela). Milk samples were collected between 09:00 and 13:00pm from the breast that had not been used to feed their infant for ≥2 h. Collected samples were transferred into labeled 2-mL cryogenic storage vials and stored at −80 °C [9].
Prior to quantifying CRP concentrations in human milk, the first author adapted and validated a protocol using a high-sensitivity ELISA kit from R&D Systems (catalog number DCRP00) to account for human milk rather than serum or urine being the biofluid. Protocols validated for measuring bioactive proteins in serum or urine are often used to quantify these same compounds in human milk, but it is not always known if this practice is valid given the potential matrix effects of human milk on ELISA assay performance [34].
For adaptation and validation of the chosen high-sensitivity CRP ELISA kit, we used the skim fraction of pooled donor milk given that CRP is a water-soluble protein [35], which would also mitigate the lipid-soluble matrix interferences with the assay. First, we generated a standard curve that extended 100% beyond the lower and upper range of the kit to detect the upper and lower limit of quantification of the assay using the CRP standard provided by manufacturer of the assay (R&D Systems). Then, we generated a standard curve of pooled milk samples spiked with standard with the concentration of 50 ng/mL. The standard curve of serially diluted milk sample that was spiked with CRP standard exhibited linearity across the full range of CRP standard dilutions assayed. Next, we determined the optimal dilution factor for human milk. Based on a literature search, the concentration of CRP in human milk is much lower than that in serum [35]. Thus, we tested assaying milk samples with 5 different dilution factors (1:99, 1:9, 1:4, 1:3, 1:2, and undiluted), in contrast to the kit’s recommended 100-fold dilution for serum samples. To validate the adapted protocol, we calculated the intra-assay precision across 4 replicates, and percent recovery of milk samples spiked with a CRP standard at the concentration of 25 ng/mL across 14 replicates at the final dilution factor. In our ELISA validity experiments, we identified a 1:4 dilution as being optimal for assaying CRP in human milk based on CRP concentrations as this dilution most consistently fell along the linear portion of the assay’s standard curve. Dilution factor–corrected mean ± SD and coefficient of variation for CRP concentration across 4 replicates of a pooled milk sample assayed at 1:4 dilution was 139.25 ± 2.17 ng/mL and 1.56%, respectively. The spike recovery of CRP in human milk ranged from 89.9% to 120.2% across 14 replicates, with a coefficient of variation of 4.27% (Supplemental Tables 1 and 2).
Once we validated our adapted protocol for assay of milk CRP, we used the same kit to assay week 4 milk samples in duplicate. Whole milk samples collected at week 4 were thawed and centrifuged at 4000 × g for 10 minutes at 4 °C, and the milk fat layer was removed to use skim milk for the CRP assay. We followed the manufacturer’s protocol, except that the dilution factor was 1:4. Specifically, skim milk samples were diluted 1:4 by mixing 10 μL skim milk and 40 μL kit-provided diluent reagent per well. After calculating the assayed concentration based on the optical density reading, we multiplied this result by 5 to obtain the skimmed milk CRP concentration given that we used a 5-fold dilution.
Statistical analysis
Maternal and infant characteristics were presented using means ± SD or medians (quartile 1, quartile 3) for continuous variables and frequencies (proportion) for categorical variables. One-way analysis of variance or Kruskal–Wallis tests and χ2 or Fisher exact tests (when cell sizes ≤5) were used to test for differences in these variables across DII and BMI groups.
To test the main effects of maternal DII and prepregnancy BMI and their interaction with the milk CRP concentrations, we used linear regression modeling with orthogonal 2 × 2 factorial design, with and without adjustment for maternal age, maternal education, and delivery mode. To perform these analyses, participants were categorized into low and high DII groups at the median DII score for the sample and categorized as those with normal (<25) or elevated (≥25) BMI based on the WHO cutoff for healthy BMI [36].
To evaluate the difference in distributions of CRP concentrations, we performed the crossclassification analysis across the 4 combinations of low/high DII and normal/elevated BMI groups as defined earlier, using the Kruskal–Wallis test. Then, in post hoc analyses, we compared the group with low DII/normal BMI with the 3 other groups combined using the Kruskal–Wallis test and with each of the 3 other groups separately using Dunn test with Bonferroni correction.
Data analysis was performed using the R software, version 4.4.1 (R Core Team) [37]. The MASS package (version 7.3.60.0.1) was used to perform variable selection for missing data imputation [33]. Statistical significance was set as a P value of <0.05, with the exception that a P value of <0.0167 was considered statistically significant for multiple comparisons with Bonferroni correction.
Results
Participant characteristics
Within the Cincinnati subcohort of the GEHM study, 120 mother–infant dyads were eligible for continued follow-up at week 4 postpartum, and 100 of these participants had dietary data and milk samples available. The DII score ranged from −3.37 to 7.02, with a median (Q1, Q3) of 0.13 (−1.54, 1.27) units. Mean ± SD maternal age was 31.85 ± 5.16 years at delivery, and median (Q1, Q3) prepregnancy BMI was 25.37 (22.77, 30.38). The proportion of Black and White women in our sample were 12.0% and 86%, respectively, with 2% Native Hawaiian or other Pacific Islander (Supplemental Table 3). Table 1 summarizes maternal and infant characteristics stratified by DII and BMI groups. Prepregnancy BMI and total DII score differed significantly across the 4 combinations of DII and BMI groupings, but no other maternal or infant characteristic reached statistical significance when compared across the 4 groups, with the exception of delivery mode.
TABLE 1.
Summary of maternal and infant characteristics across DII and BMI groups1.
| Maternal and Infant characteristics2 | Low DII/normal BMI, n = 30 | Low DII/elevated BMI, n = 20 | High DII/normal BMI, n = 17 | High DII/elevated BMI, n = 33 | P |
|---|---|---|---|---|---|
| Total DII score | −1.61 (−2.24, −0.83) | −1.54 (−2.40, −0.76) | 1.17 (0.81, 1.89) | 1.36 (0.53, 2.30) | <0.001 |
| Prepregnancy BMI (kg/m2) | 22.30 (21.53, 23.58) | 29.88 (26.66, 36.13) | 23.26 (21.52, 24.15) | 30.12 (27.01, 32.64) | <0.001 |
| Maternal delivery age (y) | 32.05 ± 5.16 | 30.20 ± 3.93 | 31.11 ± 4.53 | 33.06 ± 5.94 | 0.52 |
| Maternal race | 0.13 | ||||
| Black | 3 | 0 | 2 | 7 | |
| White | 26 | 20 | 14 | 26 | |
| Others | 1 | 0 | 1 | 0 | |
| Maternal education | 0.24 | ||||
| High school graduate | 0 | 2 | 1 | 2 | |
| College degree | 24 | 14 | 12 | 29 | |
| Postcollege education | 6 | 4 | 4 | 2 | |
| Infant sex: male, n (%) | 16 (53.3) | 9 (45) | 8 (47.1) | 20 (60.6) | 0.42 |
| Infant gestational age (wk) | 39.8 ± 1.1 | 39.7 ± 0.9 | 39.2 ± 1.1 | 39.3 ± 1.0 | 0.14 |
| Infant birthweight (kg) | 3.6 ± 0.5 | 3.6 ± 0.4 | 3.5 ± 0.4 | 3.6 ± 0.4 | 0.74 |
| Delivery mode: vaginal, n (%) | 27 (90.0) | 12 (60.0) | 16 (94.1) | 25 (75.8) | 0.03 |
Abbreviations: DII, dietary inflammation index.
Participants (n = 100) were categorized into low and high DII groups at the median and BMI groups defined as normal (<25) or elevated (≥25).
Values are mean ± SD or median (quartile 1, quartile 3) for parametric and nonparametric continuous variables, respectively, or n (%) for frequency (proportion) of categorical variables. Based on the normality of data, differences in continuous variables across the groups were compared with either 1-way analysis of variance test or Kruskal–Wallis test. Proportions of categorical variables across the DII and BMI groups were compared with Pearson χ2 test or Fisher exact tests in instances with cell sizes ≤5. Statistical significance was set as a P value of <0.05.
Human milk CRP concentrations
Human milk CRP concentrations (n = 100) ranged from 3.1 to 477.4 ng/mL, with a median (Q1, Q3) of 101.7 (54.8, 172.3) ng/mL. Table 2 summarizes the linear regression results. In the final multivariable linear regression model that included adjustment for maternal delivery age, maternal education, and delivery mode, the main effect of BMI ≥ 25 was significantly associated with higher human milk CRP concentrations than that of BMI < 25 (β ± SE: 96.19 ± 27.23 ng/mL; 95% CI: 41.87, 152.15; P = 0.0006), and the main effect of maternal DII score ≥ 0.13 (high DII) compared with DII < 0.13 (low DII) was not significantly associated with higher milk CRP concentrations (β ± SE: 35.40 ± 27.22 ng/mL; 95% CI: −18.03, 91.33; P = 0.197); furthermore, this model revealed a significant negative interaction between maternal DII and BMI groups in their joint association with milk CRP concentration (β ± SE: −80.84 ± 38.02 ng/mL; 95% CI: −161.45, −7.82; P = 0.036). The strength of association of prepregnancy BMI as a predictor of CRP was shown to be large in the low DII group (Supplemental Figure 1).
TABLE 2.
Associations of maternal DII score and prepregnancy BMI with human milk CRP concentrations at week 4 postpartum (n = 100).
| Maternal characteristic | Human milk CRP concentration (ng/mL) |
|||||
|---|---|---|---|---|---|---|
| Model 1 |
Model 2 |
|||||
| β ± SE | 95% CI (2.5%, 97.5%) | P | β ± SE | 95% CI (2.5%, 97.5%) | P | |
| DII score, ≥0.13 | 33.70 ± 27.05 | −20.01, 87.40 | 0.216 | 35.40 ± 27.22 | −18.03, 91.33 | 0.197 |
| BMI, ≥25 | 91.11 ± 25.73 | 40.05, 142.18 | 0.0006 | 96.19 ± 27.23 | 41.87, 152.15 | 0.0006 |
| DII × BMI interaction | −72.63 ± 37.01 | −146.10, 0.83 | 0.053 | −80.84 ± 38.02 | −161.45, −7.82 | 0.036 |
Estimates were derived from linear regression analysis and reported as β coefficient ± SE and 95% CI (2.5%, 97.5). The model with predictors of maternal DII and BMI was statistically significant from the model without any predictors (F test = 4.44; P = 0.0058). For the orthogonal 2 × 2 factorial design in the regression models, DII < 0.13 and BMI < 25 were taken as reference. Model 1: multivariate linear regression for maternal prepregnancy BMI, DII score, and their interaction. Model 2: Model 1 was adjusted for maternal age, maternal education, and delivery mode. Statistical significance was set as a P value < 0.05.
Abbreviations: CRP, C-reactive protein; DII, dietary inflammation index.
In crossclassification analysis, human milk CRP concentrations were significantly different across the 4 combinations of DII and BMI groups (P = 0.009) (Table 3). In post hoc analysis, women with low DII/normal BMI had significantly lower milk CRP [n = 30; 64.3 (38.2, 121.4) ng/mL] than those in the other 3 groups combined [n = 70; 124.1 (71.2, 181) ng/mL; P = 0.022]. CRP concentrations for those with low DII/elevated BMI (n = 20), high DII/normal BMI (n = 17), and high DII/elevated BMI (n = 32) were 152.4 (96.4, 205.9), 110.4 (82.5, 163.2), and 107.7 (54.3, 186.1) ng/mL, respectively (Table 3). Multiple comparisons of the low DII/ normal BMI group with each of the other 3 groups revealed a statistically significant difference for all 3 comparisons; however, the only comparison that remained significant after Bonferroni adjustment was with the low DII/elevated BMI group (adjusted P = 0.0057) (Figure 2).
TABLE 3.
Comparison of CRP concentrations in human milk across DII and BMI groups1.
| Comparison across all 4 DII × BMI groups2 |
Post hoc analysis2,3 |
|||||
|---|---|---|---|---|---|---|
| DII × BMI category | n (%) | Milk CRP (ng/mL), median (Q1, Q3) | P | n (%) | Milk CRP (ng/mL), median (Q1, Q3) | P |
| Low DII normal BMI | 30 (30) | 64.3 (38.2, 121.4) | 0.0096 | 30 (30) | 64.3 (38.2, 121.4) | 0.0022 |
| Low DII elevated BMI | 20 (20) | 152.4 (96.4, 205.9) | 70 (70) | 123.1 (73.0,180.9) | ||
| High DII normal BMI | 17 (17) | 110.4 (82.5, 163.2) | ||||
| High DII elevated BMI | 33 (33) | 107.5 (54.7, 180.4) | ||||
Abbreviations: CRP, C-reactive protein; DII, dietary inflammation index.
Statistical significance was set as a P value < 0.05.
Participants (n = 100) were categorized into low and high DII groups at the median and BMI groups defined as normal (<25) or elevated (≥25).
Group differences were compared with the Kruskal–Wallis test.
For post hoc analysis, milk CRP concentrations in women with both low DII score and normal BMI (n = 30) were compared with the other 3 categories combined (n = 70).
FIGURE 2.
Distribution of milk CRP concentrations across DII and BMI categories. Box plots showing medians and quartiles of milk CRP concentrations (ng/mL) by DII and BMI groups. Differences in milk CRP concentrations were compared using the Dunn’s test with Bonferroni adjustment. Differing letters above each group denotes significant difference prior to Bonferroni adjustment (p < 0.05). The group with both low DII and normal BMI had the lowest milk CRP concentrations, but after Bonferroni adjustment, it was only significantly different from the group with low DII and elevated BMI (95% CI: −49.25, −5.05; adjusted P = 0.0057). Statistical significance was set as an adjusted P value of <0.0167 for multiple comparisons of the group with both low DII and normal BMI with each other 3 group with Bonferroni correction. CRP, C-reactive protein; DII, dietary inflammatory index.
Discussion
This study examined the joint association of maternal dietary inflammatory potential and prepregnancy BMI with milk CRP concentrations in a Midwestern cohort. We used DII to compute the inflammatory potential of maternal diet. DII score in our study ranged from −3.37 to +7.02. Based on >160 articles published between 2014 and 2018, Hebert et al. [38] reported that the DII score generally ranges from −5.5 to +5.5 when derived from 25 to 30 parameters. Our analysis used 40 parameters, and our findings align with the effective range of DII provided by its developers [38], indicating that our method for scoring is valid.
The CRP concentrations in the week 4 GEHM milk samples reported in this study—median of 101.7 ng/mL with an IQR of 54.8–172.3—indicates a higher concentration and much wider range of human milk CRP than previously reported [18,19]. For example, Whitaker et al. [19] reported a median CRP of 24.5 and an IQR of 11.2–67.1 ng/mL for week 4 samples using an ELISA assay. Although differences may be explained by sampling different populations, we speculate that our wider range may be explained by our ability to quantify milk CRP at the edges of the limit of detection by adapting, optimizing, and validating a high-sensitivity ELISA kit for use with human milk.
In our study, the BMI × DII interaction term was statistically significant, highlighting effect modification and revealing a negative interaction. With an anti-inflammatory diet (low DII), prepregnancy BMI alone is a significant predictor of milk CRP. However, in the presence of a more proinflammatory diet (high DII), BMI does not contribute further to milk CRP concentrations.
This result aligns with the crossclassification analysis shown in Table 3, where mothers with both an anti-inflammatory dietary pattern and prepregnancy BMI of <25 had median milk CRP concentrations that were only 42%–60% of the concentrations in the other 3 combinations of DII and BMI. Our findings align with 2 previous studies that reported a positive association between maternal prepregnancy BMI and milk CRP concentrations [18,19]. However, to our knowledge, the modifying role of maternal diet quality on milk CRP concentrations has not been previously published. Ramiro-Cortijo et al. [39] found no significant association between postpartum DII score and other cytokines in human milk, including IL-8, IL-10, IL-13, and TNF-α, but did not study CRP. However, the study by Ramiro-Cortijo et al. [39] used only 20 parameters for DII scoring, which may have limited their ability to capture dietary inflammatory potential, and they did not report whether their assay protocols were validated for use in human milk. Furthermore, having only 35 dyads with milk cytokine assessment, the potential interaction of maternal BMI with diet quality was not considered.
Elevated CRP concentrations in the milk of mothers with elevated BMI and/or a proinflammatory dietary pattern can be explained by higher uptake of CRP from circulation by the mammary glands among mothers with higher blood CRP concentrations. This is because obesity can lead to chronic systemic inflammation due to increased release of proinflammatory cytokines such as TNF-α and IL-6 from adipose tissue, which induces synthesis of CRP in hepatocytes [11]. Furthermore, a poor-quality diet may upregulate inflammatory responses via increasing oxidative stress, dysbiosis of gut microbiota, and worsening glycemic control [40]. Given that poor diet quality or elevated BMI can lead to increases in inflammatory markers in circulation [12,14,15], we may speculate that mothers with high diet quality and BMI of <25 might have lower serum CRP concentrations than others. However, we could not assess serum CRP concentrations, and the correlation of CRP concentrations between serum and human milk remains unclear. Although the correlation between DII score and serum concentrations of cytokines has been demonstrated in previous studies of nonlactating populations [[21], [22], [23], [24]], little is known about the effects of maternal dietary inflammation potential on milk inflammatory components. Given the critical need for more research on the association between maternal diet and human milk composition to inform future guidelines, we believe our study contributes to the existing literature by being the first cohort study to investigate the interplay between maternal diet quality and BMI on milk CRP concentrations.
The biological impacts of human milk inflammatory markers on breastfed infants are not yet known. Given that human milk contains both proteases and antiproteases, one can infer that proteases may assist the digestion of proteins in the breastfeeding infant’s immature digestive system, although antiproteases may have evolved to protect the bioactive compounds in human milk from digestion [41]. Although the mechanistic studies that investigate the bioavailability of these bioactive compounds in infant’s digestive system are lacking, a growing number of studies have shown the potential effects of some milk inflammatory markers on infant growth outcomes [42]. Particularly, Sims et al. [18] showed that greater intake of CRP via breastfeeding during the first 9 months of life was positively associated with greater fat mass index in infants of mothers with prepregnancy BMI of ≥25, whereas it was negatively associated with fat-free mass in the infants of mothers with normal prepregnancy BMI. However, our understanding of the long-term effects of CRP intake on breastfed infants is limited.
Limitations and strengths
The present study has its limitations. The generalizability of our findings may be limited due to homogeneity of the study participants regarding breastfeeding status and sociodemographic characteristics: the predominance of participants were non-Hispanic White women. Further, BMI is not an ideal indicator of maternal adiposity and metabolic health [[43], [44], [45]], and we could not assess biological inflammatory markers in serum to confirm chronic inflammation status of mothers with elevated BMI. We were also limited by the measurement of 1 inflammatory marker in milk at 1 time point and maternal BMI only from prepregnancy. We selected CRP because of its ubiquitous associations with low-grade inflammation and milk samples from week 4 postpartum to maximize sample size.
Aside from these limitations, this study has certain strengths. The original GEHM study design was of high quality and included weekly phone surveillance, biweekly study visits, standardized milk collection procedures, and maternal 24-h diet recalls performed by research dietitians based on 3 randomly selected days. Given the scarcity of research on the diet of lactating women in the United States, especially in Midwestern settings, we believe that our study addresses a critical gap in the literature. We included 40 dietary parameters of 45 available DII parameters, strongly supporting characterization of dietary inflammatory potential. Another strength of the study is our novel examination of the joint association of maternal BMI and diet quality on milk composition. Furthermore, using a high-sensitivity ELISA kit with a protocol that we first optimized and validated for use in human milk samples enabled us to examine predictors of milk CRP across a much broader range of values that previously reported for mature milk samples, strengthening the accuracy of our results.
Conclusions
Our study showed that milk CRP concentrations at 4 weeks postpartum were lower in women with both a more anti-inflammatory diet and a normal BMI. Based on our findings, we recommend that future studies examining maternal factors that influence milk composition include assessment of maternal serum inflammatory markers, as well as multiple variables that contribute to low-grade inflammation during lactation, including diet quality and measures of maternal adiposity. Further, our data suggest that all cytokine assays used in human milk research should first be adapted and validated for use in the human milk matrix.
Author contributions
The authors’ responsibilities were as follows – HC, ALM, LAN-R: designed the study, interpreted results for this manuscript, and wrote the article; ALM is the principal investigator of the GEHM study; HC: performed milk C-reactive protein assays and dietary inflammatory index computation and analyzed the data; NJO, SSS, ALM: were responsible for data collection and management; SSS: managed dietary data; NJO, ALM, LAN-R: provided with insight for the statistical analysis approach; NAM, SDM: provided with milk samples to perform the milk C-reactive protein analysis; LAN-R: has primary responsibility for final content; and all authors: read, provided feedback, and approved final manuscript.
Funding
The Mead Johnson Nutrition funded the data and sample collection for the GEHM study. University of Cincinnati University Research Council (URC) Graduate Student Stipend and Research Cost Program for Faculty—Student Collaboration funded HC in summer 2022 to perform laboratory analysis. Reckitt/Mead Johnson Nutrition funded GEHM milk sample collection but study design, sample analysis, interpretation of results, and writing of the manuscript were fully at the discretion of the authors. HC, ALM, LNR are funded by NICHD 1R01HD109915.
Conflict of interest
NAM and SDM are employees of the Reckitt/Mead Johnson Nutrition. ALM reports financial support by Mead Johnson Nutrition. KAD is an employee of Danone Research. The other authors report no conflicts of interest.
Acknowledgments
We thank participants of GEHM study in Cincinnati site and co-investigators of GEHM study for study design, recruitment, data, and sample collection.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.tjnut.2025.06.004.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
References
- 1.Ballard O., Morrow A.L. Human milk composition: nutrients and bioactive factors. Pediatr. Clin. North Am. 2013;60(1):49–74. doi: 10.1016/j.pcl.2012.10.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Gila-Diaz A., Arribas S.M., Algara A., Martin-Cabrejas M.A., Lopez de Pablo A.L., Saenz de Pipaon M., et al. A review of bioactive factors in human breastmilk: a focus on prematurity. Nutrients. 2019;11(6):1307. doi: 10.3390/nu11061307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Dietz W.H. Critical periods in childhood for the development of obesity. Am. J. Clin. Nutr. 1994;59(5):955–959. doi: 10.1093/ajcn/59.5.955. [DOI] [PubMed] [Google Scholar]
- 4.Gillman M.W. Early infancy as a critical period for development of obesity and related conditions. Nestle Nutr. Workshop Ser. Pediatr. Program. 2010;65:13–20. doi: 10.1159/000281141. discussion 20–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gillman M.W. The first months of life: a critical period for development of obesity. Am. J. Clin. Nutr. 2008;87(6):1587–1589. doi: 10.1093/ajcn/87.6.1587. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Nommsen L.A., Lovelady C.A., Heinig M.J., Lonnerdal B., Dewey K.G. Determinants of energy, protein, lipid, and lactose concentrations in human milk during the first 12 mo of lactation: the DARLING Study. Am. J. Clin. Nutr. 1991;53(2):457–465. doi: 10.1093/ajcn/53.2.457. [DOI] [PubMed] [Google Scholar]
- 7.Kugananthan S., Gridneva Z., Lai C.T., Hepworth A.R., Mark P.J., Kakulas F., et al. Associations between maternal body composition and appetite hormones and macronutrients in human milk. Nutrients. 2017;9(3):252. doi: 10.3390/nu9030252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Leghi G.E., Netting M.J., Middleton P.F., Wlodek M.E., Geddes D.T., Muhlhausler A.B.S. The impact of maternal obesity on human milk macronutrient composition: a systematic review and meta-analysis. Nutrients. 2020;12(4):934. doi: 10.3390/nu12040934. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Dingess K.A., Valentine C.J., Ollberding N.J., Davidson B.S., Woo J.G., Summer S., et al. Branched-chain fatty acid composition of human milk and the impact of maternal diet: the Global Exploration of Human Milk (GEHM) Study. Am. J. Clin. Nutr. 2017;105(1):177–184. doi: 10.3945/ajcn.116.132464. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Calder P.C., Ahluwalia N., Brouns F., Buetler T., Clement K., Cunningham K., et al. Dietary factors and low-grade inflammation in relation to overweight and obesity. Br. J. Nutr. 2011;106(Suppl 3):S5–S78. doi: 10.1017/S0007114511005460. [DOI] [PubMed] [Google Scholar]
- 11.Ellulu M.S., Patimah I., Khaza'ai H., Rahmat A., Abed Y. Obesity and inflammation: the linking mechanism and the complications. Arch. Med. Sci. 2017;13(4):851–863. doi: 10.5114/aoms.2016.58928. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Park H.S., Park J.Y., Yu R. Relationship of obesity and visceral adiposity with serum concentrations of CRP, TNF-alpha and IL-6. Diabetes Res. Clin. Pract. 2005;69(1):29–35. doi: 10.1016/j.diabres.2004.11.007. [DOI] [PubMed] [Google Scholar]
- 13.Esposito K., Marfella R., Ciotola M., Di Palo C., Giugliano F., Giugliano G., et al. Effect of a Mediterranean-style diet on endothelial dysfunction and markers of vascular inflammation in the metabolic syndrome: a randomized trial. JAMA. 2004;292(12):1440–1446. doi: 10.1001/jama.292.12.1440. [DOI] [PubMed] [Google Scholar]
- 14.Esmaillzadeh A., Kimiagar M., Mehrabi Y., Azadbakht L., Hu F.B., Willett W.C. Dietary patterns and markers of systemic inflammation among Iranian women. J. Nutr. 2007;137(4):992–998. doi: 10.1093/jn/137.4.992. [DOI] [PubMed] [Google Scholar]
- 15.Chrysohoou C., Panagiotakos D.B., Pitsavos C., Das U.N., Stefanadis C. Adherence to the Mediterranean diet attenuates inflammation and coagulation process in healthy adults: the ATTICA Study. J. Am. Coll. Cardiol. 2004;44(1):152–158. doi: 10.1016/j.jacc.2004.03.039. [DOI] [PubMed] [Google Scholar]
- 16.Serrano-Martinez M., Palacios M., Martinez-Losa E., Lezaun R., Maravi C., Prado M., et al. A Mediterranean dietary style influences TNF-alpha and VCAM-1 coronary blood levels in unstable angina patients. Eur. J. Nutr. 2005;44(6):348–354. doi: 10.1007/s00394-004-0532-9. [DOI] [PubMed] [Google Scholar]
- 17.Dawod B., Marshall J.S. Cytokines and soluble receptors in breast milk as enhancers of oral tolerance development. Front. Immunol. 2019;10:16. doi: 10.3389/fimmu.2019.00016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sims C.R., Lipsmeyer M.E., Turner D.E., Andres A. Human milk composition differs by maternal BMI in the first 9 months postpartum. Am. J. Clin. Nutr. 2020;112(3):548–557. doi: 10.1093/ajcn/nqaa098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Whitaker K.M., Marino R.C., Haapala J.L., Foster L., Smith K.D., Teague A.M., et al. Associations of maternal weight status before, during, and after pregnancy with inflammatory markers in breast milk. Obesity (Silver Spring) 2017;25(12):2092–2099. doi: 10.1002/oby.22025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Shivappa N., Steck S.E., Hurley T.G., Hussey J.R., Hebert J.R. Designing and developing a literature-derived, population-based dietary inflammatory index. Public Health Nutr. 2014;17(8):1689–1696. doi: 10.1017/S1368980013002115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Shivappa N., Hebert J.R., Rietzschel E.R., De Buyzere M.L., Langlois M., Debruyne E., et al. Associations between dietary inflammatory index and inflammatory markers in the Asklepios Study. Br. J. Nutr. 2015;113(4):665–671. doi: 10.1017/S000711451400395X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Ren Z., Zhao A., Wang Y., Meng L., Szeto I.M., Li T., et al. Association between dietary inflammatory index, C-reactive protein and metabolic syndrome: a cross-sectional study. Nutrients. 2018;10(7):831. doi: 10.3390/nu10070831. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Vahid F., Shivappa N., Faghfoori Z., Khodabakhshi A., Zayeri F., Hebert J.R., et al. Validation of a dietary inflammatory index (DII) and association with risk of gastric cancer: a case-control study. Asian Pac. J. Cancer Prev. 2018;19(6):1471–1477. doi: 10.22034/APJCP.2018.19.6.1471. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Wirth M.D., Burch J., Shivappa N., Violanti J.M., Burchfiel C.M., Fekedulegn D., et al. Association of a dietary inflammatory index with inflammatory indices and metabolic syndrome among police officers. J. Occup. Environ. Med. 2014;56(9):986–989. doi: 10.1097/JOM.0000000000000213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Zhang Z., Wu Y., Zhong C., Zhou X., Liu C., Li Q., et al. Association between dietary inflammatory index and gestational diabetes mellitus risk in a prospective birth cohort study. Nutrition. 2021;87–88 doi: 10.1016/j.nut.2021.111193. [DOI] [PubMed] [Google Scholar]
- 26.Yang Y., Kan H., Yu X., Yang Y., Li L., Zhao M. Relationship between dietary inflammatory index, hs-CRP level in the second trimester and neonatal birth weight: a cohort study. J. Clin. Biochem. Nutr. 2020;66(2):163–167. doi: 10.3164/jcbn.19-100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Sen S., Rifas-Shiman S.L., Shivappa N., Wirth M.D., Hebert J.R., Gold D.R., et al. Dietary inflammatory potential during pregnancy is associated with lower fetal growth and breastfeeding failure: results from Project Viva. J. Nutr. 2016;146(4):728–736. doi: 10.3945/jn.115.225581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Kant A.K., Graubard B.I. Secular trends in regional differences in nutritional biomarkers and self-reported dietary intakes among American adults: National Health and Nutrition Examination Survey (NHANES) 1988-1994 to 2009-2010. Public Health Nutr. 2018;21(5):927–939. doi: 10.1017/S1368980017003743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Cusack L.K., Smit E., Kile M.L., Harding A.K. Regional and temporal trends in blood mercury concentrations and fish consumption in women of child bearing age in the United States using NHANES data from 1999-2010. Environ. Health. 2017;16(1):10. doi: 10.1186/s12940-017-0218-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.U.S. Department of Agriculture; U.S. Department of Health and Human Services . USDA; Washington, DC: 2020. Dietary Guidelines for Americans, 2020-2025. [Google Scholar]
- 31.DGA Committee . U.S. Department of Agriculture; Washington, DC: 2020. Scientific report of the 2020 Dietary Guidelines Advisory Committee: advisory report to the Secretary of Agriculture and the Secretary of Health and Human Services. [Google Scholar]
- 32.Woo J.G., Guerrero M.L., Ruiz-Palacios G.M., Peng Y.M., Herbers P.M., Yao W., et al. Specific infant feeding practices do not consistently explain variation in anthropometry at age 1 year in urban United States, Mexico, and China cohorts. J. Nutr. 2013;143(2):166–174. doi: 10.3945/jn.112.163857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Venables W.N., Ripley B.D. 4th ed. Springer; New York: 2002. Modern applied statistics with S. [Google Scholar]
- 34.Suwaydi M.A., Gridneva Z., Perrella S.L., Wlodek M.E., Lai C.T., Geddes D.T. Human milk metabolic hormones: analytical methods and current understanding. Int J Mol Sci. 2021;22(16) doi: 10.3390/ijms22168708. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Fujimori M., Franca E.L., Fiorin V., Morais T.C., Honorio-Franca A.C., et al. Changes in the biochemical and immunological components of serum and colostrum of overweight and obese mothers. BMC Pregnancy Childbirth. 2015;15:166. doi: 10.1186/s12884-015-0574-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.World Health Organization A healthy lifestyle - WHO recommendations. https://www.who.int/europe/news-room/fact-sheets/item/a-healthy-lifestyle---who-recommendations [Internet]. 2010 [cited 2025 Jan 13]. Available from:
- 37.R Core Team, R: R Foundation for Statistical Computing; Vienna, Austria: 2023. a language and environment for statistical computing. [Google Scholar]
- 38.Hebert J.R., Shivappa N., Wirth M.D., Hussey J.R., Hurley T.G. Perspective: the dietary inflammatory index (DII)—lessons learned, improvements made, and future directions. Adv. Nutr. 2019;10(2):185–195. doi: 10.1093/advances/nmy071. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Ramiro-Cortijo D., Herranz Carrillo G., Singh P., Rebollo-Hernanz M., Rodriguez-Rodriguez P., Ruvira S., et al. Maternal and neonatal factors modulating breast milk cytokines in the first month of lactation. Antioxidants (Basel) 2023;12(5):996. doi: 10.3390/antiox12050996. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Minihane A.M., Vinoy S., Russell W.R., Baka A., Roche H.M., Tuohy K.M., et al. Low-grade inflammation, diet composition and health: current research evidence and its translation. Br. J. Nutr. 2015;114(7):999–1012. doi: 10.1017/S0007114515002093. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Dallas D.C., Murray N.M., Gan J. Proteolytic systems in milk: perspectives on the evolutionary function within the mammary gland and the infant. J. Mammary Gland Biol. Neoplasia. 2015;20(3–4):133–147. doi: 10.1007/s10911-015-9334-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Brockway M.M., Daniel A.I., Reyes S.M., Gauglitz J.M., Granger M., McDermid J.M., et al. Human milk bioactive components and child growth and body composition in the first 2 years: a systematic review. Adv. Nutr. 2024;15(1) doi: 10.1016/j.advnut.2023.09.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Somi M.H., Nikniaz Z., Ostadrahimi A., Eftekhar Sadat A.T., Faramarzi E. Is normal body mass index a good indicator of metabolic health in Azar cohort population? J. Cardiovasc. Thorac. Res. 2019;11(1):53–60. doi: 10.15171/jcvtr.2019.09. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Rubino F., Cummings D.E., Eckel R.H., Cohen R.V., Wilding J.P.H., Brown W.A., et al. Definition and diagnostic criteria of clinical obesity. Lancet Diabetes Endocrinol. 2025 doi: 10.1016/S2213-8587(24)00316-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Carbone S., Canada J.M., Billingsley H.E., Siddiqui M.S., Elagizi A., Lavie C.J. Obesity paradox in cardiovascular disease: where do we stand? Vasc. Health Risk Manag. 2019;15:89–100. doi: 10.2147/VHRM.S168946. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.