Background/Purpose: Systemic inflammation in autoimmune rheumatic diseases (ARDs) may alter human milk composition, akin to changes observed in other inflammatory states, indirectly impacting infant health, yet data are nearly non-existent.

Methods: We utilized the Human Milk Biorepository (HMB), located at UC San Diego, to compare milk composition, specifically fat, carbohydrate, protein, and calories, between women with rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE) and healthy controls matched on infant and maternal age and infant sex.  Descriptive characteristics were compared using Fisher’s exact test for categorical and Mann-Whitney U test for continuous variables. After trimming outliers to normalize the distribution of macronutrients, multivariable linear regression analyses were conducted to estimate the coefficients and 95% confidence intervals (CIs) with ARDs as the exposure and macronutrient composition as the outcome. Model a was adjusted for maternal age, child age, and infant sex; Model b additionally adjusted for body mass index, parity, frequency of feeds, time of milk collection, and hours since the last feed; and Model c incorporated all these variables and the milk storage length of the milk samples.

Results: We analyzed samples from 74 mothers with ARDs (13 SLE, 58 RA, and three with both) and 31 healthy controls. The median age of mothers in the ARD and non-ARD groups were 33.85 and 35.01 years, respectively. The children had a median age of 4.8 months at the time of milk collection. Notably, the median storage duration of the analyzed samples was significantly longer in the ARD group (53 months) compared to 3 months in controls (Table 1). The median (Interquartile range [IQR]) levels of nutrients for ARD vs control groups were: carbohydrate [8.57 (7.80-8.99) vs 9.02 (8.73-9.31) g/dl, p=0.001], fat [3.62 (2.76-4.60) vs 3.45 (2.82-4.25) g/dl, p=0.94], protein [0.88 (0.64-1.09) vs 0.55 (0.37-0.69) g/dl, p=0.0004] and energy [70.81 (61.47-80.21) vs 70.74 (70.74-78.89) kCal/dl, p=0.78], respectively.

The unadjusted linear regression model revealed higher average protein content in ARD by 0.26 g/dl (95% CI 0.73 – 0.45), which remained elevated in Models a and b. However, in the fully adjusted Model c, there was no difference in milk protein between ARD and non-ARD groups (p = 0.19). Conversely, the levels of carbohydrates were consistently lower in the ARD group, by 0.71 g/dL (95% CI -1.19, -0.24) in the fully adjusted model. There were no differences in fat or total energy content between groups (Table 2).

Conclusion: ARDs may alter the nutritional composition of human milk, particularly by reducing carbohydrate content. However, as the carbohydrate levels in ARD groups are still within the normal range reported in the literature the clinical implications require further investigation. Although findings indicated higher protein in the ARD group, the difference diminished after accounting for the length of milk storage, highlighting an area for further study. This research lays the groundwork for detailed research into human milk composition in ARD patients, including for distinct ARD subgroups, and across various ARD treatments.

ACR Meeting Abstracts - https://acrabstracts.org/abstract/exploring-the-impact-of-autoimmune-rheumatic-diseases-on-human-milk-macronutrient-composition/