Association of inflammation biomarkers with food cravings and appetite changes across the menstrual cycle

Abstract

Importance: Premenstrual symptoms, including food cravings, are often a regular complaint among menstruating women. However, existing evidence regarding the biological mechanisms by which these food cravings occur remains unclear. Inflammation may play an essential role in the occurrence of these food cravings before menstruation.

Objective: The purpose of the present study was to examine the associations between inflammatory markers and the risk of moderate/severe food cravings while accounting for changes in hormone levels and stress across the menstrual cycle.

Design, setting, and participants: The BioCycle Study followed women (n = 259) aged 18-44 for two menstrual cycles. Food cravings (via questionnaire) were assessed up to four times per cycle. Each assessment corresponded to menses and mid-follicular, ovulation, and luteal phases of the menstrual cycle. A wide range of cytokine and chemokine levels (hsCRP, GCSF, GMCSF, IL-4, IL-6, RANTES, MIP1B, etc.) were assessed in blood samples collected at up to 8 visits per cycle, with visits timed using fertility monitors.

Main outcomes and measures: Cravings for chocolate, sweets, salty, and other foods, and changes in appetite were determined to estimate the odds of moderate or severe cravings. Associations between inflammatory markers and risk of reporting a moderate/severe craving symptom at each cycle visit was determined using weighted generalized linear models (e.g., marginal structural models). Models were adjusted for age, BMI, and race, as well as time-varying covariates such as estradiol, stress, leptin, and total energy intake, and accounted for repeated measures (i.e., multiple cycles per woman). Both inflammatory markers and reports of cravings were modeled to account for variation at each visit.

Results: An association between higher inflammatory biomarkers such as hsCRP, GCSF, GMCSF, IL-4, IL-6, RANTES, MIP, and increased risk of moderate/severe cravings were identified across the menstrual cycle all risk ratio>1, all CIs range 0.71-2.38. hsCRP retained statistical significance after false discovery rate correction with chocolate, sweet, and salty cravings, while GCSF, GMCSF, IL-6, and RANTES retained significance with chocolate and sweet cravings only.

Conclusion: and Relevance: The results suggest a potential role of inflammation in food cravings and appetite changes across the menstrual cycle.

Keywords: Alcohol intake; Body mass index; Food cravings; Inflammation; Menstrual cycles; Stress.