Pervasive Influence of Hormonal Contraceptives on the Human Plasma Proteome in a Broad Population Study

Abstract

Large-scale plasma proteomics in population studies offer new avenues to explore non-genetic associations, such as biomarkers for lifestyle and environmental exposure. Moreover, through longitudinal monitoring, they help in characterizing the onset and progression of diseases. Utilizing the SWATH DIA-MS-based proteomic approach with 5-minute chromatographic gradients, we examined the neat plasma proteome of 3,632 participants from the Cooperative Health Research in South Tyrol (CHRIS) study. We present precise protein quantities for 148 highly abundant plasma protein groups. Many of these proteins are challenging to quantify using affinity-based reagents. This set of highly abundant plasma proteins includes immunoglobulins, coagulation factors, metabolic proteins, and components of the innate immune system, such as the complement system. Our findings indicate that, beyond age and sex, the most significant biological variable affecting the plasma proteome was the use of hormonal contraceptives. The signature from this drug class was more pronounced compared to other common medications, which influenced only a subset of the measured proteins and exhibited smaller effect sizes. Notably, Angiotensinogen (AGT) levels were strongly influenced by contraceptive use, suggesting that AGT levels might serve as a potential biomarker for contraceptive usage. Due to its high prevalence, and apparently strong influence, this type of medication should be accounted for in any epidemiological or clinical study of the plasma proteome to avoid spurious findings.

Menopause Hormone Replacement Therapy and Lifestyle Factors affect Metabolism and Immune System in the Serum Proteome of Aging Individuals

“…We recently reported that HCU has a major impact on the high abundant fraction of the plasma proteome, in women under 40 years enrolled in the CHRIS study 21 , and as aforementioned, we detected a similar signature in the young women within the BASE-II cohort. With the focus on aging, we herein explored the effect of menopausal hormonal replacement therapy (MHT) in older women participating in the BASE-II.…”

“…Large-scale serum and plasma proteome studies 21,43,80,81 have revealed disease prevalence, risk factors, and healthcare needs across diverse populations. These studies have identified putative biomarkers of aging 1,[21][22][23][24]80 , common lifestyle factors like Type II Diabetes, and administered drugs 43,82,83 . Advancing age significantly heightens the risk for various chronic ailments alongside physical and cognitive decline, ultimately leading to mortality.…”

“…While most of the proteins are normally distributed, some proteins, such as lipoprotein (a) (Lp(a)) and complement factor H-related protein 1 (CFHR1), exhibited tailed or almost bimodal distributions (Figures S59, S158). Tailed distributions in certain proteins, particularlypronounced among the younger age group, were found for proteins that we recently associated to the intake of hormonal contraceptives in younger women (including angiotensinogen (AGT), fetuin B (FETUB), sex hormone binding globulin (SHBG), plasminogen (PLG) or peptidoglycan recognition protein 2 (PGLYRP2)) 21 . (Figures S2 -S254 for Log2 transformed protein abundance distributions of all proteins across the entire cohort, as well as stratified by age groups).…”

Phenotypic spectrum ofFAM47E-SHROOM3haplotype composition in a general population sample

“…28 Data generation and processing was described elsewhere. 29 After merging with the genetic data, 3535 participants remained for the analysis ( Suppl. Tab.…”

Role of the complement system in Long COVID

“…When comparing the mass-spectrometry based proteomes of the control (non-PACS) and the PACS group without balancing for these demographic factors, CS components indeed appear to be up-regulated in PACS (Figure 1A). However, the CS is known to be age and BMI dependent ( 7, 8 ). In their data analysis, Cervia-Hasler et al attempt a correction for demographic effects using linear regression ( 5 ).…”