Study: Nonlinear dynamics of multi-omics profiles during human aging. Image Credit: tomertu / Shutterstock
In a recent study published in the journal Nature Aging, researchers in Singapore and the United States conducted comprehensive profiling of a longitudinal cohort (n = 108) using next-generation multi-omics techniques to reveal the nonlinear dynamics of human aging. The study cohort comprised individuals residing in California between the ages of 25 and 75, followed up for up to 6.8 years (median = 1.7 years).
The study revealed that only 6.6% of molecular markers showed linear age-associated changes, whereas a substantial 81% exhibited nonlinear patterns, highlighting the complexity of the aging process. Molecular markers analyzed during the study revealed that human aging is not a linear process, with chronological ages of around 44 and 60 demonstrating dramatic dysregulation of specific biological pathways, such as alcohol and lipid metabolism during the 40-year transition and carbohydrate metabolism and immune regulation during the 60-year transition. These findings provide unprecedented insights into the pathways (both biological and molecular) associated with human aging and present a significant leap in identifying therapeutic interventions against age-associated chronic diseases.
Background
Aging is defined as the time-related deterioration of physiological functions associated with health and survival. Decades of research have identified that these physiological changes strongly correspond with the risk and incidence of chronic diseases, including diabetes, neurodegeneration, cancers, and cardiovascular diseases (CVDs).
Recent research using next-generation, system-level, high-throughput omics technologies suggests that, unlike previously believed, aging is not a linear process. The study utilized techniques such as transcriptomics, proteomics, metabolomics, and microbiome analysis to uncover the complexity of aging at a molecular level. Specific chronological ages may serve as thresholds corresponding to significant nonlinear metabolism rates and molecular profile alternations. For example, both neurological diseases and CVDs are known to demonstrate substantial spikes in population-level prevalence at ~40 and ~60 years.
Unfortunately, despite this relatively novel knowledge, the literature has hitherto mainly investigated the biology of aging with the assumption that aging is a linear process. This approach has potentially masked mechanistic insights essential for developing therapeutic interventions against age-related diseases, hindering the quest for extended human lifespans and healthier old ages.
About the study
The present study aims to address this gap in the literature by using a battery of deep multi-omics profiling technologies to investigate the specific alternations in biological and molecular pathways associated with different adult age groups. The study was conducted on a cohort of healthy adult volunteers from California, United States (US), between the ages of 25 and 75. Participants were eligible for the study if they lacked a clinical history of chronic diseases, including anemia, CVD, cancer, psychiatric illness, or bariatric surgery.
Baseline data collection included a modified insulin suppression test, fasting plasma glucose (FPG) test, and hemoglobin A1C (HbA1C) test to establish participants' insulin sensitivity, diabetes, and average glucose levels, respectively. Furthermore, participants' body mass indices (BMIs) were recorded at study enrolment and follow-up.
"…5,405 biological samples (including 1,440 blood samples, 926 stool samples, 1,116 skin swab samples, 1,001 oral swab samples and 922 nasal swab samples) were collected. 135,239 biological features (including 10,346 transcripts, 302 proteins, 814 metabolites, 66 cytokines, 51 clinical laboratory tests, 846 lipids, 52,460 gut microbiome taxons, 8,947 skin microbiome taxons, 8,947 oral microbiome taxons and 52,460 nasal microbiome taxons) were acquired, resulting in 246,507,456,400 data points."
The battery of multi-omics tests comprised seven distinct evaluations, namely 1. transcriptomics (using RNA extracted from flash-frozen peripheral blood mononuclear cells [PBMCs]), 2. proteomics (using liquid chromatography of participants' plasma samples), 3. untargeted metabolomics (using plasma-derived metabolite profiles generated via reverse-phase liquid chromatography [RPLC] and hydrophilic interaction chromatography [HILIC]), 4. cytokine data (derived from Luminex-based multiplex assays of participants' plasma), 5. plasma lipidomics (using differential mobility spectrometry), 6. microbiome analysis (using genomic sequencing of participants' stool, skin, oral, and nasal samples), and 7. standard clinical laboratory tests (metabolic panel, complete blood counts, kidney and liver panels, high-sensitivity C-reactive protein [hsCRP], etc.).
Study findings
The included study cohort comprised 108 participants (51.9% female) between the ages of 25 and 75 (median 55.7). Participants were sampled for multi-omics data every 3-6 months (median follow-up period = 1.7 years, maximum = 6.8 years). This rigorous longitudinal analysis allowed the researchers to capture both linear and nonlinear molecular changes associated with aging. Mulit-omics findings highlighted the importance of nonlinear approaches in characterizing biological aging by revealing that of the investigated molecules, only 6.6% demonstrated linear age-associated changes, while 81.0% demonstrated nonlinear patterns.
Importantly, these molecular patterns were surprisingly consistent across all seven multi-omics investigations, suggesting that these changes have deep biological implications. A trajectory clustering analysis approach employed to group molecules by their temporal similarity revealed the presence of three distinct clusters (clusters 5, 2, and 4).
The first comprised a mRNA and autophagy-associated transcriptomics module exhibiting a dramatic increase around 60 years of age. This pathway maintains cellular homeostasis and demonstrates increased aging-related disease risk. The second comprises a phenylalanine metabolism pathway encapsulating serum/plasma glucose and blood urea nitrogen, both of which substantially increase at around age 60, highlighting reduced kidney function and increased CVD risk. The third includes pathways related to caffeine metabolism and unsaturated fatty acid biosynthesis, critical to cardiovascular health.
To better elucidate peaks in microbiome and molecule dysregulation across the adult aging process, researchers employed a modified Differential Expression Sliding Window Analysis (DE-SWAN) algorithm. Analysis findings highlight the presence of two prominent peaks (crests) corresponding to ~40 and ~60 years, consistent across the full range of multi-omics profiles (particularly proteomics). Modules in the first peak were found to be strongly correlated with alcohol and lipid metabolism. In contrast, those in the second peak were strongly correlated with immune dysfunction, kidney function, and carbohydrate metabolism.
Conclusions
The present study highlights the highly nonlinear nature of the biological and molecular processes associated with human aging, as demonstrated by seven distinct multi-omics investigations. The study is noteworthy in that it additionally identifies specific patterns in the aging process that dramatically increase at around 40 and 60 years, corresponding to biologically meaningful dysregulation of alcohol and lipid metabolism (at ~40) and immune dysfunction, kidney performance, and carbohydrate metabolism (at ~60).
"These comprehensive multi-omics data and the approach allow for a more nuanced understanding of the complexities involved in the aging process, which we think adds value to the existing body of research. However, further research is needed to validate and expand upon these findings, potentially incorporating larger cohorts to capture the full complexity of aging."