Allostatic Load: The Role of Chronic Stress in Predicting Early-Onset Cancer Risk

Early-onset cancers, defined as cancers diagnosed in individuals younger than 50 years, have emerged as a growing public health concern worldwide, with incidence rates increasing across multiple tumor types despite advances in screening and prevention. Traditional cancer risk prediction models that rely on genetic factors, family history, or isolated clinical characteristics have struggled to identify younger patients who may be at elevated risk, highlighting the need for more comprehensive approaches that incorporate additional biological and environmental factors.

Allostatic load, a measure of the cumulative physiologic effects of chronic stress on the body, has emerged as a potential biomarker for predicting early-onset cancer risk. Research presented at the 2026 American Society of Clinical Oncology (ASCO) Annual Meeting has taken an important step toward establishing allostatic load as a potential future component of future risk prediction models.

In an interview with Targeted Oncology^TM, Ravi B. Parikh, MD, MPP, associate professor in the Department of Hematology and Medical Oncology at the Emory University School of Medicine and Winship Cancer Institute, discussed this research, exploring the implications of the findings for cancer prevention and risk assessment in community oncology settings.

Targeted Oncology: What was the rationale for conducting this research?

Ravi B. Parikh, MD, MPP: The rationale for conducting our research was largely to develop a better risk prediction model for early-onset cancers, writ large. We know that there have been many more patients screened for cancers across the board—breast, colorectal, lung—but we know that there's been a tremendous epidemic of early-onset cancers, cancers diagnosed before age 50 in the general population, particularly among certain cancers like colorectal cancer and noncolorectal gastrointestinal cancers.

We know that those traditional cancer risk prediction models, ones that are based either on singular genetic variants or on isolated clinical or social risk factors, do quite poorly for predicting early-onset cancer. In some cases, cancers before age 50 are not even screened for unless they have a strong family history, and so…these existing traditional risk prediction models don't explain the fact why early-onset cancer has risen by 79% [globally] over the past 2 decades,² far outpacing the general incidence of cancers that have [described]. So, we sought to use 2 large population-level databases to try to answer that question about whether we could better incorporate other nontraditional risk factors to explain this early-onset cancer [epidemic].

What are some of the key findings that were presented at ASCO?

We found that allostatic load—a physiologic embedding of chronic stress put upon the body, which has been shown to be predictive of poor cancer outcomes among patients who have cancer—is an independent predictor of early-onset cancer risk in the general population in 2 very different population-level databases: one based in the UK using the UK Biobank, and one based in the United States using the US NIH All of Us database. Allostatic load independently explained about 9% to 20% depending on the cohort of early-onset cancer risk, even when accounting for traditional, what we called “exposome-related” or environmental risk factors.¹ We think that highlights how allostatic load may, whether it's causal or not, be a strong predictive signal of early-onset cancer that could fit into a potential general purpose early-onset cancer risk prediction model in the future.

How should findings from this study influence the way that community oncology thinks about risk and prevention?

Allostatic load, as mentioned, is a manifestation of chronic stress in the body. It's largely calculated from very simple, routinely collected laboratory values like hemoglobin A1c, glucose, and from certain vital signs like body mass index, so it is not a difficult, “protected” variable to collect. That's in some sense the beauty of it, because it could be reproduced from all patients who are routinely seen in a community oncology setting.

I think the way this largely influences community oncology practice is probably more so in the need for additional validation of this variable. Now that we've shown this strong signal, we need to design more prospectively collected cohort studies to further prove out allostatic load, because we think that right now it's addressing a very unmet need of the lack of risk prediction models in early-onset cancer. So, rather than waiting for these cancers to pop up and treating them, oftentimes when they are in a stage where they can't necessarily be cured in these younger patients, we ought to be incorporating allostatic load into general risk prediction models as a way to better identify these patients. Maybe the way it influences community practice today is largely familiarizing yourself with what allostatic load is and potentially creating routinely collected markers of that in rpractice, so that we know which patients are at higher potential risk. But that being said, I think it's going to be up to guideline bodies to start incorporating this metric into their cancer risk prevention or risk assessment guidelines as we get further validation.

How much of the allostatic load signal do you think reflects social determinants of health that may be addressable at the community level?

In our study, we used a concept called the exposome framework to try to formalize a metric of exposures, including things like air pollution, chronic secondhand smoke, income disparities, and other elements that contribute to socioeconomic disparity. We tried to isolate those factors because allostatic load, we feel, is a biologically derived risk factor of chronic stress. That being said, we acknowledge the fact that a certain proportion of the socioeconomic determinants’ impact on early-onset cancer incidence may be explained through allostatic load, as opposed to other genomic or other risk factors.

In fact, we were able to quantify that many social determinants of health, their impact on early-onset cancer incidence could be explained if not partially, almost completely fully through the allostatic load variable. In that sense, we feel positive about it, because it's really difficult to operationalize social determinants of health in the standard cancer risk prediction models. Social determinants are insufficiently collected; they may be heterogeneously reported by patients, but allostatic load is an objective marker, and so if we regard it as an embedding of how the social determinants may impact cancer risk, then perhaps we're getting at the core of what really leads to this early-onset cancer, and in that sense may be able to avoid all the measurement errors of social determinants by formalizing it through this allostatic load concept.

REFERENCES

1. Parikh RB, Girard A, Moore TM, et al. Allostatic load as a mediator of exposomic risk pathways to early-onset cancer: A cross-cohort validation study. J Clin Oncol. 2026;44(16_suppl):10500-10500. doi:10.1200/jco.2026.44.16_suppl.10500

2. Zhao J, Xu L, Sun J, et al. Global trends in incidence, death, burden and risk factors of early-onset cancer from 1990 to 2019. BMJ Oncology. 2023;2(1). doi:10.1136/bmjonc-2023-000049