Prevalence of Multiple Chronic Conditions Among Adults in the All of Us Research Program: Exploratory Analysis 2025 Li et al

Abstract

Background:The growing prevalence of multiple chronic conditions (MCC) has significant impacts on health care systems and quality of life. Understanding the prevalence of MCC throughout adulthood offers valuable insights into the evolving burden of chronic diseases and provides strategies for more effective health care outcomes.

Objective:This study estimated the prevalence and combinations of MCC among adult participants enrolled in the All of Us (AoU) Research Program, especially studying the variations across age categories.

Methods:We conducted an exploratory analysis using electronic health record (EHR) data from adult participants (N=242,828) in the version 7 Controlled Tier AoU Research Program data release. Data analysis was conducted using Python in a Jupyter notebook environment within the AoU Researcher Workbench. Descriptive statistics included condition frequencies, the number of chronic conditions per participant, and prevalence according to age categories. The presence of a chronic condition was determined by documentation of one or more ICD-10 (International Statistical Classification of Diseases, Tenth Revision) codes for the respective condition. Age categories were established and aligned with diagnosis dates and stages of adulthood (early adulthood: 18-39 years; middle adulthood: 40-49 years; late middle adulthood: 50-64 years; late adulthood: 65-74 years; advanced old age: 75-89 years).

Results:Our findings demonstrated that approximately 76% (n=183,753) of AoU participants were diagnosed with MCC, with over 40% (n=98,885) having 6 or more conditions and prevalence increasing with age (from 33.78% in early adulthood to 68.04% in advanced old age). The most frequently occurring MCC combinations varied by age category. Participants aged 18-39 years primarily presented mental health–related MCC combinations (eg, anxiety and depressive disorders; n=845), whereas those aged 40-64 years frequently had combinations of physical conditions such as fibromyalgia, chronic pain, fatigue, and arthritis (204 in middle adulthood and 457 in late middle adulthood). In late adulthood and advanced old age, hyperlipidemia and hypertension were the most commonly occurring MCC combinations (n=200 and n=59, respectively).

Conclusions:We report notable prevalence of MCC throughout adulthood and variability in MCC combinations by age category in AoU participants. The significant prevalence of MCC underscores a considerable public health challenge, revealed by distinct condition combinations that shift across different life stages. Early adulthood is characterized predominantly by mental health conditions, transitioning to cardiometabolic and physical health conditions in middle, late, and advanced ages. These findings highlight the need for targeted, innovative care modalities and population health initiatives to address the burden of MCC throughout adulthood.

Open access

 

This seems a misleading attempt at creating a pathway where there isn't one, and is probably more a function of what, and who, gets diagnosed when by physicians, a shift from "you're too young to be disabled, must be psychological" ceding some ground after a certain age. Plus probably another function of how young people consult physicians much less often, where multiple conditions can remove the constant nagging "could it be psychological?" algorithm they all run, where attribution to a diagnosis can serve to put it below the threshold. Not all of it, of course, but the tendency to misdiagnose health problems as psychological is definitely as much a factor of sex as it is of age. Old people don't suffer from brain fog, they suffer from dementia, or old age. Or something like it. They do, of course, it's all about what gets attributed how by whom.

It tells us far too little about the health problems people experience, and much more about what the perception and decisions of the medical profession, the health care industry and health politics do. At least in part. How big of a part is impossible to know.

I basically think of it as hallucinations in LLMs. How do we know they're hallucinations? Because in most cases we know the real facts, and how they differ from what the LLM spits out. 6 fingers on a hand? Unlikely. But switching between 5 and 6 is impossible. But in the case of health issues there is no 'ground truth', no one knows what all these things are, so hallucinations can't be differentiated from facts. Until they eventually are, usually much, much later than they would have been had the early hallucinations not been taken as facts beyond correction. But there is never any back-propagation correcting the mistakes, the process of making the system learn from mistakes. Instead mistakes are simply swept under the rug, covered up, guaranteeing more of them in the future. As a choice. What a terribly self-defeating choice.

 

How much of that is just officially categorizing normal aspects of life? Some of the "conditions" might only exist when an official questionnaire is created.

 

That was my thought. The paper seems to say 'old people get old'.