[DSRP Evidence](https://dsrpevidence.org/)

# The Impact of Definitions of Disease on Overdiagnosis

## Details

**Authors** Tikkinen, Kari A. O. and Halme, Alex L. E. and Guyatt, Gordon H. and Glasziou, Paul

**Year** 2025

**Publisher** JAMA Internal Medicine

**Kind of work** article

**Discipline** Medicine

**Secondary disciplines** Public Health

**Applied** false

[Read it at the publisher](https://doi.org/10.1001/jamainternmed.2025.1727) 
10.1001/jamainternmed.2025.1727

## In authors' words

### What they set out to do (purpose)

To show how the way a disease is defined shapes who is labeled as diseased and how much overdiagnosis follows.

### Who or what was studied (sample)

US adults aged 20 and older in NHANES 2017-2020 (for threshold illustrations); a population survey of 60 conditions cited from prior work.

### How they did it (methods)

Viewpoint article with illustrative prevalence calculations at different diagnostic cutoffs.

### What they found (results)

Lowering diagnostic cutoffs sharply raises the share labeled as diseased: total cholesterol 10% at 240 mg/dL versus 54% at 180 mg/dL; fasting glucose 12% at 126 mg/dL versus 61% at 100 mg/dL; systolic pressure 14% at 140 mm Hg versus 75% at 110 mm Hg. In the cited survey only 12 of 60 conditions were judged diseases by at least 80% of respondents, and physicians were more likely than laypeople to call a condition a disease. Wider gestational diabetes criteria roughly doubled prevalence without measurable improvement in outcomes.

## Commentary

### In short

Moving the line that defines a disease moves millions of people across it, and patients, doctors and the public do not agree on where the line belongs.

### Response

The threshold figures show that where a boundary is drawn changes what counts as a case (D), and the survey results show that patients, physicians and laypeople place it differently (P). Both are carried by the authors' own data and argument. The piece is a short viewpoint, so the evidence is illustrative rather than a controlled test, and the outcome claim rests on a single example (gestational diabetes).

**Patterns it shows** D, P

Shows it makes a difference

**Added** 2026-11-08

**How to cite this** Tikkinen, Kari A. O. and Halme, Alex L. E. and Guyatt, Gordon H. and Glasziou, Paul (2025). The Impact of Definitions of Disease on Overdiagnosis. JAMA Internal Medicine.
