All traits

Inflammatory bowel disease

Polygenic score · Health · 137 variants

Inflammatory bowel disease (Crohn's and ulcerative colitis) involves chronic gut inflammation. This score sums common variants affecting susceptibility.

What this inflammatory bowel disease score reads

This is PGS004105, one of the scores from Monti and colleagues' 2024 benchmarking study that ran the same polygenic methods across five separate biobanks. It uses 139 variants selected by clumping and thresholding from a GWAS of 59,957 people.

Inflammatory bowel disease covers both Crohn's disease and ulcerative colitis, and this score treats them as one trait even though they differ in genetics, location and behaviour.

How well it predicts, and why the answer depends on the biobank

The benchmarking design gives an unusually honest picture. Across the five biobanks the same score produced odds ratios per standard deviation ranging from 1.24 to 1.63, areas under the curve from 0.56 to 0.63, and variance explained from 0.7% to 3.7%.

That spread is the paper's central finding: performance varied more between biobanks than between the statistical methods used to build scores. A percentile from a score like this is therefore not a fixed property of you, it depends on which population you are being compared against and how cases were ascertained there.

How to read a high or low score

Inflammatory bowel disease is uncommon, affecting a few hundred people per 100,000 in Western countries, so even a doubling of relative risk leaves absolute risk low. A high percentile is not a reason for testing in the absence of symptoms.

Persistent diarrhoea, blood in the stool, unexplained weight loss, night-time symptoms or ongoing abdominal pain are what warrant medical attention, and they warrant it regardless of any score.

What this score does not capture

The NOD2 coding variants, which are the strongest common genetic risk factors for Crohn's disease, are not in this score. This site reads them separately on the Crohn's disease page, which is worth looking at alongside this one.

Nor does it capture smoking, which is one of the strangest and best-documented environmental factors here: it raises Crohn's disease risk while being associated with lower ulcerative colitis risk. Antibiotic exposure, appendicectomy, diet and the gut microbiome all sit outside the score too.

Ancestry and accuracy

The score was developed in European-ancestry data and evaluated mainly there, with one South Asian biobank included; in that biobank the odds ratio per standard deviation was 1.35 to 1.40, within the range seen in European cohorts but from a much smaller case count. Ashkenazi Jewish ancestry carries notably higher inflammatory bowel disease risk, which a score of this kind does not represent.

Variants in this score you can read on their own

This score includes a variant that this site also explains individually, where you can see what each genotype means on its own: Autoimmune protection (IL23R R381Q) (rs11209026).

How much of this score your file covers

137 of the published score's 139 variants are present in a standard consumer DNA file (99%). The remaining 2 cannot be read from one, so a result computed here approximates the published score rather than reproducing it.

Common questions

Does this score tell me whether I would get Crohn's disease or ulcerative colitis? No, it treats them as a single trait. The two conditions share much of their genetic architecture but differ in important places, including NOD2, which contributes to Crohn's disease and is not in this score.

Why do different tools give me different inflammatory bowel disease percentiles? Partly because they use different variant sets, and partly because a percentile depends on the reference population. This score's own benchmarking paper found performance varied more between biobanks than between methods, which is a good reason to hold any single percentile loosely.

Related

Alzheimer's disease · Body mass (BMI) tendency · Breast cancer (female) · Colorectal cancer risk (8q24) · Fasting glucose / type 2 diabetes (MTNR1B) · Systolic blood pressure · Uric acid / gout tendency

References: PGS Catalog · PubMed

Educational and informational only, not medical advice.

How we interpret results · How to read your raw DNA data

Already have your 23andMe, AncestryDNA, or FamilyTreeDNA file?

See your own result.