# Covariate Relationships

**URL:** <https://discourse.pumas.ai/t/covariate-relationships/1428>\
**Category:** Uncategorized\
**Created:** [July 21, 2026, 6:25am UTC](https://discourse.pumas.ai/t/covariate-relationships/1428 "2026-07-21T06:25:47Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Abhiram](https://avatars.discourse-cdn.com/v4/letter/a/43a26b/32.png) [@Abhiram](https://discourse.pumas.ai/u/Abhiram)\
**Post date:** [July 21, 2026, 6:25am UTC](https://discourse.pumas.ai/t/covariate-relationships/1428/1 "2026-07-21T06:25:47Z")

</div>

We build a covariate model to find the effect of any covariates on a pk parameter. Many times it has been observed that covariates has statistically significant effect while modelling, but in real life the covariate effects were modest (≤ 20%) and not clinically meaningful. Why so?  
Can anyone explain this?