Thread by @ukuv

Tweetorial on addictions and obesity, based on our recent paper “Obesity has limited behavioural overlap with addiction and psychiatric phenotypes” @NatureHumBehav

Paper: doi.org/10.1038/s41562…
Preprint: psyarxiv.com/jt2b7/
1/12

Are we loosing control over eating to the extent that we are addicted to food? Food addiction concept has been popularised since 1950-s and is debated to this day.@PaulPcf22

History of food addiction: ncbi.nlm.nih.gov/pmc/articles/P…
Debate: nature.com/articles/s4138…
2/12

There is even a questionnaire to measure food addiction, which I find very similar to trait Uncontrolled Eating (UE). UE = difficulties in regulation of eating. See tweetorial on UE here:

https://twitter.com/ukuv/status/1101500474459635713

3/12

We have reviewed overlap between addiction and obesity. They both correlated with most impulsivity domains. Similarly, they correlated with brain activity in impulse-control related areas. But this was a narrative review and N of domains was limited..
frontiersin.org/articles/10.33…
4/12

To quantify behavioural overlap between addictions and obesity we used personality profiles. Profiles came from prior papers correlating obesity/addictions with NEO PI-R personality facets. Behavioural overlap = personality correlation = correlation between squiggly lines.
5/12

When analysing only 5 personality domains, obesity (OB) and UE are rather similar to addictions (ADD). Obesity has as high similarity with addictions as addictions have with each other. Education (EDU) and gender (GEN) are baseline phenotypes.
This replicates our review.
6/12

However, behavioural overlap across 30 facets is more nuanced. We found a gradient: obesity has some similarity with addictions. UE’s similarity with addictions is considerably higher. Therefore the addiction model has some use for obesity, but it is not the whole picture.
7/12

We also zoomed in to the facets driving the association. We saw that addictions and obesity have similar correlations with N5: Impulsiveness and facets from Conscientiousness. However, addictions associate with E5: Sensation-Seeking, whereas obesity does not.
8/12

When exploring behavioural overlap between obesity and psychiatric conditions (PSYCH), we found similarities with mood disorders, obsessive-compulsive and cluster B personality disorders 😮. Again, facets (right) provided more information than domains (left).
9/12

The personality disorder-obesity link is explained by similar correlations with E3: Assertiveness. Not sure why this facet relates to obesity, but this association is robust! We also found this facet in a follow-up meta analysis. Any ideas welcome!
doi.org/10.1111/obr.12…
10/12

In summary, correlating personality profiles enables quantifying behavioural similarity between phenotypes. Using detailed profiles is better! Obesity has some overlap with addictions, but this is not the whole story.

Links:
doi.org/10.1038/s41562…
psyarxiv.com/jt2b7/
11/12

I wrote a “behind the paper” blog post, outlining how we got the idea for the analysis. Big thanks to @DecisionPark and @neuroecb for writing a comment to the paper, and to @marike_cogneuro for patient editing.
Blog: socialsciences.nature.com/users/325122-u…
Comment: nature.com/articles/s4156…
/fin