Background

Obesity classifications by body mass index (BMI) or metabolically abnormal obesity (MAO) fail to address the heterogeneity of obesity. We proposed a novel classification based on identifying actionable phenotypes (PMID: 25486131). Our aim was to examine specific characteristics and uniqueness of phenotypes within this novel classification of human obesity.

Methods

In a homogenous, phenotype-defined population of 283 patients with obesity (age 36(28-46) y, BMI 35(32-38) kg/m2, 75% females, non-MAO), we examined unique specific phenotypic characteristics within each subgroup of obesity.

Results

Obesity can be sub-grouped into the following categories: abnormal satiation (14%), abnormal satiety (16%), hedonic (13%), slow metabolism (11%), mixed group (34%) and other (14%). In abnormal satiation group, there is high caloric intake prior to reaching fullness measured by ad libitum buffet meal and hypothalamic blood flow measured by PASL fMRI remains low in the postprandial period, compared to lean controls and other groups of obesity. In abnormal satiety group, there is a suboptimal gut response to food intake with decreased VAS-fullness over 120 min post-meal, accelerated gastric emptying measured by scintigraphy, decreased postprandial levels of GLP-1 and PYY hormones. In hedonic group, there are increased levels of anxiety, depression, cravings and low serum tryptophan compared to other groups. The slow metabolism group had low resting energy expenditure measured by indirect calorimetry and decreased muscle mass measured by DEXA compared to other groups. As phenotyping is limited to few academic centers, we developed (n=180) and validated (n=120) a fasting blood multi-omic (genetic, hormones, metabolites) test that predicted the obesity subgroups (ROC >90% AUC).

Conclusions

Actionable phenotypes can be identified within a heterogeneous, complex cohort of obesity; we have developed and validated a novel, simple, fasting, blood-based multi-omic biomarker test to identify obesity phenotypes.