AIMS & HYPOTHESES

Aim 1: To use latent class analysis to empirically derive obesity subtypes

Hypothesis 1: Distinct subtypes of obesity exist that can be characterized by a combination of behavioral, physiologic, environmental, genetic and other familial factors.

Aim 2: To examine whether the following groups are distinct obesity phenotypes: onset of obesity before 5 years of age; obesity onset before age 18 and at least one obese parent; obesity with loss of control eating; obesity with emotional eating

Hypothesis 1: There will be overlap between the empirically derived subtypes in aim 1 and the proposed subgroups in aim 2.

Hypothesis 2: Risk factors (birthweight, maternal weight status, high sugar-sweetened soda or fast food intake, low physical activity, depressive symptoms, internalizing and externalizing symptoms, experiencing trauma, early puberty, screen time) for and cognitive (low inhibition, high impulsivity) and biological correlates (high systolic blood pressure, low leptin, low HDL levels, poor sense of smell) of obesity may vary by subtype.

Aim 3: To examine the patterns of association between the obesity subtypes derived in aim 1 and the obesity SNPs identified or confirmed in previous large genetic epidemiology consortia7,9

Hypothesis 3: FTO and MC4R will be more strongly associated with subtypes focused on unhealthy eating behaviors (i.e., loss of control eating and emotional eating).

As secondary aims we will explore whether the empirically derived classifications vary by developmental stage (childhood vs. adolescence) or gender. In other secondary aims we will investigate if fMRI results among girls show differences in selective attention and working memory across the obesity phenotypes identified.