Michael Cook

Ph.D.

NCI, DCEG, HREB

Investigator

PLCO (Learn more about this study)

PLCO-92

Jun 23, 2014

An analysis of body mass index trajectories in relation to esophageal and gastric cardia adenocarcinoma in the PLCO and NIH-AARP cohorts

Background. Over the last two decades, esophageal and gastric cardia adenocarcinoma (EA/GCA) have been among the most rapidly increasing cancer types in the United States (U.S.) and other Western countries. EA and GCA have overlapping pathogeneses and are often analyzed as a single clinical entity; they are both glandular epithelial cancers originating in or near the gastroesophageal junction and have similar risk factor profiles and 5-year survival rates. Increasing rates of obesity in the general population have paralleled the increasing rates of EA/GCA. Elevated body mass index (BMI, kg/m2) has consistently been shown to be a risk factor for EA and a somewhat weaker risk factor for GCA. While the exact mechanisms are unclear, it is hypothesized that central adiposity may increase intra-abdominal pressure. This promotes gastroesophageal reflux disease and the transition to Barrett’s esophagus, the precursor lesion of EA. No studies to date have examined early adulthood adiposity exposures or trajectories of adiposity in relation to EA/GCA risk. One recent study examined combined obesity-related cancers (i.e., colorectal, renal, pancreatic, and esophageal) and found an increased risk of these cancer types with increased BMI at age 21 and a more modest association with adult weight gain. However, this study was underpowered to examine EA/GCA as distinct outcomes. The objective of this study is to analyze adiposity at discrete ages, weight change, and weight trajectories in relation to the distinct outcomes of EA and GCA.

Study Populations. As EA and GCA are rare cancer types, we propose to pool data from the PLCO and NIH-AARP study to increase sample size.

Exposure. Specifically, we will analyze discrete adiposity (i.e., BMI at ages 20, 35, 50, and current), weight change (≤2, -1.9 to <5, 5 to <15, ≥15 kg), and weight trajectories in relation to EA/GCA.

Outcomes. Tumours with ICD-10 codes of C15.0–C15.9 will be classified as esophageal cancers and C16.0 will be classified as gastric cardia cancers; only adenocarcinomas (ICD-O-3: 8140–8575) will be included in these analyses.

Statistical Analysis. For analyses of age-specific weights and—separately—weight change between two time points, Cox proportional hazards regression will be utilized. Simply adjusting a model of weight change for baseline weight can produce spurious results, as they are not independent variables. Thus, we propose to examine both weight change and weight change adjusted for average weight, which are orthogonal variables. Variables will be retained in the adjusted model if they are significant predictors of the outcomes (p<0.05). The proportional hazards assumption will be tested utilizing an interaction with log(time). To examine district group weight trajectory patterns in relation to EA/GCA, a latent class trajectory modeling approach will be used, which will enable us to define clusters of individuals with similar weight change patterns.

In sum, pooling data from PLCO and NIH-AARP will allow us to adequately examine the proposed exposures, as well as examine potential effect measure modification. The prospective nature of these cohorts also provides distinct advantages over case-control designs, most notably in that it provides some protection against reverse causation.

Specific Aim 1: Determine if adiposity at discrete ages is associated with EA/GCA.
Hypothesis: There will be a positive association between higher BMI and EA/GCA incidence, which may vary by age.
Rationale: The incidence of EA and GCA has increased 300–575% over the last 30–40 years, which has paralleled the Increasing rates of obesity in the general population. Elevated BMI has consistently been shown to be a risk factor for EA and a somewhat weaker risk factor for GCA.
No studies to date have examined early adulthood adiposity exposures in relation to EA/GCA risk. One recent study examined combined obesity-related cancers (i.e., colorectal, renal, pancreatic, and esophageal) and found an increased risk of these cancer types with increased BMI at age 21. However, this study was underpowered to examine EA/GCA as distinct outcomes.

Specific Aim 2: Determine if weight change is associated with EA/GCA.
Hypothesis: There will be a positive association between weight gain and EA/GCA incidence.
Rationale: While BMI has consistently been shown to be a risk factor for EA and GCA, no studies to date have examined adult weight change in association with these tumors. One recent study examined combined obesity-related cancers (i.e., colorectal, renal, pancreatic, and esophageal) and found a modest increased risk of these cancer types with higher levels of adult weight gain. However, this study was underpowered to examine EA/GCA as distinct outcomes.

Specific Aim 3: Determine if weight trajectories are associated with EA/GCA.
Hypothesis: There will be a positive association between weight gain trajectories and EA/GCA incidence.
Rationale: While BMI has consistently been shown to be a risk factor for EA and GCA, no studies to date have examined trajectories of weight change in association with these tumors. It is unclear if weight cycling, a pattern of losing and regaining weight, among overweight or obese individuals is associated with adverse outcomes.