Premium
Anthropometric factors and Breslow thickness: prospective data on 2570 cases of cutaneous melanoma in the population‐based Janus Cohort
Author(s) -
Stenehjem J.S.,
Veierød M.B.,
Nilsen L.T.,
Ghiasvand R.,
Johnsen B.,
Grimsrud T.K.,
Babigumira R.,
Støer N.C.,
Rees J.R.,
Robsahm T.E.
Publication year - 2018
Publication title -
british journal of dermatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.304
H-Index - 179
eISSN - 1365-2133
pISSN - 0007-0963
DOI - 10.1111/bjd.16825
Subject(s) - medicine , cohort , breslow thickness , prospective cohort study , population , melanoma , dermatology , anthropometry , cohort study , surgery , cancer , environmental health , sentinel lymph node , breast cancer , cancer research
Summary Background Breslow thickness is the most important prognostic factor of localized cutaneous melanoma ( CM ), but associations with anthropometric factors have been sparsely and incompletely investigated. Objectives To examine prediagnostic body mass index ( BMI ), body surface area ( BSA ), and height, weight and weight change in relation to Breslow thickness, overall and by anatomical site and histological subtype; and to assess possible nonlinear associations between these anthropometric factors and Breslow thickness. Methods CM s in the Janus Cohort were identified between 1972 and 2014. Linear regression was used to estimate geometric mean ratios ( GMR s) of Breslow thickness with 95% confidence intervals ( CI s) according to anthropometric factors. Restricted cubic splines in generalized linear models predicted adjusted mean Breslow thickness, and were used to assess possible nonlinear relationships. Results Of 2570 cases of CM , obese patients had a GMR of 1·16 (95% CI 1·04–1·30) of Breslow thickness vs. normal‐weight patients. For BSA and weight, quintile 5 showed GMR s of 1·13 (95% CI 1·00–1·27) and 1·17 (95% CI 1·03–1·33) of Breslow thickness vs. quintile 1, respectively. Associations seemed restricted to superficial spreading melanomas and CM s on the trunk and lower limbs. The associations plateaued at an adjusted mean Breslow thickness of about 2·5 mm ( BMI 29 kg m −2 , BSA 2·05 m 2 and weight 90 kg), before declining for the highest values. No associations were found for height and weight change. Conclusions This large case‐series of incident CM demonstrated positive associations between BMI , BSA , weight and Breslow thickness, and suggested that behavioural or other mechanisms apply at high values.