Author:
|
Lønnebotn, Marianne; Svanes, Cecilie; Igland, Jannicke; Franklin, Karl A.; Accordini, Simone; Benediktsdottir, Bryndís; Bentouhami, Hayat; Blanco, José A. G.; Bono, Roberto; Corsico, Angelo Guido; Demoly, Pascal; Dharmage, Shyamali C.; Dorado Arenas, Sandra; García Aymerich, Judith; Heinrich, Joachim; Holm, Mathias; Janson, Christer; Jarvis, Deborah; Leynaert, Bénédicte; Martínez Moratalla Rovira, Jesús; Nowak, Dennis; Pin, Isabelle; Raherison, Chantal; Sánchez Ramos, José Luis; Schlunssen, Vivi; Skulstad, Svein Magne; Dratva, Julia; Gómez Real, Francisco
|
Abstract:
|
Life course data on obesity may enrich the quality of
epidemiologic studies analysing health consequences of obesity.
However, achieving such data may require substantial resources.
We investigated the use of body silhouettes in adults as a tool
to reflect obesity in the past. We used large population-based
samples to analyse to what extent self-reported body silhouettes
correlated with the previously measured (9-23 years) body mass
index (BMI) from both measured (European Community Respiratory
Health Survey, N = 3 041) and self-reported (Respiratory Health
In Northern Europe study, N = 3 410) height and weight. We
calculated Spearman correlation between BMI and body silhouettes
and ROC-curve analyses for identifying obesity (BMI >/=30) at
ages 30 and 45 years. Spearman correlations between measured BMI
age 30 (+/-2y) or 45 (+/-2y) and body silhouettes in women and
men were between 0.62-0.66 and correlations for self-reported
BMI were between 0.58-0.70. The area under the curve for
identification of obesity at age 30 using body silhouettes vs
previously measured BMI at age 30 (+/-2y) was 0.92 (95% CI 0.87,
0.97) and 0.85 (95% CI 0.75, 0.95) in women and men,
respectively; for previously self-reported BMI, 0.92 (95% CI
0.88, 0.95) and 0.90 (95% CI 0.85, 0.96). Our study suggests
that body silhouettes are a useful epidemiological tool,
enabling retrospective differentiation of obesity and
non-obesity in adult women and men. |