POS1418 SEPARATING THE EFFECTS OF CHILDHOOD AND ADULT BODY SIZE ON INFLAMMATORY ARTHRITIS: A MENDELIAN RANDOMISATION STUDY

Abstract

BackgroundObesity is strongly associated with inflammation and inflammatory arthritis, but most studies to date have only examined adulthood body mass index (BMI). Whether childhood obesity has a direct and long-term influence on risk of inflammatory arthritis is unclear. This question is important as childhood obesity becomes a growing and global public health crisis [1], but traditional epidemiologic approaches may be limited by reverse causation (obesity can be cause and consequence of disease) and confounding. Mendelian randomisation (MR) uses genetic variants as instrumental variables to estimate the effect of an exposure on an outcome. Since variants are randomly allocated at conception, MR is less susceptible to reverse causation and confounding.ObjectivesTo examine whether childhood body size affects risk of inflammatory arthritides – rheumatoid arthritis (RA), ankylosing spondylitis (AS), psoriatic arthritis (PsA), gout and systemic lupus erythematosus (SLE) – after accounting for the effect of adult body size using two-sample univariable and multivariable MR.MethodsGenetic instruments for childhood (age 10 years) and adult body size were derived using data from 453,169 individuals from the UK biobank study (313 and 580 variants respectively; r2<0.001, p<5x10-8), which have been previously validated using BMI data from three independent populations [2]. Participants recalled their body size at age 10 as “thinner”, “about average” or “plumper”. Adult body size was determined using BMI at a mean age of 57 years and, to facilitate comparison, split into three groups to match the proportions for childhood body size. Genome-wide association data comprised of 22,350 RA, 9,069 AS, 3,609 PsA, 13,179 gout and 5,201 SLE cases. We conducted univariable MR to estimate the total effects of childhood and adult body size on outcomes and multivariable MR to examine the independent effect of childhood body size after accounting for adult body size. F statistics were calculated for each exposure with a value >10 considered suggestive of adequate instrument strength.ResultsThe F statistic for childhood and adult body size instruments were 30 and 23 respectively. Genetically predicted childhood body size had a total effect on risk of PsA (OR 2.18 per change in body size category; 95%CI 1.43, 3.31), gout (OR 2.18; 1.43, 3.31) and SLE (OR 2.44; 1.14, 5.22), but not RA (OR 0.95; 0.70, 1.29) or AS (OR 0.96; 0.61, 1.52). After accounting for adult body size, the direct effect of childhood body size was little changed for PsA (OR 1.92; 1.14, 3.25) and SLE (OR 2.69; 1.24, 5.87), but was attenuated for gout (OR 1.40; 0.94, 2.09).Figure 1.Mendelian randomisation estimates of the effect of childhood and adult body size on inflammatory arthritis.ConclusionResults of this study suggest that the risk conferred from having a larger body size during childhood may not be fully reversable even when a healthy size is achieved in adulthood. How the immune system is altered by early life adiposity to increase risk of PsA and SLE is unclear. Differential effect estimates of body size on each inflammatory arthritis (PsA vs AS, RA vs SLE) may help shed light on their unique pathophysiology in future studies.