Article Text

PDF
OP0275 Birth Characteristics and Childhood Infections Predict Ankylosing Spondylitis. A National Register Based Nested Case-Control Study
  1. U. Lindström1,
  2. H. Forsblad-d'Elia1,
  3. J. Askling2,
  4. L.E. Kristensen3,4,
  5. E. Lie1,5,
  6. S. Exarchou3,
  7. L. Jacobsson1
  1. 1Institute of Medicine, Sahlgrenska Academy, Rheumatology and Inflammation Research, Gothenburg
  2. 2Rheumatology Unit, Department of Medicine Solna, Karolinska Institutet, Stockholm
  3. 3Section of Rheumatology, Department of Clinical Sciences, Lund University, Lund, Sweden
  4. 4Department of Rheumatology, The Parker Institute, Frederiksberg and Bispebjerg Hospital, Denmark
  5. 5Diakonhjemmet Hospital, Oslo, Norway

Abstract

Background Ankylosing spondylitis (AS) is strongly associated with genetic factors, in particular the HLA-B27 genotype. In contrast, studies on environmental risk factors for developing AS are few. For the related disease entity reactive arthritis the triggering effect of infections is well established. For other autoimmune diseases, such as diabetes and rheumatoid arthritis, epidemiological studies have demonstrated that birth weight may predict disease development later in life. For allergic diseases, including asthma, having older siblings seems to decrease the risk, possibly explained by the “hygiene hypothesis”.

Objectives To determine if birth characteristics and childhood hospitalization due to infections predict a diagnosis of AS.

Methods Data from several national registers were used for this study. The National Patient Register was used to identify AS cases, defined as individuals born in Sweden after 1972 with ≥1 visit to specialized health care with an ICD-code for AS, and first AS ICD-code occurring after the age of 16 years. For each AS case 5 matched (sex, age, county) controls were retrieved from the population register. Cases and controls were linked to the Medical Birth Register to retrieve data on birth characteristics, and the National Patient Register to identify hospitalization due to infections (based on ICD-codes) up to 16 years of age. Univariate and multivariate conditional logistic regression models were used to assess increased risk (odds ratios) for a later diagnosis of AS. Exposures assessed were birth weight, gestational age, type of birth (single/multiple), number of older siblings and exposure to infections. The following covariates were also included in the multivariate analysis: mothers' marital status, mothers' birth country and size of the delivery unit.

Results By univariate analyses statistically significant increased risks were observed for low birth weight (<3000g) (18% vs 15%, OR: 1.3 (95% Cl: 1.1–1.6)), having older siblings (63% vs 58%, OR: 1.3 (95% Cl: 1.1–1.4), and for having been hospitalized due to infections at age 5-12 (5% vs 3%, OR: 1.4 (95% Cl: 1.0–1.8) and at age 13-16 (2% vs 1%, OR: 1.6 (95% Cl: 1.0–1.5)). Similar point estimates were found in a multivariate analysis including birth weight, gestational age, number of older siblings, infections, mothers' marital status, mothers' birth country and size of the delivery unit (fig 1). Point estimates were similar for men and women as well as in sensitivity analyses excluding subjects with inflammatory bowel disease or psoriasis prior to first AS diagnosis.

Conclusions A diagnosis of AS was predicted by low birth weight, having older siblings and having been hospitalized due to infection during age 5-16 years, suggesting that these factors may be of importance in the disease pathogenesis.

Disclosure of Interest None declared

Statistics from Altmetric.com

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.