Background The effects of latitude gradients on diseases serve as surrogates for environmental factors to generate hypotheses for further testing. The age of onset in rheumatoid arthritis (ao-RA) is one important outcome predictor. Some reports suggest that ao-RA may be different depending on latitude gradients.

Objectives To assess the ao-RA at preselected cities worldwide.

Methods A proof-of-concept, worldwide survey. We selected major cities in each 15° quadrant (latitude and longitude) worldwide and found rheumatologists' contact information through rheumatology associations (e.g. ACR, PANLAR). An invitation e-mail was sent to rheumatologists requesting to directly interview 20 consecutive RA patients (ACR criteria) who attended their clinics about ao-RA. A second e-mail with instructions and a data-log sheet was sent to those who accepted. The information was limited to patient ID (initials or consecutive numbers), date of birth (month and year), date of RA diagnosis and date of RA onset (when the patient first noted a swollen joint). This data was considered nonsensitive, so no informed consent was required as per local regulations. Our goal was to have 3 participating rheumatologists in each selected city from 40 countries around the world. Statistical analysis: intra-city consistency was considered if no significant differences in ao-RA among rheumatologists' data were found (non-parametric tests). Then the ao-RA was compared per latitudes (each 10°) and longitudes by continents, and worldwide (one-way ANOVA with post-hoc Scheffé for multiple comparisons). Frequency distribution per ao-RA etary groups (10 year intervals) was also analyzed by country and quadrant. Significant differences were set at p<0.05.

Results Data from 2,481 patients were obtained from 126 rheumatologists in 78 cities in 42 countries. Intra-city variability could be assessed in 29 (37%) cities, and consistency was ascertained in 72.5% of these. The overall frequency distribution by age group showed that in 28% of patients the disease started before 35 yrs of age, 23% between 36 and 45, 26% between 46 and 55, 15% between 56 and 65, and 8% after 65 yrs of age; the mean ao-RA was 44±14 years (95% CI 44 to 45) and 84% were women. The linear chart of the mean ao-RA by worldwide latitudes showed a U shape, with the first peak in southern latitudes (>30°), the second peak in northern latitudes (>41°) and the base at the Tropic of Cancer. The ao-RA was 9 years older in northern latitudes when compared to the Tropic of Cancer (48 ± 15 vs. 39 ± 12 years, p<0.001), and the proportion of patients with ao-RA after the age of 55 was also higher (38% vs. 20%, p<0.001); these figures were consistent when analyzing data by continent.

Conclusions RA should be considered a disease mainly affecting young people, with almost one-third beginning before age 36 and half before the age of 45 worldwide. ao-RA appears to be influenced in some way by latitudes, mainly in the Tropic of Cancer (younger) and northern latitudes (older). Studies to assess environmental factors in these latitudes may be worth pursuing further.

Disclosure of Interest None declared