Background/Purpose: Patients with rheumatoid arthritis (RA) often attribute an exacerbation of their symptoms to changes in the weather. There are several studies examining this relationship. We conducted a systematic review of the existing literature on the effect of weather on symptoms in RA, and a follow up study that characterized patients with RA that reported being “weather sensitive”.

Methods: A systematic review of Medline and Embase from inception until November 2011 looked for articles that report the effect of any aspect of weather on any symptom or outcome of patients with arthritis and 765 articles qualified. Then only included were non-duplicative articles in English with RA patients. We identified a subset of patients as “weather sensitive”(WS) based on patients reporting worse symptoms with increased weather. We observed that the percentage of WS patients in the studies reviewed, along with the difference in study methods, played a role in determining the final results. In order to identify WS patients and determine if they differ from other patients, we queried a cohort of 6901 US rheumatic disease patients (including 4655 RA, 1206 osteoarthritis (OA), and 374 fibromyalgia (FMS)) participating in an observational study and compared their demographic and clinical characteristics.

Results: Only 19 articles (N=3357 RA patients) qualified for our review. These studies did not find a consistent association between weather and RA symptoms. For example, 4 studies found a positive correlation between temperature and symptoms like pain and stiffness, 4 others found a negative correlation, and 4 more found no correlation. We had similar results for 11 articles studying humidity and RA outcomes: 6 positive, 1 negative, and 4 had no correlation. For atmospheric pressure, 3 were positive, 1 was negative, and 6 had no correlation. Of the 4655 RA patients we surveyed, 63% reported being WS (compared to 71% of OA and 78% of FMS), 20% were not, and the remaining 17% did not know. Columns 2-3 of Table 1 list their characteristics. Patients in the WS group were younger, less male, less educated, and had shorter RA duration. They also had worse RA symptoms as measured by HAQ, pain VAS, and SF36. The logistic multivariable regression for WS odds ratios are provided in column 4. We accounted for possible over reporting of symptoms by including a patient symptom count and polysymptomatic distress scale.

Conclusion:  Although widely believed, no consistent association in the literature between weather and RA symptoms was seen.  Patients with worse function and pain are more likely to report being WS even after controlling for sociodemographic status and general symptoms. Future studies aimed at objectively measuring WS are needed to further evaluate whether subjective symptoms are more correlated with weather possibly through increased pain sensitization.

Table. Characteristics of RA patients by weather sensitivity and multivariable logistic regression.

	Not weather sensitive (N=926) (SD)	Weather sensitive (N=2926)	Odds ratio (95%CI)
Age (years)	64.8 (12.0)	62.1 (12.3)	0.98 (0.97, 0.99)
Male (%)	25.6	15.0	0.71 (0.58, 0.88)
Education (0-17 years)	14.7 (2.2)	14.1 (2.3)	0.91 (0.88, 0.95)
RA duration (years)	19.7 (12.1)	18.6 (SD 12.3)	0.99 (0.99, 1.00)
HAQ (0-3)	0.61 (0.7)	1.10 (SD 0.7)	1.26 (1.02, 1.55)
Pain VAS (0-10)	2.16 (2.4)	3.97 (SD 2.7)	1.09 (1.04, 1.15)
SF36 Physical CS Score	44.3 (11.4)	36.0 (SD 10.8)	0.97 (0.96, 0.99)
Symptom count (0-37)	4.9 (4.6)	8.4 (6.0)	1.05 (1.03, 1.08)
Polysymptomatic distress scale (0-31)	6.2 (6.5)	11.5 (7.9)	1.02 (1.00, 1.04)

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/the-effect-of-weather-on-patient-symptoms-in-rheumatoid-arthritis-a-systematic-review-of-the-literature-and-exploration-of-weather-sensitivity/