Clinical and Experimental Rheumatology 2000; 18: 375-378. BRIEF PAPER Palindromic ABSTRACT eventually develop RA (1-3). Another Objective etiology, namely “allergic rheumatism” rheumatism: Effect of Evaluation of the contribution of dietary was suggested by Hench and Rosenberg components in triggering the attacks of in their initial description of the syn- dietary manipulation palindromic rheumatism (PR), and the drome (1). However, they could not effect of dietary manipulation on the fre- prove such an allergic origin in their G. Nesher, M. Mates quency and severity of PR attacks. cases. Methods Following a clinical observation of a Rheumatology Service, Department of Sixteen patients (10 males, 6 females) patient who experienced a complete re- Intermal Medicine, Shaare-Zedek Medical were diagnosed as having PR during mission of his PR attacks while on a low- Center, Jerusalem, Israel. 1994-8 in one center. Their mean age calorie formula preparation, it was hy- Gideon Nesher, MD, Senior Lecturer, was 45 ± 6, duration of symptoms prior pothesized that antigens in food may be Hebrew University Medical School, and to diagnosis was 4 ± 1.4 years, and fre- the triggering factors of PR attacks. Adjunct Associate Professor, St. Louis quency of PR attacks were 3.1 ± 1.8/ Thus, a prospective study was initiated University School of Medicine; Michal month. All patients were instructed to to evaluate the possible relationship be- Mates, M.D. make a list of the food that was consumed tween food and PR episodes. Please address correspondence and daily and to specify the dates of PR epi- reprint requests to: Dr. Gideon Nesher, sodes. Data were evaluated after a pe- Patients and methods Rheumatology Service, Department of riod of 2-4 months in each patient. During a period of 55 months, 16 patients Intermal Medicine, Shaare-Zedek Medical Results were diagnosed as having PR according Center, P.O.Box 3235, Jerusalem 91031, Israel. E-mail: [email protected] In 5 patients (31%) there was an asso- to the diagnostic criteria of Mahonen et ciation between episodes of PR and cer- al. (4). These criteria include: recurrent © Copyright CLINICAL AND tain foods that were consumed within 36 attacks of sudden-onset mono- or pol- EXPERIMENTAL RHEUMATOLOGY 2000. hours prior to PR episodes. These were yarthritis or of para-articular soft tissue fish (2 patients), eggs, canned vegeta- inflammation lasting from a few hours Key words: Palindromic rheumatism, bles and processed cheese (each in one up to one week, verification of at least diet, food. case). Elimination of the relevant food one attack by a physician, subsequent from each patient’s diet resulted in com- attacks at least in three different joints, plete cessation of the PR attacks in two and exclusion of other forms of arthriti- of the cases, while the other three had des. milder, infrequent attacks. Four patients All patients were instructed to make a were rechallenged with the offending daily list of the food that was consumed, food. In all cases it resulted in recurrence and to specify the time and date of the of the PR attacks. No association be- PR attacks, their duration (in hours) and tween PR episodes and prior consump- their severity (using a 1-5 scale). Data tion of certain foods could be docu- were evaluated for each patient after a mented in the other 11 patients. period of 7-14 weeks. In each case the Conclusion food that was reported to be consumed In some PR patients ingestion of certain during the 48 hours prior to each attack foods, specific for each case, can trig- of PR was marked. A search was then ger the typical attack. It is suggested that made to find which of these food prod- this association should be looked for in ucts were consumed prior to at least 80% any PR patient, as elimination of the of- of the PR episodes in each patient. These fending food from the diet may help in food products were marked as possibly preventing the PR attacks. associated with the PR attacks. Another search was then made, to exclude food Introduction products that were consumed repeatedly Palindromic rheumatism (PR), is char- during the attack-free intervals. Patients acterized by recurring afebrile episodes were then treated and followed for a pe- of acute arthritis and periarthritis in one riod of 12 - 42 months. or several joints, lasting from a few hours To evaluate the specificity of the food to one week, leaving no residual dam- diary data, a control group of 7 patients age (1). The etiology of PR is obscure. with familial Mediterranean fever It has been debated whether PR is part (FMF), another “periodic disease”, was of the spectrum of rheumatoid arthritis included. These patients were also in- (RA), representing an early stage in the structed to make a daily list of the food evolution of RA, as many cases of PR that was consumed, and to specify the 375 BRIEF PAPER Palindromic rheumatism and food / G. Nesher & M. Mates Table I. Effect of dietary manipulation on 5 patients with palindromic rheumatism. Age, gender 47, M 45, F 48, M 42, F 44, M Trigger Fish Processed cheese Canned vegetables Whole egg Canned fish Food-attack interval 12 - 18 hrs 18 - 36 hrs 12 - 24 hrs 18 - 24 hrs 18 - 36 hrs Attack severity*, 2.7 ± 0.8 3.4 ± 0.9 3.8 ± +0.8 4.5 ± 0.5 3.8 ± 0.8 frequency 3/month 3/month 2/month 2/month 1.5/month Attack frequency after elimination 0 0.1/month 0.2/month 0.3/month 0 Rechallenge yes yes no yes yes Recurrence with rechallenge yes yes yes yes *Severity of attacks was self-recorded by the patients, using a scale of 1-5. time of the FMF attacks. Their food di- riod. There were no skin or respiratory yellow cheese in the second case, canned ary data were evaluated as described for allergic reactions during the PR attacks. tuna fish in the third, and salmon and the PR patients. ESR and CRP were elevated during the carp (on two different occasions) in the attacks in only one case. IgE levels and fourth patient. In all cases a typical PR Results eosinophil counts were evaluated during attack was observed by the treating phy- There were 10 men and 6 women, their some of the attacks and were within the sician within 24 hours of ingestion. In mean age was 45 ± 6. The mean dura- normal range. No association was found the third patient it was associated with tion of PR symptoms prior to diagnosis between specific food products and FMF elevation of the ESR and CRP to 58 mm/ was 4 ± 1.4 years. The frequency of PR attacks in the control group. h and 3.4 mg/dl, respectively (normal < attacks was 3.1 ± 1.8 per month, and the Each of these five PR patients was in- 30, and < 1, respectively). The first pa- mean duration of each attack was 30 ± structed to eliminate the specific food tient was also rechallenged with sepa- 16 hours. An elevated sedimentation rate from the diet. Patients were then fol- rated egg-yolk and egg-white: the egg- (ESR) or C-reactive protein (CRP) dur- lowed for a mean of 2.1 ± 1.1 years. white triggered a typical attack while ing an attack was documented in 4/15 During that period 2 of the 5 patients ingestion of egg-yolk resulted in very patients (27%). Two patients (12%) had were completely free of PR episodes, mild wrist arthralgia that resolved within a positive rheumatoid factor (RF). Pa- while the other 3 had milder, infrequent two hours. The second patient experi- tients had neither symptoms of FMF nor attacks (Table I). Taken together, prior enced PR attacks with 3 different brands a family history of this disease. None of to the dietary manipulation these 5 pa- of yellow cheese. There were no attacks the patients had a history of allergic dis- tients had a mean of 2.2 ± 0.8 attacks following ingestion of white cheese or eases, gastrointestinal diseases, elevated per month which decreased to 0.1 ± 0.1. other dairy products. The third patient IgE levels or IgA deficiency. None of the The mean duration of the PR attacks had no attacks following ingestion of patients had a family history of RA. All decreased from 31.2 ± 6.6 hours to 5 ± non-canned fish, or canned food other patients had normal levels of uric acid. 6.5 hours, and the mean self-reported than fish (such as corn, peas, mushrooms The joint areas most commonly involved severity score decreased from 3.6 ± 0.7 and olives). In addition to a mild PR at- were the metacarpo-phalangeal and to 1 ± 1.4. tack following the ingestion of carp, the proximal interphalangeal, each in 12 Four of the patients agreed to have a fourth patient had also a very mild at- patients, followed by the wrists in 10 rechallenge with the specific offending tack following the eating of salmon, but patients and the knees in 8. The mean food: a whole boiled egg in one case, the ingestion of canned tuna fish did not number of joint areas involved in each patient was 3.5 ± 1. In 5 patients (31%), an association was Table II. Comparison between palindromic rheumatism patients with food-triggered attacks found between PR episodes and con- (FT) and non food-triggered attacks (non FT). sumption of certain foods within 36 hours prior to the attack (table 1). These FT (n = 5) Non FT (n = 11) were fish, canned fish (tuna and sar- Male: female ratio 3: 2 7: 4 dines), canned vegetables (corn and green peas), eggs, and processed (yel- Mean age at the start of symptoms 41 ± 2 41 ± 6 low) cheese, each in one patient.