Skip Navigation


Rheumatology Advance Access originally published online on September 26, 2006
Rheumatology 2006 45(11):1370-1375; doi:10.1093/rheumatology/kel328
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
45/11/1370    most recent
kel328v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (6)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Wolfe, F.
Right arrow Articles by Chakravarty, E. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wolfe, F.
Right arrow Articles by Chakravarty, E. F.
Related Collections
Right arrow Rheumatoid Arthritis
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Rates and predictors of herpes zoster in patients with rheumatoid arthritis and non-inflammatory musculoskeletal disorders

F. Wolfe1,2, K. Michaud1,3 and E. F. Chakravarty4

1National Data Bank for Rheumatic Diseases and 2University of Kansas School of Medicine, Wichita, KS, 3Centre for Primary Care and Outcomes Research, Stanford University, Stanford and 4Division of Immunology and Rheumatology, Stanford University School of Medicine, Palo Alto, CA, USA.

Correspondence to: F. Wolfe, MD, National Data Bank for Rheumatic Diseases, Arthritis Research Center Foundation, 1035 N. Emporia, Suite 230 Wichita, KS 67214, USA. E-mail: fwolfe{at}arthritis-research.org


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 References
 
Objectives. Herpes zoster (HZ) is a common disorder that causes substantial pain and morbidity. We examined its rate and predictors in rheumatoid arthritis (RA) and non-inflammatory musculoskeletal (MSK) disorders to determine if HZ was increased in RA and whether treatment contributed to the risk of HZ.

Methods. After excluding patients witzh prior HZ, we assessed 10 614 RA and 1721 MSK patients by semi-annual questionnaires during 33 825 patient-years of follow-up. Predictors of HZ were determined by Cox regression and expressed as hazard ratios (HR) and 95% confidence intervals (CI).

Results. The annualized incidence rate per 1000 patient-years was 13.2 (95% CI 11.9–14.5) in RA and 14.6 (95% CI 11.2–18.1) in MSK, and did not differ significantly after adjustment for age and sex. HZ was predicted by impaired functional status, as measured by the Health Assessment Questionnaire (HAQ), [HR 1.3 (95% CI 1.1–1.5)] and by the use of COX-2-specific non-steroidal anti-inflammatory drugs (NSAIDs) [HR 1.3 (95% CI 1.1–1.6)] in RA and MSK. In multivariable analyses in patients with RA, cyclophosphamide HR 4.2 (95% CI 1.6–11.5), azathioprine HR 2.0 (1.2–3.3), prednisone HR 1.5 (1.2–1.8), leflunomide HR 1.4 (1.1–1.8) and COX-2 NSAIDs HR 1.3 (95% CI 1.1–1.6) were significant predictors of HZ.

Conclusion. The incidence of HZ is increased in RA and MSK compared with population-based rates. However, the rate of HZ in RA is not increased compared with MSK. After adjustment for severity, various treatments, but not methotrexate or biologics, were risk factors for HZ.

KEY WORDS: Herpes zoster, Incidence, Rheumatoid arthritis, Treatment


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 References
 
Herpes zoster (HZ) is a common disorder caused by reactivation of latent varicella-zoster virus (VZV). Affecting 10–30% of people over their lifetime [1, 2], HZ causes substantial pain and morbidity. Although HZ is most clearly associated with aging [1, 2], people with medical disorders associated with decreased cell-mediated immunity, such as HIV [3, 4], lymphoma [5] and other malignancies [6], are also at a higher risk for HZ. In addition to specific disorders, treatments which attenuate cell-mediated immunity increase the risk of HZ.

Rheumatoid arthritis (RA) is a common immunological disorder in which immuno suppressive and immunomodulatory treatments are frequently used. However, there is essentially no scientific literature about the risk of HZ in patients with RA. In contrast, the association of HZ and systemic lupus erythematosus is well-documented [7–10]. A number of drugs that have been used for treatment of RA have been implicated in the development of HZ, including cyclophosphamide, azathioprine and methotrexate (MTX). However, there are only rare studies in RA describing such associations and only one report that provided quantitative data regarding the risk and rate of HZ association with RA treatment (MTX) [11]. The issue of treatment effect takes on additional relevance with increasing use of biologic immunomodulatory agents in patients with RA.

As HZ is often reported as an adverse event in clinical trials, it is important to understand baseline HZ rates. In addition to treatment effect, the question as to whether patients with RA are at an increased risk for HZ is not known, regardless of whether the comparison group is the general population [1, 2, 12] or control populations with other chronic illnesses. In the report that follows, we investigated rates and risk factor for HZ in patients with RA and non-inflammatory musculoskeletal (MSK) disorders.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 References
 
Patients in this study were participants in the National Data Bank for Rheumatic Diseases (NDB) longitudinal study of rheumatic disease outcomes. NDB participants are recruited on an ongoing basis from the practices of US rheumatologists, and are followed prospectively with semi-annual, detailed, 28-page questionnaires, as previously described [4–7]. Referrals to the data bank were made from the practices of 995 US rheumatologists. In this report, we studied patients with RA who had completed at least one semi-annual questionnaire during the period between 1 January 2001 and 30 June 2005. We chose this starting date to be sure that the participants had the opportunity for prescription of biologic RA treatments. For comparison purposes we also studied rheumatic disease patients who did not have RA or other inflammatory disorders. These patients with non-inflammatory MSK disorders included those diagnosed with osteoarthritis, back pain syndromes and similar disorders. To avoid the possibility of misclassification due to over-reporting, we excluded patients diagnosed with fibromyalgia. This study was approved by the Via Christi Institutional Review Board (IRB), Wichita, KS, USA. All participants signed an IRB-approved informed consent.

The presence of HZ was assessed in questions regarding previous and current HZ. At the time of entry into the study, patients reported whether they had ever been diagnosed with HZ. New cases of HZ were determined by patient self-report. The specific question asked to all participants was ‘Did you have or were you treated for Shingles (Herpes Zoster) ... between ... [dates of the 6 months period prior to questionnaire administration were inserted here].’ In 2005, we conducted validation studies of patients who had reported a diagnosis of HZ between 2001 and 2005. We used a standardized questionnaire which included a text field for a full description of the skin lesions. From 429 incident self-reports of HZ that were able to be validated, we determined that the false positive rate was 20.5%, and that this rate was consistent over the 5-y period. Of this group, 88 were reclassified as non-cases, leaving 341 true cases. We were unable to evaluate 138 reports because of withdrawal of consent, non-response or death.

We adjusted for the false positives in incidence rate analyses by generating random deviates from a binomial distribution based on 341 true cases of zoster out of 429 reported cases and then repeatedly sampling from this distribution to probabilistically re-classify the 138 positive cases for which validation data was not available. This random process was repeated 1000 times for incidence rates, and the results and confidence intervals reported are bootstrap estimates that account for uncertainty due to misclassification. Under the assumption that 20.5% of unvalidated cases were misclassified, the overall correct classification is 94.2% [(341 correctly classified + 79.5% x 138)/(341 + 138)]. For analysis of the incident cases, all patients with prior history of HZ were excluded. Of 17 243 patients available during the study period, 12 335 had not experienced HZ previously, and this group was used for the incidence rate studies. For patients who reported more than one case of HZ, only the first case (incident case) was analysed.

At each questionnaire assessment, we recorded socioeconomic and demographic variables as well as treatments. Patients also reported functional status using the Stanford Health Assessment Questionnaire (HAQ) [13, 14]. We determined pain, global severity and fatigue by visual analogue scales (VAS) [15]. Global measures of health included the Medical Outcomes Study Short Form 36 (SF-36) physical component scale (PCS) and mental component scale (MCS) [16, 17]. The effect of comorbidity was assessed by a comorbidity score, which is the sum of 11 present or past comorbid conditions reported by the patient. Conditions include cancer, stroke, fracture, renal, neurological, endocrine, gastrointestinal, cardiovascular, pulmonary, genito-urinary and psychiatric problems.

Although we examined the data in preliminary analyses with covariates that included pain, patient global and SF-36 variables, inclusion of these variables did not change the results of the analyses nor did they perform better for the control of the disease activity as compared with using only the HAQ. Therefore, we report results for the HAQ and not the other variables in the interest of simplicity and clarity. The HAQ is included in all the clinical trials of RA and observational studies.

Only 37 patients used cyclophosphamide during the study period 2001–05 and unstable HR estimates were obtained. To provide a useful estimate of cyclophosphamide effect, we extended the analyses of cyclophosphamide to begin in 1998 instead of 2001. In the multivariable analyses that included cyclophosphamide, all of the covariates of Table 1 were included, but we did not include specific biologics (etanercept, infliximab and adalimumab) as dummy variables, as these drugs were not in use in 1998. The result of this analysis is reported in the univariate analysis table (Table 2) with an appropriate foot note.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Multivariate treatment predictors of HZ in RAa

 

View this table:
[in this window]
[in a new window]

  		TABLE 2. Treatment predictors of HZ in RA and MSKa

 
To enhance comparability, age-adjusted incidence rates (Table 3) were adjusted to the US 2000 population [18]. Adjustment of incidence rates across all age and sex strata was performed by weighting NDB data for individuals within each age and sex strata to reflect the expected proportion in the general population.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Incidence of HZ in rheumatic disease patients

 
Tests of group differences for Table 4 were performed using t-tests or chi-square tests, as appropriate. Lifetime prevalence data were analysed by Poisson regression with robust variance estimators and adjusted for age at the last observation (end of follow-up). This adjustment is approximate as the age at which zoster occurred is not known. Incidence rate comparisons were performed using generalized linear models with a Poisson link. Predictors of HZ were examined with Cox regression using time-varying covariates. Predictor variables were lagged such that treatments in the previous 6-month period were used to predict HZ in the following 6-month period.


View this table:
[in this window]
[in a new window]

  		TABLE 4. Demographic, illness severity and treatment status of 10 614 RA patients and 1721 patients with MSK at the study onset

 
Figure 1 is a running line smooth of the y variable (lifetime zoster) on all x variable predictors (age and sex) simultaneously; that is, the smooth is adjusted for all the other covariates [19]. Using a simple type of backfitting, the resulting smoother is a locally linear function of the predictors for each observation.


Figure 1
View larger version (19K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 1. The lifetime prevalence of HZ as a function of age for both sexes combined. Lines above and below the centre line represent 95% CI.

 
Data were analysed using Stata (College Station, TX, USA) version 9.1. Statistical significance was set at the 0.05 level, confidence intervals were established at 95% and all the tests were two-tailed.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 References
 
Demographics and clinical status
The characteristics and clinical status of the 12 335 rheumatic disease patients who had not had HZ at the time of their entry into the study are shown in Table 4. There are a number of noteworthy differences between the RA and non-RA patients. With respect to demographics, patients with MSK were older, 67.0 (S.D. 11.4) vs 60.1 (12.8) yrs and were less likely to be current (7.2 vs 13.0%) or lifetime smokers (44.9 vs 55.9%). Their HAQ scores were also lower, with a difference in HAQ score between the groups of 0.12 (95% CI 0.09–0.16), P < 0.001.

Among treatment variables, more than two-thirds of all the patients used non-steroidal anti-inflammatory drugs (NSAIDs). MSK patients more often used NSAIDs overall (72.6 vs 67.0) and COX-2 NSAIDs (44.6 vs 33.8%), but not non-COX-2 NSAIDs (33.0 vs 45.4%). In the RA cohort, MTX was the most commonly used treatment (56.7%) followed by prednisone (41.1%). Cyclophosphamide was used by 0.2%.

Lifetime prevalence
Including patients with prior HZ, the lifetime prevalence of HZ in the 17 243 participants was: all patients 16.9% (95% CI 16.3–17.4), RA 16.5% (95% CI 15.9–17.1) and MSK 19.2% (95% CI 17.8–20.8). The lifetime prevalence of HZ increased with each 10 yrs of age: prevalence ratio 1.12 (95% CI 1.09–1.15) per year, P < 0.001 (Fig. 1). The prevalence ratio was increased for women compared with men [1.3 (95% CI 1.2–1.4), P < 0.001). Adjusted for age and sex, the relative risk for HZ in RA compared with MSK was 0.92 (95% CI 0.83–1.02), P = 0.105.

Incidence rates
The crude incidence of new cases of HZ per 1000 RA and MSK patient-years was 13.4 (95% CI 12.2–14.6) for all the patients combined, 13.2 (95% CI 11.9–14.5) for those with RA and 14.6 (95% CI 11.2–18.1) for those with MSK. Adjusted for age and sex, the incidence rate ratio for the development of HZ was 1.0 (95% CI 0.7–1.3) for RA compared with MSK, P = 0.812. The incidence rates for the groups standardized to the age and sex of the general population are presented in Table 3. The wider confidence intervals around the standardized rates reflect uncertainty regarding rates in younger patients.

Risk factors for the development of herpes zoster
As the predictors and coefficients for the development of HZ were similar between RA and MSK, we present predictors of HZ for both groups together except for RA treatment variables. Zoster was more common in women [HR 1.3 (95% CI 1.0–1.6)] (Table 5). In addition, each 10-yr increase in age was associated with a 10% increase in the risk of developing HZ. Persons who developed HZ were less likely to have diabetes [HR 0.7 (95% CI 0.5–0.1] P = 0.034]. Lagged functional status, as measured by the HAQ, increased the risk of HZ significantly [HR 1.3 (95% CI 1.1–1.5) P = 0.001].


View this table:
[in this window]
[in a new window]

  		TABLE 5. Univariable predictors of HZ (all patients)

 
Zoster was associated with NSAID use overall [HR 1.3 (95% CI 1.1–1.6)] and with COX-2 NSAIDS [HR 1.3 (95% CI 1.1–1.6)], but not with non-COX-2 NSAIDs [HR 0.9 (95% CI 0.8–1.1)] (Table 2). Among RA patients, the largest treatment effect was seen with cyclophosphamide [HR 4.2 (95% CI 1.6–11.5)] followed by azathioprine [HR 2.1 (95% CI 1.3 to 3.3)], prednisone [HR 1.6 (95% CI 1.3–1.9)] and leflunomide [HR 1.4 (95% CI 1.1–1.8)]. The HR was reduced for adalimumab [HR 0.4 (95% CI 0.2–0.7)]. To control for illness severity, we also performed these analyses after adding HAQ as an explanatory variable (column 4). As there was little effect on the HRs of adding HAQ, we have omitted the additional HRs from the table. However, the P-values from the added HAQ covariates are displayed in column 4 of the table. They are essentially unchanged from the P-values of the model without HAQ as a covariate.

These analyses were extended multivariately in Table 1. In addition to the treatment variables included in the table, age, sex, education, smoking, HAQ, diabetes and comorbidity were also included in the model. Positive associations were noted between HZ and prednisone, COX-2 NSAIDs, leflunomide and azathioprine. MTX, infliximab and etanercept were not associated with the development of HZ.


   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
References
 
The age-adjusted annualized incidence of HZ is around 3.2–3.3 per 1000 person-years in large population-based studies [6, 12], and the incidence of HZ increases with age. From the MarketScan US health plan enrollment study of 9152 incident cases and 2.4 million person-years of follow-up [6], robust estimates of the rates in various age groups can be obtained that are relevant to patients with arthritis in the current study. In the MarketScan study, the age-dependent annualized incidences (95% CI) per 1000 person-years were: ages 40–49 yrs: 2.9 (2.7–3.0), 50–59 yrs: 4.6 (4.4–4.8), 60–69 yrs: 6.9 (6.6–7.2), 70–79 yrs: 9.5 (9.0–9.9) and ≥80 yrs: 10.9 (10.2–11.6). The age-related increase in incidence can be attributed to the gradual loss of cell-mediated immunity that occurs with aging [1, 20]. In the current study, the mean age of participants was 60.1 yrs and the annualized incidence rate for HZ was 13.4 (12.2–14.6) per 1000 person-years, a rate that is approximately double that is noted by Insinga and colleagues [6], considering the relevant age categories.

In addition to the effect of age as demonstrated in population-based studies, HZ is also known to be increased in persons with disorders and treatments known to decrease cell-mediated immunity, including cancer, transplantation, HIV, other immunosuppressive disorders and/or immunosuppressive treatments. Incidence rates in these settings have ranged from ~9 or 10 [6, 21] to as high as 25–91 [1] per 1000 person-years, depending on the condition and study methodology. The incidence found in this study of ~12–14 per 1000 person-years, falls within expected rates for individuals with diminished cell-mediated immunity.

There is evidence that cell-mediated immunity is generally impaired in people with chronic illnesses generally [20]. Such individuals have increased rates of influenza and impaired response to vaccination [22–24]. In addition, persons failing to meet a rigorous definition of perfect health because of the lack of exercise or the use of medications for hypertension and osteoarthritis (‘deterioration of health status’) have lower levels of IL-2 and IL-6 [20, 25]. These findings have considerable relevance to the current study in which all participants had chronic illness and impaired health status, and almost all were using medications.

With respect to rheumatic diseases, published data indicate that the risk of HZ is a combined function of age, overall health status, specific therapies and disorders. However, there have been no studies on the risk of development of HZ in general populations of patients with RA. In part, this is due to the overall low incidence of HZ and the consequent requirement of very large samples of RA patients. Instead, most studies of RA report on patients treated with specific therapies in controlled trials or from cohorts with special characteristics or from case reports. Selection characteristics of patients in these reports do not allow simple extrapolation to the general population of RA patients. However, they can provide useful insights as well as estimates of the risk of HZ in patients treated with specific therapies. We describe some of those reports below.

Of 543 patients (mean age of 54 yrs at the onset of treatment) who received lymphocytotoxic monoclonal antibody (mAb) therapy causing prolonged lymphopenia, the annualized rate of HZ was ~65 per 1000 patient-years of follow-up [26]. McCarty et al. [27] combined MTX, azathioprine and anti-malarials in the treatment of 169 patients. They reported that the most striking finding was the development of HZ (17 patients) and second attacks of varicella (two patients) during an average follow-up of 7 yrs. The rates for HZ and the combined group were 14.4 and 16.1 per 1000 patient-years, respectively.

Among 1540 patients treated with anti-tumour necrosis factor (TNF) therapy in a Spanish biologics registry [28], the annualized rate of HZ during 1.1 yrs of follow-up was 47 per 1000 patient-years. The mean age of this group was 51 yrs and 82.1% had RA. The clinical (severity) characteristics of these patients were not described. In contrast, we did not find an association between infliximab or etanercept use and HZ, and this is in agreement with clinical trial results and the package inserts for these products that list no warnings. Although the use of adalimumab was associated with a reduced risk of HZ in the current study, adalimumab was only used very late in the course of the study, and the adalimumab result is a statistical artefact relating to differential censoring. However, we present the data for completeness.

Although MTX is often considered to be a risk factor for the development of HZ, this impression appears to be the result of a few case reports [29, 30] and the warning seen in the US package insert for Rheumatrex (MTX) which states only that ‘There have been reports of infections [including] ... herpes zoster’ in patients receiving MTX. Two other studies may contribute to the apparent association, the McCarty et al. [27] study and a report from Antonelli et al. [11]. In the only quantitative study of the association between MTX and HZ, these authors reported on 187 RA patients receiving MTX [11]. They found that HZ occurred in ‘14.5 cases per 1000 patient-years in our MTX-treated RA patients, as compared with the general population incidence of 1.3–4.8 cases per 1000 patient-years.’ Although MTX is now used in all biologic trials, there are no reports of increased rates of HZ among patients receiving these therapies. In the current study, the overall rate of HZ was 14.7 per 1000 patient-years, and the univariate and multivariate HR was 1.1 (95% CI 0.9–1.3) and 1.0 (95% CI 0.8–1.3) for the use of MTX. It would seem fair to conclude that there is no demonstrable association between MTX and the development of HZ. With the availability of additional data regarding the rate of HZ in RA, the results of HZ observed by Antonelli et al. [11] can be reinterpreted to show no association between MTX and HZ.

The US package insert for leflunomide indicates that HZ may occur in 1–3% of the patients. Adjusted for covariates, in the current study we found an increased risk of HZ for leflunomide [HR 1.4 (95% CI 1.1–1.8)]. However, this increase is small and falls within the observed incidence noted in the package insert.

Azathioprine is another therapy for which non-quantitative associations with HZ are described in the literature [31–33]. Although the number of patients in the current study receiving azathioprine was relatively small (329), the risk of HZ was doubled in those patients after adjusting for other treatments and disease severity: HR 2.0 (95% CI 1.2–3.3).

Cyclophosphamide use is associated with increased rates of HZ in lupus [9, 34], and HZ has been noted in systematic reviews of RA patients treated with cyclophosphamide [35] and in case reports of multi-drug therapy [33]. However, there are no quantitative data on the degree of association between cyclophosphamide and HZ. In the current study, we found a substantial increase in HZ risk in cyclophosphamide-treated patients [HR 4.2 (95% CI 1.6–11.5)], after adjustment for age, sex and functional status.

Prednisone use was common in this cohort (41.1%), and in multivariable analyses was associated with an increased risk of HZ [HR 1.5 (95% CI 1.2–1.8)]. COX-2 NSAIDs have been associated with HZ in <2% of the patients in clinical trials. We observed that use of COX-2-specific NSAIDs was associated with development of HZ [HR 1.3 (95% CI 1.1–1.6)]. This effect was noted both in RA and MSK patients. However, a rate of <2% noted in clinical trials is not different from the results of the current study.

Among the important findings of this study was that the age- and sex-adjusted incidence rate of HZ did not appear to differ between RA and MSK patients. While it might be suspected that RA as an ‘immunological disorder’ would have a higher rate of HZ, we were unable to find any data in the medical literature concerning this hypothesis. The results of the current study, however, support the conclusion that patients with RA are not at an increased risk to develop HZ compared with patients with other chronic non-inflammatory rheumatic disorders. However, the rate in RA patients and those with MSK (Table 3) is greater than that found in the general population: 3.2–3.3 per 1000 person-years [6, 12] overall and 6.9 (6.6–7.2) for the 60–69-year-old age group [6]. These data offer support to the hypothesis that chronic illness, in general, may predispose to HZ [1, 20].

We also found that patients with diabetes were less likely to develop HZ [HR 0.7 (95% CI 0.5–1.0)]. However, a prior study of HZ in diabetics found no association between the disorders [36]. We did not find that current cigarette smoking was associated with the development of HZ [HR 0.6 (95% CI 0.8–1.2)]; nor was lifetime smoking associated with HZ [HR 1.0 (95% CI 0.8–1.2)]. Although smoking can have a deleterious effect on cell-mediated immunity in vitro [37], it was found to have no effect on the risk of HZ in a clinical study of the British general practice database [1] and to have a protective effect in a stratified probability sample of community-dwelling persons more than 65 yrs of age, with a relative risk of 0.47 (95% CI 0.25–0.89) [38]. Apparent beneficial effects of unhealthy practices or of illness can sometimes be attributed to the healthy worker survivor effect [39, 40].

We did not describe HIV infection in our analyses (Tables 4 and 5) as no patients in our cohort reported this infection and none were using HIV treatments. The absence of HIV reflects the older ages of patients with rheumatic diseases in this study. In addition, the overall rate of cancer did not differ between RA patients and controls in this report, although cancer rates are increased for lymphoma and lung cancer in RA patients in this cohort.

Among the potential limitations of this report is the use of patient diagnoses of HZ. Schmader et al. [41] studied the accuracy of self-reported diagnoses of ‘shingles’ using a validated questionnaire and physician diagnosis [41]. Of 31 subjects who reported HZ in their study, the false positive rate compared with physician diagnosis was of 3.2% (95% CI 0–6.1). Two of the 42 subjects reported HZ, but that diagnosis was not confirmed by the questionnaire [false positive rate 4.8% (95% CI 0.64)]. However, one of the two cases was confirmed by a physician. No false negatives were noted. The overall false positive rate in that study was 4.6 (95% CI 0.08–17.2). The Behavioral Risk Factor Surveillance System has continued to use this self-report methodology in epidemiological reports [42]. In the current study, we determined the false positive rate to be 20.5% and we corrected the incidence rates to reflect that misclassification in patients in whom validation was not possible. We did not correct for misclassification in our predictive analyses of Tables 5, 2 and 1 because we did not know exactly which patients were misclassified, and we estimate a 5.8% misclassification rate overall. However, the confidence intervals reported include the effect of misclassification. We also conducted sensitivity analyses by excluding patients who might be misclassified. There were no clinically significant differences noted. We also studied 200 consecutive patients who did not report HZ. No false negatives were noted.

In conclusion, the incidence of HZ is increased in RA and MSK compared with population-based rates. However, the rate of HZ in RA does not appear to be increased compared with MSK. After adjustment for severity, cyclophosphamide, azathioprine, leflunomide, prednisone and COX-2 NSAIDs, but not MTX or biologics, were the risk factors for HZ.

The authors have declared that the National Data Bank for Rheumatic Diseases has received funding support from pharmaceutical companies, including Merck, Pfizer, Tap, Wyeth-Australia, Aventis, Novartis, Bristol–Myers Squibb, Amgen and Centocor.


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 

  1. Thomas SL and Hall AJ. (2004) What does epidemiology tell us about risk factors for herpes zoster? Lancet Infect Dis 4:26–33.[CrossRef][ISI][Medline]
  2. Brisson M, Edmunds WJ, Law B, et al. (2001) Epidemiology of varicella zoster virus infection in Canada and the United Kingdom. Epidemiol Infect 127:305–14.[Medline]
  3. Veenstra J, Krol A, van Praag RM, et al. (1995) Herpes zoster, immunological deterioration and disease progression in HIV-1 infection. AIDS 9:1153–8.[ISI][Medline]
  4. Buchbinder SP, Katz MH, Hessol NA, et al. (1992) Herpes zoster and human immunodeficiency virus infection. J Infect Dis 166:1153–6.[ISI][Medline]
  5. Guinee VF, Guido JJ, Pfalzgraf KA, et al. (1985) The incidence of herpes zoster in patients with Hodgkin's disease. An analysis of prognostic factors. Cancer 56:642–8.[CrossRef][ISI][Medline]
  6. Insinga RP, Itzler RF, Pellissier JM, Saddier P, Nikas AA. (2005) The incidence of herpes zoster in a United States administrative database. J Gen Intern Med 20:748–53.[CrossRef][ISI][Medline]
  7. Park HB, Kim KC, Park JH, et al. (2004) Association of reduced CD4 T cell responses specific to varicella zoster virus with high incidence of herpes zoster in patients with systemic lupus erythematosus. J Rheumatol 31:2151–5.[ISI][Medline]
  8. Pope JE, Krizova A, Ouimet JM, Goodwin JL, Lankin M. (2004) Close association of herpes zoster reactivation and systemic lupus erythematosus (SLE) diagnosis: case-control study of patients with SLE or noninflammatory musculoskeletal disorders. J Rheumatol 31:274–9.[ISI][Medline]
  9. Kang TY, Lee HS, Kim TH, Jun JB, Yoo DH. (2003) Clinical and genetic risk factors of herpes zoster in patients with systemic lupus erythematosus. Rheumatol Int.
  10. Manzi S, Kuller LH, Kutzer J, et al. (1995) Herpes zoster in systemic lupus erythematosus. J Rheumatol 22:1254–8.[ISI][Medline]
  11. Antonelli MA, Moreland LW, Brick JE. (1991) Herpes zoster in patients with rheumatoid arthritis treated with weekly, low-dose methotrexate. Am J Med 90:295–8.[ISI][Medline]
  12. Donahue JG, Choo PW, Manson JE, Platt R. (1995) The incidence of herpes zoster. Arch Intern Med 155:1605–9.[Abstract]
  13. Fries JF, Spitz PW, Young DY. (1982) The dimensions of health outcomes: the health assessment questionnaire, disability and pain scales. J Rheumatol 9:789–93.[ISI][Medline]
  14. Fries JF, Spitz PW, Kraines RG, Holman HR. (1980) Measurement of patient outcome in arthritis. Arthritis Rheum 23:137–45.[ISI][Medline]
  15. Wolfe F, Hawley DJ, Wilson K. (1996) The prevalence and meaning of fatigue in rheumatic disease. J Rheumatol 23:1407–17.[ISI][Medline]
  16. Stewart AL, Hays RD, Ware JE. (1988) The MOS short-form general health survey. Med Care 26:724–35.[ISI][Medline]
  17. Ware JEJ and Kosinski M. (2001) The SF-36® Physical and Mental Health Summary Scales: A Manual for User's of Version 1, Second Edition 1st (QualityMetric, Inc., Lincoln).
  18. US Census Bureau. 2000 resident population estimates of the United States by age and sex. US Census Bureau [2004 Available from: URL:http://wonder.cdc.gov/census.shtml.
  19. Royston P and Cox NJ. (2005) A multivariable scatterplot smoother. Stata J 5:405–12.
  20. Castle SC. (2000) Clinical relevance of age-related immune dysfunction. Clin Infect Dis 31:578–85.[CrossRef][ISI][Medline]
  21. di Luzio PU, Arpinelli F, Visona G. (1999) Herpes zoster and its complications in Italy: an observational survey. J Infect 38:116–20.[CrossRef][ISI][Medline]
  22. Burns EA and Goodwin JS. (1997) Immunodeficiency of aging. Drugs Aging 11:374–97.[ISI][Medline]
  23. Mullooly JP, Bennett MD, Hornbrook MC, et al. (1994) Influenza vaccination programs for elderly persons: cost-effectiveness in a health maintenance organization. Ann Intern Med 121:947–52.[Abstract/Free Full Text]
  24. Gravenstein S, Drinka P, Duthie EH, et al. (1994) Efficacy of an influenza hemagglutinin-diphtheria toxoid conjugate vaccine in elderly nursing home subjects during an influenza outbreak. J Am Geriatr Soc 42:245–51.[ISI][Medline]
  25. Mysliwska J, Bryl E, Foerster J, Mysliwski A. (1999) The upregulation of TNF alpha production is not a generalised phenomenon in the elderly between their sixth and seventh decades of life. Mech Ageing Dev 107:1–14.[CrossRef][ISI][Medline]
  26. Issacs JD, Greer S, Sharma S, et al. (2001) Morbidity and mortality in rheumatoid arthritis patients with prolonged and profound therapy-induced lymphopenia. Arthritis Rheum 44:1998–2008.[CrossRef][ISI][Medline]
  27. McCarty DJ, Harman JG, Grassanovich JL, Qian C, Klein JP. (1995) Combination drug therapy of seropositive rheumatoid arthritis. J Rheumatol 22:1636–45.[ISI][Medline]
  28. Gomez-Reino JJ, Carmona L, Valverde VR, Mola EM, Montero MD. (2003) Treatment of rheumatoid arthritis with tumor necrosis factor inhibitors may predispose to significant increase in tuberculosis risk: a multicenter active-surveillance report. Arthritis Rheum 48:2122–7.[CrossRef][ISI][Medline]
  29. Ching DWT. (1995) Severe, disseminated, life threatening herpes zoster infection in a patient with rheumatoid arthritis treated with methotrexate. Ann Rheum Dis 54:155.[Free Full Text]
  30. Shiroky JB, Frost A, Skelton JD, Haegert DG, Newkirk MM, Neville C. (1991) Complications of immunosuppression associated with weekly low dose methotrexate. J Rheumatol 18:1172–5.[ISI][Medline]
  31. Korelitz BI, Fuller SR, Warman JI, Goldberg MD. (1999) Shingles during the course of treatment with 6-mercaptopurine for inflammatory bowel disease. Am J Gastroenterol 94:424–6.[CrossRef][ISI][Medline]
  32. Speerstra F, Boerbooms AM, Van de Putte LB, Van Beusekom HJ, Kruijsen MW, Vandenbroucke JP. (1982) Side-effects of azathioprine treatment in rheumatoid arthritis: analysis of 10 years of experience. Ann Rheum Dis 41:Suppl. 1, 37–9.
  33. Csuka M, Carrera GF, McCarty DJ. (1986) Treatment of intractable rheumatoid arthritis with combined cyclophosphamide, azathioprine, and hydroxychloroquine. A follow- up study. JAMA 255:2315–9.[Abstract]
  34. Takada K, Illei GG, Boumpas DT. (2001) Cyclophosphamide for the treatment of systemic lupus erythematosus. Lupus 10:154–61.[Abstract/Free Full Text]
  35. Suarez-Almazor ME, Belseck E, Shea B, Wells G, Tugwell P. (2000) Cyclophosphamide for treating rheumatoid arthritis. Cochrane Database Syst Rev 4:CD001157.[Medline]
  36. Ragozzino MW, Melton LJ, Kurland LT. (1983) Herpes zoster and diabetes mellitus: an epidemiological investigation. J Chronic Dis 36:501–5.[CrossRef][ISI][Medline]
  37. Moszczynski P, Zabinski Z, Moszczynski P Jr, Rutowski J, Slowinski S, Tabarowski Z. (2001) Immunological findings in cigarette smokers. Toxicol Lett 118:121–7.[CrossRef][ISI][Medline]
  38. Schmader K, George LK, Burchett BM, Hamilton JD, Pieper CF. (1998) Race and stress in the incidence of herpes zoster in older adults. J Am Geriatr Soc 46:973–7.[ISI][Medline]
  39. Richardson D, Wing S, Steenland K, McKelvey W. (2004) Time-related aspects of the healthy worker survivor effect. Ann Epidemiol 14:633–9.[CrossRef][ISI][Medline]
  40. Arrighi HM and Hertz-Picciotto I. (1994) The evolving concept of the healthy worker survivor effect. Epidemiology 5:189–96.[ISI][Medline]
  41. Schmader K, George LK, Newton R, Hamilton JD. (1994) The accuracy of self-report of herpes zoster. J Clin Epidemiol 47:1271–76.[CrossRef][ISI][Medline]
  42. Yih WK, Brooks DR, Lett SM, et al. (2005) The incidence of varicella and herpes zoster in Massachusetts as measured by the Behavioral Risk Factor Surveillance System (BRFSS) during a period of increasing varicella vaccine coverage, 1998–2003. BMC Public Health 5:68.[CrossRef][Medline]
Submitted 6 July 2006; revised version accepted 15 August 2006.
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
45/11/1370    most recent
kel328v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (6)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Wolfe, F.
Right arrow Articles by Chakravarty, E. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wolfe, F.
Right arrow Articles by Chakravarty, E. F.
Related Collections
Right arrow Rheumatoid Arthritis
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?