PATIENTS AND METHODS
- Top of page
- PATIENTS AND METHODS
The patients participated in a larger cross-sectional study of disease activity parameters in various inflammatory rheumatic diseases (4). Hospital inpatients, 70 of whom had RA according to the 1987 revised criteria of the American College of Rheumatology (formerly the American Rheumatism Association) (5), were included consecutively during the period from December 1991 to May 1992. The patients responded separately to each of the 6 questions regarding sicca symptoms that are used in the preliminary European criteria for SS (6).
Details regarding baseline inclusion and exclusion, clinical evaluations, laboratory tests, and the sicca symptoms questionnaire have been described previously (4, 7). In addition to a diagnosis of RA, inclusion criteria required patients to be between 18 and 80 years of age and in functional class I to III (8). Patients with severe infections, malignancy, pregnancy, or surgery within the last 4 weeks were excluded from the study. Recorded data included the number of swollen and tender joints, duration of morning stiffness, the Health Assessment Questionnaire (HAQ) score, pain last week as recorded on a 100-mm visual analog scale (VAS), patient's and investigator's overall assessments of disease activity (by VAS), current medication use, Westergren erythrocyte sedimentation rate, C-reactive protein, rheumatoid factor (RF), antinuclear antibodies (ANA), anti–SSA/SSB antibodies, and immunoglobulin G. The cut-off levels for a positive RF and positive ANAs were set at titers of 128 (9, 10). Also recorded were the number of years of formal education, stressful life events last year as recorded on a VAS, the presence or absence of extraordinary fatigue, and the Larsen score for radiographs of the hands (11).
In 1997, all RA patients who had been evaluated at baseline were invited by letter to participate in the present study. A questionnaire identical to the one at baseline was administered and answered under the same conditions as during the primary investigations.
Clinical and laboratory evaluations at followup.
The patients were subjected to identical evaluations and investigational procedures as they were at baseline. In addition, a Schirmer's test (12) was performed on both eyes. This test had not been part of the investigations at inclusion.
Data were frequently not normally distributed and we used the median and the interquartile range as measures of central tendency and dispersion. Differences between groups were tested with the Mann-Whitney test. A chi-square test was applied to dichotomous variables. Yates' continuity correction was adopted whenever cross-table cell numbers were small. The McNemar test was used to assess the general increase of sicca symptoms during the study period.
We used multiple logistic regression to study potential baseline risk factors for sicca symptoms at followup. To reduce the number of variables included in the multivariate model, we selected risk factors associated with sicca symptoms in the univariate analyses with P values less than 0.10. A stepward, forward algorithm was used to select factors for the multivariate models, and the odds ratio (OR) and 95% confidence interval (95% CI) were computed. In this analysis, pain was used as a continuous variable (0–100 mm), and the OR expressed the increased risk of sicca symptoms at followup, per mm on the VAS scale for pain. The above statistical analyses were performed using SPSS release 9.0.0 software (SPSS, Chicago, IL).
Confidence intervals for the proportions of patients with positive responses both at baseline and at followup, and for those negative at baseline and positive at followup, were computed using the EpiCalc 2000 version 1.02 computer program (Brixton Health, Llanidloes, UK).
- Top of page
- PATIENTS AND METHODS
The present prospective study of sicca symptoms in patients with RA showed that symptoms were remarkably constant over time. We found that 84.2% of patients reporting symptoms at baseline also reported them at followup. During the study period, there was a substantial increase (52.6%) in the number of patients reporting sicca symptoms.
Of the 14 patients lost to followup, 3 (21.4%) had sicca symptoms at inclusion. This is somewhat lower than the prevalence among the patients followed (33.9%), but not significantly different (P = 0.29). If the 14 patients had been included, the overall percentage of sicca symptoms at inclusion would have been lower than 33.9%. The large relative increase of sicca symtoms during followup is thus probably real, and not influenced positively by a of loss to followup bias.
Previous studies have reported prevalences of sicca symptoms in RA in the range of 25% to 65% (13–17). Our findings of prevalences of 33.9% at inclusion and 51.8% at followup are within this range.
The prevalence in the general population is generally lower (18–23), but seems to increase with age. Thomas et al (18) reported a prevalence of 21% (both for dry eyes and dry mouth) for those younger than 55 years, and 27% (dry mouth) and 30% (dry eyes) for those older than 55 years. Schein et al (20) reported a prevalence as high as 27.4% in an elderly population, with 25.3% in the age group 65–69 years.
However, some studies of normal populations indicate that although the prevalence of mouth dryness seems to increase in a linear manner with age, the prevalence of eye dryness levels out in elderly age groups (19, 20, 23). This contrasts with our findings of a major increase of ocular sicca symptoms during followup, and constitutes an interesting difference between RA patients and the general population in the development of sicca symptoms over time. Previous cross-sectional studies support the notion of a disease-specific increased expression of ocular sicca symptoms in RA patients. Andonopoulos et al (14) found that 38.2% of RA patients had subjective xerophthalmia, whereas only 5.9% had subjective xerostomia. Patients with RA may have other ocular manifestations, most notably scleritis. However, the prevalence of scleritis in patients with RA is only between 0.67% and 6.3%, and symptoms are usually different from sicca symptoms (24). The presence of scleritis is thus an unlikely explanation for the large time-dependent increase of ocular sicca symptoms in the present study.
There may be several reasons for the increase in the prevalence of sicca symptoms during the study period. Previous cross-sectional studies of RA patients have found associations between sicca symptoms and disease duration and age (17, 25). Age and the use of antidepressant drugs are associated with sicca symptoms in the general population (19, 20). These variables were not significantly associated with sicca symptoms in the cross-sectional analyses in our study (Table 4). However, sicca symptoms are unspecific, and presumably there are a number of factors with varying or minor predisposing effects that may not be identified statistically, except in large-scale studies. In addition, no record was made of the use of medications that may induce dryness, with the exception of antidepressant drugs.
We found no significant associations between ocular sicca symptoms and the Schirmer test in the cross-sectional analyses. This result is, however, not incompatible with the results of other studies of both RA patients and the general population, where only weak associations have been found (17, 21).
Sicca symptoms at baseline were, not surprisingly, identified as strong predictive factors for sicca symptoms at followup. Pain at baseline was also identified as an independent risk factor. In addition, the baseline variables of antidepressant drug use, number of swollen joints, use of oral steroids, and HAQ score were significantly associated with sicca symptoms at followup in the univariate analyses. With the exception of antidepressant drug use, these variables are associated with RA disease activity (26). It may be hypothesized that sicca symptoms in patients with RA have some association with disease activity, and that this may be one reason pain was identified as a predictive factor. It may also be speculated that some patients have a low threshold for reporting subjective symptoms, and that this may have contributed to the association between pain and sicca symptoms. Because of the size of the RA population in the present study, a cautious interpretation of the findings regarding risk factors are warranted. However, pain has also been found to be strongly associated with sicca symptoms in a previous cross-sectional study (17).
In conlusion, the results of this study showed that sicca symptoms in patients with RA are remarkably constant over time. There was also a considerable temporal increase of the proportion of patients having such symptoms. In contrast to findings in general populations, this tendency was especially pronounced for the ocular sicca symptoms. Sicca symptoms and pain at baseline were identified as risk factors for future sicca symptoms.