SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Objective

It has been found that women with rheumatoid arthritis (RA) have a poorer prognosis than men. However, the impact of age at symptom onset is unclear. We investigated the relationship between these factors and functional disability in patients with recent-onset inflammatory polyarthritis (IP).

Methods

A total of 3,666 patients (66% women) were registered with the Norfolk Arthritis Register between 1990 and 2008. Functional disability was assessed using the Health Assessment Questionnaire (HAQ), adjusted for age at HAQ completion. Linear random-effects models were used to examine HAQ score over time, by sex and age at symptom onset (early = age <55 years, late = age 55–74 years, very late = age ≥75 years).

Results

Women had higher HAQ scores over time than men (mean difference 0.29; 95% confidence interval [95% CI] 0.25, 0.34). Men with late-onset IP had lower baseline HAQ scores than men with early onset (mean difference −0.14; 95% CI −0.29, −0.001). Women had comparable baseline HAQ scores at all ages of onset. Both sexes showed the greatest rate of disability progression in patients with very late onset. Those with early onset had a steady level of disability over time. Adjustment for treatment received, comorbidities, and RA subgroup analysis produced results that were largely similar to the initial analysis.

Conclusion

Female patients have higher HAQ scores than male patients; patients with early symptom onset show the smallest sex difference. Older age at symptom onset is associated with an increasingly steep trajectory of disability progression. The impact of sex on outcome is evident at baseline, whereas the impact of age at symptom onset becomes apparent during long-term followup.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

It is widely accepted that the prevalence of rheumatoid arthritis (RA) is higher among women than men (1). It has also been reported that, compared to men, women have a poorer outcome in terms of functional outcome and disease activity (2–11), although no sex difference in terms of radiographic damage has been reported (5, 10).

A large-scale, multinational, cross-sectional study of more than 6,000 RA patients from 25 countries (Quantitative Standard Monitoring of Patients with RA) reported higher mean Health Assessment Questionnaire (HAQ) scores and Disease Activity Scores in 28 joints (DAS28), and a lower remission rate in women than men (9). The authors suggested that while there was no sex difference in drug treatments, their findings may be due to sex bias in the disease outcome measures used, especially HAQ score. For example, evidence from the general population shows that physical strength is associated with functional ability (12) and men are generally stronger than women; this is supported by studies that failed to find a radiographic difference between the sexes (5, 10). Another large study also reported higher mean HAQ and DAS28 scores for women with RA compared to men of a similar age. A study of 4,823 RA patients in Japan reported that women had higher overall disease activity and a faster rate of disease progression, in terms of disability, than men (8).

There is less agreement on the impact of age at symptom onset on outcome in patients with RA. A number of observational studies have reported that patients with late symptom onset have a more favorable outcome than those with early-onset RA (2, 13–16). However, of growing concern in an aging population are findings that patients with late-onset RA have a worse prognosis than early-onset patients, in terms of DAS28 score, HAQ score, and degree of radiographic damage (3, 5). On the other hand, there has also been a number of reports that early-onset and late-onset RA patients have a comparable prognosis in terms of radiographic damage, inflammation (as measured by C-reactive protein [CRP] level), disability (as measured by HAQ score), and disease activity (as measured by DAS28 score) (2, 17, 18).

We have previously reported a linear relationship between age at symptom onset and HAQ score; that women have significantly higher HAQ scores than men at years 1, 2, and 3 of followup; and that female sex and late onset (age ≥64 years) are key risk factors for development of disability (HAQ score ≥1.0) (6, 7, 19, 20). In this analysis, we have investigated the separate and combined impact of sex and age at symptom onset on functional outcome, modeled as trajectories of HAQ scores over time, in a cohort of prospectively-followed patients with recent-onset inflammatory polyarthritis (IP). Within this objective, we aimed to investigate the hypothesis that there are different phenotypes of IP among patients by sex or by age at symptom onset.

Significance & Innovations

  • By investigating the impact of sex, age at symptom onset, and the interaction between these 2 factors on disease outcome over time, we have presented a novel perspective on these relationships.

  • The unique approach of this investigation led to the finding that the impact of sex on functional outcome is evident from shortly after symptom onset, whereas the impact of age at symptom onset becomes evident only with long-term prospective followup.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Setting.

The Norfolk Arthritis Register (NOAR) is a large primary care–based inception cohort of patients with recent-onset IP presenting to a physician with ≥2 swollen joints persisting for ≥4 weeks. A detailed description of the register has been published elsewhere (1).

Patients.

The cohort included in the current investigation comprised consecutive patients registered with the NOAR between January 1990 and December 2008, who were followed to March 2010. Patients underwent a standardized assessment by a research nurse at baseline and years 1–3, 5, 7, 10, and 15. For this analysis, patients who joined the NOAR during 1995–1999 were included until year 2 only, as these patients followed a different pattern of assessment, as described previously (7).

Data collection.

At baseline assessment, demographic details, date of symptom onset, and medical history were recorded. A joint examination (51 joints) was carried out and blood samples were taken, from which serum was later tested for rheumatoid factor (RF; latex method, positive at titer ≥1:40) and anti–cyclic citrullinated peptide antibodies (Axis-Shield DIASTAT kit, positive at ≥5 units/ml). The DAS28 was calculated at baseline using the serum CRP level (mg/liter, by end-point immunoturbidimetric agglutination method) (21). Patients completed the British version of the Stanford HAQ (22) at baseline and years 1–5, 7, 8, 10, and 15. The HAQ is scored between 0 and 3, with 3 indicating the greatest degree of functional disability. At each assessment, patients were also asked to report any comorbid diseases and the name of any IP treatment they were currently taking. The clinical care of NOAR patients, including treatment regimen, is not determined by their participation in the NOAR, and is managed completely by their existing physician, in either primary or secondary care. The 1987 American College of Rheumatology (ACR) classification criteria for RA (23) were applied at each assessment.

Statistical analysis.

Differences in baseline characteristics between men and women were tested using Mann-Whitney and chi-square tests, as appropriate to the data characteristics. Average HAQ scores were compared over time by sex using a linear random-effects (LRE) model. LRE models overcome a number of the potential issues in analyzing longitudinal data. For example, LRE models permit the inclusion of individuals with different lengths of followup, and so data from patients who have been lost to followup or study cohorts with multiple study entry dates can be used and data wastage is minimized. In addition, LRE models can adjust for correlations between repeated measures, accounting for both between- and within-individual variability. As part of the LRE modeling process, we established whether there was a significant interaction between age at symptom onset and year of followup, and between sex and year of followup. The interaction with sex was not statistically significant (P = 0.33). However, the interaction with age at symptom onset was significant (P < 0.001); therefore, this was adjusted for in all LRE analyses reported here. We also adjusted our analyses for age at HAQ completion and calendar year of NOAR registration.

Further LRE models, stratified by sex, were used to compare HAQ scores over time by age at symptom onset as a categorical variable (“early onset” = age <55 years, “late onset” = age 55–74 years, “very late onset” = age ≥75 years), again unadjusted, and then adjusted for age at HAQ completion and calendar year of NOAR registration. These LRE models were then adjusted for treatment (measured as whether patients were taking nonbiologic or biologic disease-modifying antirheumatic drugs [DMARDs] or steroids at each assessment) and then additionally the number of comorbid conditions, both as time-varying factors. Another LRE model that included all patients was used to compare HAQ scores by age at symptom onset and sex simultaneously, adjusted for age at HAQ completion and calendar year of NOAR registration. Finally, this model was then carried out again including only the subgroup of patients who met the 1987 ACR criteria for RA at anytime during followup.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Cohort characteristics.

Of the total cohort, 2,402 patients (65.5%) were women. Women were more likely to have “early” symptom onset than men (age <55 years: 53.6% versus 41.3%, age 55–74 years: 37.6% versus 46.0%, age ≥75 years: 8.8% versus 12.7%); this difference was significant (P < 0.001) (Table 1). At registration, women had a longer symptom duration than men (median 6.8 versus 5.9 months; P = 0.005). Women were more likely to meet the 1987 ACR criteria for RA than men (44.8% versus 39.1%; P = 0.001). However, women were less likely to be positive for RF (32.6% versus 36.2%; P = 0.04). Women had lower median baseline concentrations of CRP (8.0 mg/liter versus 9.8 mg/liter; P = 0.005), but a higher median number of swollen and tender joints (2 versus 1; P < 0.001) and higher median DAS28 (3.8 versus 3.5; P < 0.001) and HAQ scores (0.88 versus 0.63; P < 0.001) than men. Men were more likely than women to receive a DMARD within the first 6 months following symptom onset (19.2% versus 15.1%; P = 0.002).

Table 1. Baseline characteristics by sex*
 Female patients (n = 2,402)Male patients (n = 1,264)P for difference
  • *

    IP = inflammatory polyarthritis; IQR = interquartile range; ACR = American College of Rheumatology; RA = rheumatoid arthritis; RF = rheumatoid factor; anti-CCP = anti–cyclic citrullinated peptide antibodies; CRP = C-reactive protein; DAS28 = Disease Activity Score in 28 joints; HAQ = Health Assessment Questionnaire; DMARD = disease-modifying antirheumatic drug.

  • Significant difference.

Age at IP onset, years  <0.001
 <55 (early onset), no. (%)1,288 (53.6)522 (41.3) 
 55–74 (late onset), no. (%)903 (37.6)581 (46.0) 
 ≥75 (very late onset), no. (%)211 (8.8)161 (12.7) 
Symptom duration, median (IQR) months6.8 (3.2–16.0)5.9 (3.0–13.5)0.005
Met 1987 ACR criteria for RA, no. (%)1,076 (44.8)494 (39.1)0.001
Positive for RF, no./total (%)693/2,128 (32.6)417/1,152 (36.2)0.04
Positive for anti-CCP, no./total (%)593/1,978 (30.0)345/1,050 (32.9)0.10
CRP concentration, total/median (IQR) mg/liter1,876/8.0 (2.0–19.0)998/9.8 (2.0–20.1)0.005
Swollen and tender joint count (51 joints), total/median (IQR)2,402/2 (0–6)1,264/1 (0–4)< 0.001
DAS28 score, total/median (IQR)1,876/3.8 (2.9–4.8)998/3.5 (2.5–4.6)< 0.001
HAQ score, total/median (IQR)2,375/0.88 (0.38–1.63)1,248/0.63 (0.13–1.25)< 0.001
DMARD treatment during followup, no. (%)1,373 (57.2)745 (58.9)0.30
DMARD treatment <6 months post–symptom onset, no./total (%)360/2,382 (15.1)239/1,248 (19.2)0.002

Sex and HAQ score.

On average, women had higher HAQ scores over time than men (unadjusted difference 0.24 [95% confidence interval (95% CI) 0.20, 0.29], difference adjusted for age at HAQ completion and calendar year of NOAR registration 0.29 [95% CI 0.25, 0.34]). This difference did not vary significantly with followup year. Additional adjustment for baseline symptom duration (difference 0.29; 95% CI 0.24, 0.36), and then for DMARD treatment within 6 months of symptom onset (difference 0.29; 95% CI 0.25, 0.34), had little impact on the sex difference in HAQ score over time.

Age at symptom onset and HAQ score.

Results from the unadjusted and adjusted (for age at HAQ completion) LRE models of HAQ score, by age at onset, are summarized for male patients in Table 2; the unadjusted estimates of HAQ score are shown only to illustrate the crude relationship between age at onset and HAQ score and the impact of adjustment for age at HAQ completion, and so are not subsequently referred to. Compared to men with early onset, men with late onset had lower HAQ scores for the first 10 years of followup. Men with very late onset had comparable HAQ scores to those with early onset up to year 5, and then higher HAQ scores thereafter. Compared to men with late onset, those with very late onset had higher HAQ scores at each year of followup, and the difference increased over time.

Table 2. Estimated HAQ score by age at symptom onset in male patients*
Year of followupAge at symptom onset <55 yearsAge at symptom onset 55–74 yearsAge at symptom onset ≥75 yearsDifference in HAQ score, <55 years vs. 55–74 yearsDifference in HAQ score, <55 years vs. ≥75 yearsDifference in HAQ score, 55–74 years vs. ≥75 years
No.Mean HAQ score (95% CI)No.Mean HAQ score (95% CI)No.Mean HAQ score (95% CI)Unadjusted (95% CI)Adjusted (95% CI)Unadjusted (95% CI)Adjusted (95% CI)Unadjusted (95% CI)Adjusted (95% CI)
  • *

    <Age 55 years = early onset, age 55–74 years = late onset, age ≥75 years = very late onset. HAQ = Health Assessment Questionnaire; 95% CI = 95% confidence interval.

  • Adjusted for age at HAQ completion and Norfolk Arthritis Register registration year.

  • Significant difference compared to patients ages <55 years at symptom onset.

  • §

    Significant difference compared to patients ages 55–74 years at symptom onset.

05220.68 (0.61, 0.74)5810.80 (0.74, 0.87)1611.07 (0.95, 1.19)0.13 (0.04, 0.22)−0.14 (−0.28, 0.003)0.40 (0.26, 0.53)−0.04 (−0.27, 0.19)0.27 (0.13, 0.40)§0.10 (−0.05, 0.25)
14500.60 (0.53, 0.66)5120.71 (0.64, 0.77)1421.02 (0.89, 1.14)0.11 (0.02, 0.20)−0.16 (−0.30, −0.01)0.42 (0.28, 0.56)−0.02 (−0.25, 0.21)0.31 (0.17, 0.45)§0.14 (−0.02, 0.29)
51950.71 (0.62, 0.79)1950.81 (0.73, 0.89)461.39 (1.22, 1.56)0.10 (−0.02, 0.22)−0.16 (−0.33, −0.001)0.68 (0.49, 0.87)0.25 (−0.01, 0.51)0.58 (0.40, 0.77)§0.41 (0.21, 0.61)§
101060.82 (0.71, 0.92)831.01 (0.90, 1.12)121.80 (1.53, 2.08)0.19 (0.04, 0.34)−0.07 (−0.26, 0.11)0.99 (0.69, 1.28)0.55 (0.21, 0.90)0.79 (0.50, 1.09)§0.63 (0.32, 0.93)§
15850.84 (0.73, 0.95)631.30 (1.18, 1.42)33.09 (2.58, 3.60)0.46 (0.29, 0.62)0.19 (−0.01, 0.39)2.24 (1.72, 2.77)1.81 (1.25, 2.36)1.79 (1.26, 2.31)§1.62 (1.09, 2.15)§

The results of the LRE analysis for female patients are summarized in Table 3. Women in all age groups had comparable HAQ scores at baseline. Women with late onset had significantly higher HAQ scores at years 10 and 15 than women with early onset. Women with very late onset had higher HAQ scores than women with early onset from year 1 onward, and this difference generally increased over time, becoming significant by year 10. Compared to women with late onset, those with very late onset had higher HAQ scores at each followup anniversary and generally showed a faster rate of disability progression, although the difference was only significant at the first and tenth anniversaries, which may reflect small numbers of patients with very late onset.

Table 3. Estimated HAQ score by age at symptom onset in female patients*
Year of followupAge at symptom onset <55 yearsAge at symptom onset 55–74 yearsAge at symptom onset ≥75 yearsDifference in HAQ score, <55 years vs. 55–74 yearsDifference in HAQ score, <55 years vs. ≥75 yearsDifference in HAQ score, 55–74 years vs. ≥75 years
No.Mean HAQ score (95% CI)No.Mean HAQ score (95% CI)No.Mean HAQ score (95% CI)Unadjusted (95% CI)Adjusted (95% CI)Unadjusted (95% CI)Adjusted (95% CI)Unadjusted (95% CI)Adjusted (95% CI)
  • *

    Age <55 years = early onset, age 55–74 years = late onset, age ≥75 years = very late onset. HAQ = Health Assessment Questionnaire; 95% CI = 95% confidence interval.

  • Adjusted for age at HAQ completion and Norfolk Arthritis Register registration year.

  • Significant difference compared to patients ages <55 years at symptom onset.

  • §

    Significant difference compared to patients ages 55–74 years at symptom onset.

01,2880.87 (0.83, 0.92)9031.14 (1.09, 1.19)2111.43 (1.33, 1.54)0.26 (0.20, 0.33)−0.07 (−0.17, 0.03)0.56 (0.45, 0.67)0.01 (−0.16, 0.18)0.29 (0.18, 0.41)§0.08 (−0.05, 0.21)
11,1250.79 (0.75, 0.84)8011.03 (0.98, 1.08)1811.43 (1.32, 1.54)0.24 (0.17, 0.30)−0.10 (−0.20, 0.004)0.64 (0.52, 0.75)0.09 (−0.09, 0.26)0.40 (0.28, 0.52)§0.18 (0.05, 0.31)§
55100.90 (0.84, 0.95)3781.29 (1.23, 1.35)571.61 (1.46, 1.76)0.39 (0.31, 0.47)0.06 (−0.05, 0.17)0.71 (0.55, 0.87)0.16 (−0.04, 0.36)0.32 (0.16, 0.48)§0.10 (−0.07, 0.27)
102850.98 (0.92, 1.04)1681.50 (1.42, 1.58)132.01 (1.74, 2.27)0.52 (0.42, 0.62)0.19 (0.06, 0.32)1.03 (0.75, 1.30)0.47 (0.17, 0.78)0.50 (0.23, 0.78)§0.28 (0.002, 0.57)§
152231.08 (1.01, 1.15)961.67 (1.57, 1.77)42.14 (1.70, 2.58)0.59 (0.48, 0.71)0.26 (0.12, 0.40)1.06 (0.61, 1.51)0.51 (0.05, 0.98)0.47 (0.01, 0.92)§0.25 (−0.21, 0.71)

Estimated HAQ scores, adjusted for treatment and number of comorbidities, are shown for male and female patients in Figure 1. The additional adjustment made little impact, for either sex, upon the relationship between age at symptom onset and HAQ score.

thumbnail image

Figure 1. Difference in Health Assessment Questionnaire (HAQ) score by age at symptom onset compared to patients with early onset (age <55 years), adjusted for age at HAQ completion and Norfolk Arthritis Register registration year, treatment, and number of comorbidities.

Download figure to PowerPoint

Age and sex and HAQ score.

The left panel of Figure 2 shows that the HAQ scores of all patients generally fluctuated over the first 5 years of followup, before worsening thereafter. In patients with early onset, the sex difference was least pronounced (women versus men: 0.20; 95% CI 0.13, 0.27) and both trajectories shared the same fairly flat shape; this age group had higher HAQ scores than most others at baseline but remained at the same level of disability over time. The greatest sex difference was in the late-onset group (women versus men: 0.38; 95% CI 0.30, 0.45), although the shape of the trajectories over time was the same. Among patients with very late onset, again there was a notable sex difference in HAQ scores (women versus men: 0.33; 95% CI 0.19, 0.47) and a similar shape of trajectory. The women with very late onset had the highest HAQ scores of all throughout followup, whereas the men with very late onset had lower baseline HAQ scores than men with early onset. However, by 10 years, men with very late onset had higher HAQ scores than men with early onset (difference 0.42; 95% CI 0.18, 0.66).

thumbnail image

Figure 2. Mean Health Assessment Questionnaire (HAQ) scores by age at symptom onset and sex, adjusted by age at HAQ completion and Norfolk Arthritis Register registration year. ACR = American College of Rheumatology; RA = rheumatoid arthritis.

Download figure to PowerPoint

Looking just at women, all 3 groups had similar baseline HAQ scores (early onset: 1.08 [95% CI 1.02, 1.13], late onset: 1.03 [95% CI 0.98, 1.08], very late onset: 1.09 [95% CI 0.97, 1.21]); although those with very late onset showed the fastest rate of increase in disability, the late-onset group had the greatest postbaseline improvement. Those with early onset and late onset had similar HAQ scores for the first approximately 7 years of followup. Thereafter, women with early onset had the lowest HAQ scores (e.g., year 10; early onset: 1.05 [95% CI 0.99, 1.11], late onset: 1.20 [95% CI 1.11, 1.28], very late onset: 1.59 [95% CI 1.37, 1.82]).

Compared to the other men, those with early onset had the highest baseline HAQ scores (early onset: 0.87 [95% CI 0.80, 0.94], late onset: 0.65 [95% CI 0.59, 0.72], very late onset: 0.75 [95% CI 0.62, 0.88]), although men with very late onset soon began to deteriorate faster, with a sharp increase in functional disability over time. Men with late onset had the lowest HAQ scores of all for at least the first 10 years of followup (e.g., year 5; early onset: 0.82 [95% CI 0.75, 0.90], late onset: 0.68 [95% CI 0.61, 0.76], very late onset: 0.92 [95% CI 0.76, 1.08]), but had a faster rate of increasing disability than men with early onset.

The right panel of Figure 2 shows the HAQ score trajectories by age at onset and sex, confined to the subgroup of patients who met the 1987 ACR criteria for RA at any time during followup. At baseline, this included 1,902 patients (595 male, 1,307 female), at year 5 there were 847 patients (242 male, 605 female), and at year 10 there were 396 patients (104 male, 292 female) in this subgroup. Although we show the mean estimated HAQ scores up to year 15, it should be noted that there were only 2 very late–onset men and 2 very late–onset women in this subgroup who had reached their year 15 followup (i.e., were age >90 years at this point). As can be seen from Figure 2, there was little difference between the HAQ score trajectories of the patients in this subgroup and the entire IP cohort. Overall, these patients had higher HAQ scores than the entire cohort and there was a marginally smaller difference between the HAQ scores of late-onset and very late–onset men throughout followup.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

In this study we investigated the impact of age and sex on the trajectory of functional disability in a cohort of patients with IP. At baseline, women were more likely to meet the ACR criteria for RA and had higher DAS28 and HAQ scores and more swollen and tender joints, but were less likely to be positive for RF and had lower concentrations of CRP than men. We also found that overall, female patients had a worse outcome than male patients, evident even shortly after symptom onset. Increasing age at symptom onset was associated with a faster rate of increasing disability, which did not vary with sex. These relationships were maintained in the subgroup of patients who met the 1987 ACR criteria for RA and independent of treatment and comorbidity measures, although there may have been residual confounding, as the adjustment was relatively crude. In particular, we were unable to adjust for the severity of comorbid conditions.

Our findings add further support to previous studies that reported a more favorable outcome for male RA patients compared to female RA patients (2–11). We have also highlighted the importance of examining such questions using longitudinally collected data, and with the aid of novel analytical techniques. Further, we have shown that the poor prognosis previously reported for patients with late-onset RA may be confined to those with a very late onset (age ≥75 years) (3, 5). Some of the conflict between previous studies may be related to methodologic differences, such as the use of different definitions of “early” and “late” symptom onset (3) or inclusion of patients with late and very late onset within the same group. The definition of early and late onset used in previous studies ranged from 45 years to 65 years; our distinction at 55 years is the midpoint of this range.

Our study has a number of strengths: it was a large inception cohort of patients with recent-onset IP, who were recruited largely from primary care and followed prospectively, having their functional disability assessed using the standardized HAQ for up to 15 years. The majority of the differences in HAQ score reported here, either by sex or age at symptom onset, are greater than the widely accepted minimum clinically important difference of 0.24 (24), thus highlighting the clinical significance of our findings. This is the first time that this technique of longitudinal analysis has been used to identify relative differences in disease trajectory by age and sex, which we have done by adjusting for age at outcome measurement. Other strengths of LRE modeling include adjustment for correlations between repeated measures and the use of all outcome data collected, regardless of the total number of measurements collected per patient.

Our findings should be interpreted alongside, rather than in direct comparison to, those from studies that have not adjusted for age at outcome measurement. Rather than a study of age and sex as predictors of outcome, we have investigated whether sex and age at symptom onset are associated with different IP phenotypes over time. Adjustment for age at HAQ completion enabled us to separate the relationship between disability and aging in general from the relationship between disability and age at IP onset. The study design and findings shown here also highlight the importance of adjusting for sex when studying functional disability.

Approximately 75% of NOAR patients cumulatively satisfy the 1987 ACR criteria for RA by 5 years from symptom onset (7). The results of subgroup analysis, including only patients who met the 1987 ACR criteria for RA during followup, were very similar to our findings for the entire IP cohort. Our cohort therefore represents the entire spectrum of disease severity.

A weakness of this study is that, as might be expected, relatively few patients with very late symptom onset had more than 5 years of followup. One area of compromise in our methodology is that the estimates produced by the longitudinal model (LRE) that we used here relate to mean HAQ score. The drawback of this is that, as a non–normally distributed variable, the median HAQ score is more representative of the cohort average than the mean. However, it is also widely accepted that repeated measures of HAQ score are highly correlated, something for which the LRE model does adjust.

A number of theories of sex bias in outcome measures have been proposed to explain the reported sex differences in DAS28 and HAQ score. For example, it is possible that men and women perceive pain differently (10), that male patients may be better at compensating for disability than female patients (9), or that men may underreport their disability. The sex differences we found did not change with time, suggesting that varying “time to compensation for disability” is not an explanation for the findings. It has been suggested that the HAQ is subjective and thus open to sex bias in patients' approaches to questionnaire completion (25). In addition, we observed that women had less inflammation on average, as measured by the concentration of CRP, and were less likely to be positive for RF than men. However, the same factors have been shown to predict HAQ score and radiographic joint damage in both sexes (3). Another limitation of our study design was that data regarding radiographic damage were not available; therefore, we were unable to explore whether the difference in HAQ score by sex was related to different rates of radiographically measured joint damage.

Sex differences in treatment received may explain our findings; the men in our cohort were more likely to have received DMARD treatment within 6 months of their symptom onset than the women in our cohort. It is possible that men seek medical care (i.e., the route of referral into the NOAR) sooner after symptom onset than women, as suggested by our observation of shorter symptom duration at baseline for men. However, it has been reported that male and female RA patients experience approximately the same delay before initially having their symptoms assessed in primary care (26). Alternatively, male patients may relatively underestimate/underreport the duration of their symptoms. However, adjustment for DMARD treatment within 6 months of symptom onset had no impact on the difference in HAQ score that we found between men and women.

In summary, our findings suggest that the impact of sex on disease outcome is evident at baseline, whereas the impact of age at symptom onset becomes apparent during long-term followup. From shortly after symptom onset, women with IP have higher HAQ scores than men with IP, and this relationship does not vary significantly over time. The key relationship between age at symptom onset and functional outcome relates to the rate of disease progression; higher age at symptom onset is associated with a faster rate of increasing disability. Patients with onset after the age of 75 years have a particularly poor prognosis. Some of the sharp increase in disability in this age group may be related to comorbidity, although crude adjustment for number of comorbidities made little impact on our findings. Nevertheless, this would suggest that there is a special urgency to get disease activity under control in this age group as they have less reserve to cope with added disability. Among the men in our cohort, there was a larger proportion with very late onset than among the women in our cohort. In the general population, there is a larger number of women over the age of 75 years than men; therefore, our cohort suggests that the relative incidence of IP within this age group is higher among men than women. Our earlier estimates of the incidence of RA, by sex, in this age group confirm this (27).

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Symmons had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Camacho, Symmons.

Acquisition of data. Bunn.

Analysis and interpretation of data. Camacho, Verstappen, Lunt, Symmons.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

The support of clinical staff at the Norfolk and Norwich University Hospital and the local primary care physicians is gratefully acknowledged. Data management by the team in Manchester is also much appreciated.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES
  • 1
    Symmons DP, Barrett EM, Bankhead CR, Scott DG, Silman AJ. The incidence of rheumatoid arthritis in the United Kingdom: results from the Norfolk Arthritis Register. Br J Rheumatol 1994; 33: 7359.
  • 2
    Pease CT, Bhakta BB, Devlin J, Emery P. Does the age of onset of rheumatoid arthritis influence phenotype? A prospective study of outcome and prognostic factors. Rheumatology (Oxford) 1999; 38: 22834.
  • 3
    Kuiper S, van Gestel AM, Swinkels HL, de Boo TM, Da Silva JA, van Riel PL. Influence of sex, age, and menopausal state on the course of early rheumatoid arthritis. J Rheumatol 2001; 28: 180916.
  • 4
    Hallert E, Thyberg I, Hass U, Skargren E, Skogh T. Comparison between women and men with recent onset rheumatoid arthritis of disease activity and functional ability over two years (the TIRA project). Ann Rheum Dis 2003; 62: 66770.
  • 5
    Tengstrand B, Ahlmen M, Hafstrom I. The influence of sex on rheumatoid arthritis: a prospective study of onset and outcome after 2 years. J Rheumatol 2004; 31: 21422.
  • 6
    Symmons D, Harrison B. Early inflammatory polyarthritis: results from the Norfolk Arthritis Register with a review of the literature. I. Risk factors for the development of inflammatory polyarthritis and rheumatoid arthritis. Rheumatology (Oxford) 2000; 39: 83543.
  • 7
    Symmons DP, Silman AJ. The Norfolk Arthritis Register (NOAR). Clin Exp Rheumatol 2003; 21: S949.
  • 8
    Iikuni N, Sato E, Hoshi M, Inoue E, Taniguchi A, Hara M, et al. The influence of sex on patients with rheumatoid arthritis in a large observational cohort. J Rheumatol 2009; 36: 50811.
  • 9
    Sokka T, Toloza S, Cutolo M, Kautiainen H, Makinen H, Gogus F, et al. Women, men, and rheumatoid arthritis: analyses of disease activity, disease characteristics, and treatments in the QUEST-RA Study. Arthritis Res Ther 2009; 11: R7.
  • 10
    Ahlmen M, Svensson B, Albertsson K, Forslind K, Hafstrom I. Influence of gender on assessments of disease activity and function in early rheumatoid arthritis in relation to radiographic joint damage. Ann Rheum Dis 2010; 69: 2303.
  • 11
    Forslind K, Hafstrom I, Ahlmen M, Svensson B. Sex: a major predictor of remission in early rheumatoid arthritis? Ann Rheum Dis 2007; 66: 4652.
  • 12
    Krishnan E, Sokka T, Hakkinen A, Hubert H, Hannonen P. Normative values for the Health Assessment Questionnaire disability index: benchmarking disability in the general population. Arthritis Rheum 2004; 50: 95360.
  • 13
    Oka M, Kytila J. Rheumatoid arthritis with the onset in old age. Acta Rheumatol Scand 1957; 3: 24958.
  • 14
    Adler E. Rheumatoid arthritis in old age. Isr J Med Sci 1966; 2: 60713.
  • 15
    Ehrlich GE, Katz WA, Cohen SH. Rheumatoid arthritis in the aged. Geriatrics 1970; 25: 10313.
  • 16
    Corrigan AB, Robinson RG, Terenty TR, Dick-Smith JB, Walters D. Benign rheumatoid arthritis of aged. Br Med J 1974; 1: 4446.
  • 17
    Calvo-Alen J, Corrales A, Sanchez-Andrada S, Fernandez-Echevarria MA, Pena JL, Rodriguez-Valverde V. Outcome of late-onset rheumatoid arthritis. Clin Rheumatol 2005; 24: 4859.
  • 18
    Radovits BJ, Fransen J, van Riel PL, Laan RF. Influence of age and gender on the 28-joint Disease Activity Score (DAS28) in rheumatoid arthritis. Ann Rheum Dis 2008; 67: 112731.
  • 19
    Wiles N, Dunn G, Barrett E, Silman A, Symmons D. Associations between demographic and disease-related variables and disability over the first five years of inflammatory polyarthritis: a longitudinal analysis using generalized estimating equations. J Clin Epidemiol 2000; 53: 98896.
  • 20
    Bukhari M, Lunt M, Barton A, Bunn D, Silman A, Symmons D. Increasing age at symptom onset is associated with worse radiological damage at presentation in patients with early inflammatory polyarthritis. Ann Rheum Dis 2007; 66: 38993.
  • 21
    Prevoo ML, van 't Hof MA, Kuper HH, van Leeuwen MA, van de Putte LB, van Riel PL. Modified disease activity scores that include twenty-eight–joint counts: development and validation in a prospective longitudinal study of patients with rheumatoid arthritis. Arthritis Rheum 1995; 38: 448.
  • 22
    Kirwan JR, Reeback JS. Stanford Health Assessment Questionnaire modified to assess disability in British patients with rheumatoid arthritis. Br J Rheumatol 1986; 25: 2069.
  • 23
    Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988; 31: 31524.
  • 24
    Goldsmith CH, Boers M, Bombardier C, Tugwell P. Criteria for clinically important changes in outcomes: development, scoring and evaluation of rheumatoid arthritis patient and trial profiles. J Rheumatol 1993; 20: 5615.
  • 25
    Van Vollenhoven RF. Sex differences in rheumatoid arthritis: more than meets the eye…; BMC Med 2009; 7: 12.
  • 26
    Kumar K, Daley E, Carruthers DM, Situnayake D, Gordon C, Grindulis K, et al. Delay in presentation to primary care physicians is the main reason why patients with rheumatoid arthritis are seen late by rheumatologists. Rheumatology (Oxford) 2007; 46: 143840.
  • 27
    Wiles N, Symmons DP, Harrison B, Barrett E, Barrett JH, Scott DG, et al. Estimating the incidence of rheumatoid arthritis: trying to hit a moving target? Arthritis Rheum 1999; 42: 133946.