Older age of rheumatoid arthritis onset is associated with higher activation status of peripheral blood CD4⁺ T cells and disease activity

Abstract

Rheumatoid arthritis (RA) is a chronic inflammatory disease, with a clinical manifestation both systemic and in joints. It has been suggested that age at disease onset and/or patients' age have influence on disease activity and clinical outcome. The reasons for the different course of RA in older people are not known; however, the activation status of peripheral blood lymphocytes could be responsible. Our aim was to relate expression of activation markers in peripheral blood CD4⁺ T cells of RA patients with patients' age and/or onset age and disease activity measured by DAS28. Seventy RA patients were included into the immunological study. Two separation criteria were performed: based on age of RA onset and on the biological age of patients. We examined different activation markers, CD69, CD25, CD95 and human leucocyte antigen D-related (HLA-DR), on the CD4⁺ T cell surface. Division of RA patients in 10-year intervals at 40, 50 and 60 years revealed that RA patients with later disease onset were characterized by higher DAS28. This phenomenon was not limited to the division at 60 years of age but, surprisingly, the major differences were found for the 40-year onset division. Analysis of all four components of DAS28 revealed that disease activity in older disease onset was dependent on all components. Older-onset RA patients had a higher percentage of CD4⁺CD25⁺ and CD4⁺CD95⁺ T cells. Summarizing the major differences in DAS28 and activation status of CD4⁺ T cells observed for onset of disease at 40 years seems to be the most informative about the immunological status of RA patients.

Introduction

Rheumatoid arthritis (RA) is a chronic inflammatory disease with the dominant clinical component of joint involvement, reflected by the fact that three of the four elements of disease activity score in 28 joints (DAS28) come from the number of affected joints and pain associated with local inflammation. However, RA is not only a local disease, and systemic inflammation assessed by erythrocytes sedimentation rate (ESR) and/or C-reactive protein (CRP) measurement is included into the DAS28 calculation. During the last few years it was found that RA is characterized by profound changes in the immune system, both local (synovitis) and systemic (changed phenotypes and functions of peripheral blood CD4 and CD8 T lymphocytes) [1,2]. The direction of changes in the immune system as reported in the literature seems to be somewhat confusing, as both more activation of immune system and immunosenescence of CD4⁺ T cells have been reported in RA patients [3]. However, immunosenescence in healthy people is also associated with increased systemic inflammation and the appearance of more activated T cells in the blood [4]. Therefore, in our opinion there are two avenues for analysis of immune changes in RA: the age of disease onset and age of the patients at the moment of study.

Both have been suggested recently to influence the disease activity and clinical outcome, although there are few data from this field [5]. Usually patients are divided into either late-onset RA (LORA) including subjects aged more than 60, years and young-onset RA (YORA), including patients aged less than 60 years [6]. LORA is characterized by equal gender distribution, more frequent acute onset, more frequent involvement of large joints and lower frequency of positive results for RF (rheumatoid factor) compared to YORA. The causes of different courses of RA in elderly people are not yet known because, for example, the frequency of anti-citrullinated peptide antibodies (anti-CCP), erosion score or serum CRP levels do not differ between LORA and YORA patients [7].

In spite of still-prevalent opinion, RA can no longer be considered a disease of elderly people, as reviewed in [8]; currently, the majority of RA patients are middle-aged – importantly, earlier diagnosis is the reason for this fact. However, the questions remain as to how to predict the aggressiveness of the disease course and how changes in the cellular immune system are associated with disease activity in patients of younger and older age. There is no information in the literature on how younger and middle-aged RA patients differ in their disease activity and which parameter is more important for the disease activity: age of patients or age of disease onset. We decided to look for cellular aspects of immunity differentiating RA patients with regard to age of onset, individual age and disease activity.

The aim of this study was to determine if the activation status of peripheral blood T cells obtained from RA patients is different depending on patients' age and/or age of RA onset and disease activity.

In this study we analysed different types of CD4⁺ T cell activation markers: CD69, early activation marker; CD25 and CD95, middle activation markers; and human leucocyte antigen D-related (HLA-DR), late activation marker. The alpha-chain of the interleukin (IL)-2 receptor (CD25) is present both on activated and regulatory CD4⁺ T cells [9]. Our previous findings also showed that the CD4^lowCD25^high subpopulation expresses high levels of forkhead box P3 (FoxP3) transcription factor, and functions as regulatory CD4⁺ T cells in humans [10]. Therefore, we included both activated and regulatory CD4⁺CD25⁺ T cells into our analysis.

Material/methods

Subjects

All 70 RA patients (62 women, eight men) included into the immunological studies were consulted in the Outpatient Rheumatology Clinic of Medical University in Gdańsk, Poland. The patients were informed about the goal of the study and gave written consent, and the Bioethical Committee of Medical University of Gdańsk approved the study. Patients' diagnosis was performed according to the 1987 American College of Rheumatology (ACR) criteria. Disease activity measured by DAS28 was performed on the same day as the immunological studies. Distribution of age of RA patients (patients' age, PA) and age of disease onset (ADO) revealed that mean ADO was 42 ± 14·6 years and mean PA of these patients was 44 ± 16·3 years. There were no differences in either mean ADO or mean PA between men and women (36 ± 10 versus 40 ± 15; 39 ± 12 versus 44 ± 17 years, respectively). The mean disease duration was 5·25 ± 5·59 years (range from 3 months to 22 years). Based on the age and disease onset distribution in RA patients we decided to divide our subjects into groups differing by 10 years, to determine if any differences both in disease activity or immunological status could be found and possibly correlated between such cohorts.

Thus, all patients were divided into three groups according to their mean PA:

1. >40 (n = 43) and ≤40 years (n = 27);

2. >50 (n = 35) and ≤50 years (n = 35); and

3. >60 (n = 11) and ≤60 years (n = 59);

and into three groups according to their mean ADO:

1. >40 (n = 41) and ≤40 years (n = 29);

2. >50 (n = 19) and ≤50 years (n = 41); and

3. >60 (n = 8) and ≤60 years (n = 62).

All patients were treated only with methotrexate (not exceeding 15 mg/week), or with a combination of methotrexate with glucocorticosteroids (Encorton), 10 mg daily. None of the RA patients included into the study underwent any biological therapy, to exclude any possible influence on immune system function.

Flow cytometry of T cell subpopulations in peripheral blood

Samples of peripheral blood were stained according to a standard procedure, simultaneously with fluorochrome-conjugated monoclonal antibodies: anti-CD3 fluorescein isothiocyanate (FITC) and anti-CD4 R-phycoerythrin-cyanine 5 (RPE-Cy5) (Dako, Glostrup, Denmark), and either anti-CD25 phycoerythrin (PE), anti-HLA-DR PE, anti-CD69 PE or anti-CD95 PE (BD Bioscience, San Jose, CA, USA); appropriate isotype control antibodies were used [11]. Samples of 10 000 cells were acquired and analysed by flow cytometry [fluorescence activated cell sorter (FACScan); BD Bioscience], using WinMDI™ 2·9 software. T lymphocytes were first identified by forward- and side-scatter gating and by CD3 expression. Only CD3-positive lymphocytes were analysed further and expression of activation antigens were analysed on gated CD4⁺ T cells. To distinguish activated from regulatory T cells we used our recently described method [9], according to which CD4^low25^high T cells were classified as regulatory, and CD4⁺CD25⁺ were considered activated CD4⁺ T cells. T lymphocytes were considered CD4^low if they conformed to the definition described in our previous study [12].

Determination of disease activity

DAS28 was calculated for each patient based on number of tender (TEN28) and swollen joints (SW28), ESR level and visual analogue scale, global health (VAS/GH). Radiographic status was estimated by Steinbrocker's stage score.

Statistical analysis

The significant differences in percentage of peripheral blood T cell subpopulations between compared groups was assessed by Student's t-test; when variable distribution was not symmetrical (as confirmed by Shapiro–Wilks' W-test) or the groups of patients differed greatly in size, Mann–Whitney's U-test was applied. The relation between variables was measured by either Pearson's or Spearman's correlation tests. Cluster analysis was applied for non-homogeneous distribution and type of correlation and slope of regression lines were measured in confirmed clusters. Multiple logistic regression analysis was performed with appropriate age subgroups as a dependent variable, which was coded as ‘0’ for younger subgroups and ‘1’ for older subgroups. Data were presented as mean ± standard deviation (s.d.); P < 0·05 was considered significant. The results are presented as odds ratios (OR) and 95% confidence intervals (CI).

Results

Duration of RA does not correlate with activation of CD4⁺ T cells or age of onset

Duration of disease correlated positively with patients' age, but not with onset age. Moreover, RA patients with mean PA more than 40 years and more than 50 years had significantly higher RA duration than the under-40 and under-50 year age groups. The mean duration time of RA did not differ between the subgroups divided according to ADO (Fig. 1). Disease duration was not different between men and women (7·25 ± 6·54 versus 4·98 ± 5·69 years, respectively).

Fig. 1.

Relations between patients' age, onset age and rheumatoid arthritis (RA) duration. Duration of RA correlates positively with patients' age (○), but not with age of disease onset (□).

The relations between the disease duration, patients' age and age of onset (Fig. 1) were analysed by cluster analysis, which allowed for a cut-off at 10 years of disease duration. We investigated whether these two RA patient groups, differing in disease duration (3·17 ± 2·16 versus 17·0 ± 3·60 years, P = 0·001), were characterized by different ADO, mean PA or different expression of activation markers on peripheral blood CD4⁺ T cells. Those two groups of RA patients did not differ in either mean onset age (ADO: 40·27 ± 17·69 versus 35·18 ± 4·45 years, P = 0·58) or mean patient age (PA: 43·44 ± 17·96 versus 52·18 ± 3·51 years, P = 0·21). Also, the percentages of different CD4⁺ T cells subpopulations, given here as mean % ± s.d. CD4⁺CD69⁺ (4·31 ± 4·20 versus 4·72 ± 3·73), CD4⁺CD25⁺ activated (26·88 ± 11·99 versus 27·30 ± 10·17), CD4^lowCD25^high (4·27 ± 1·79 versus 4·19 ± 1·41), CD4⁺CD95⁺ (72·44 ± 14·24 versus 74·32 ± 9·58) and CD4⁺HLA-DR⁺ (5·19 ± 2·73 versus 4·66 ± 2·27) did not differ significantly between those two RA patient groups. Finally, the disease activity as measured by DAS28 also did not differ between those two RA patient groups (4·16 ± 1·04 versus 4·31 ± 0·74, P = 0·58).

Based on these results, we concluded that RA duration had no influence on disease activity or CD4⁺ T cell activation status; therefore, despite heterogeneous distribution of duration of disease, we included all subjects into the next analysis.

The differences in DAS28 value and its component parameters and Steinbrocker stage score between subgroups divided according to patients' age and RA onset

Divisions of RA patients according to ADO revealed higher disease activity for later onset but, surprisingly, the most significant difference was observed for groups divided at ADO = 40 years. When the separation cut-offs were made at either 50 or 60 years, similar, albeit not statistically significant, tendencies were noticed (Table 1). We also found that all DAS28 component parameters are responsible for higher disease activity in patients with ADO aged more than 40 years. With regard to patients' age, statistically significant differences were observed when a cut-off of 50 years was applied regarding all parameters of DAS28 except ESR (Table 1).

Table 1.

The differences in DAS28 value, components of DAS28 and Steinbrocker score between subgroups divided according to patients' age and onset age.

Criterion of division	≤40 years	>40 years	≤50 years	>50 years	≤60 years	>60 years
	DAS28
Patients' age	3·93 ± 1·2	4·36 ± 1·0	3·89 ± 1·1	4·55 ± 0·9*	4·13 ± 1·1	4·50 ± 0·8
Onset age	3·91 ± 1·3	4·81 ± 1·1*	4·22 ± 1·3	4·66 ± 1·2†	4·3 ± 1·3	5·0 ± 0·9†
	TEN28
Patients' age	6 ± 4	9 ± 5*	6 ± 4	10 ± 5*	7 ± 5	9 ± 6
Onset age	8 ± 5	11 ± 7*	9 ± 6	10 ± 6	9 ± 6	11 ± 6
	SW28
Patients' age	4 ± 3	5 ± 3†	4 ± 3	5 ± 3*	4 ± 3	5 ± 4
Onset age	4 ± 3	6 ± 5*	5 ± 4	5 ± 4	5 ± 4	6 ± 3
	VAS
Patients' age	3·9 ± 2·2	5·0 ± 2·0†	3·9 ± 2·1	5·5 ± 1·8*	4·5 ± 2·2	5·4 ± 1·4
Onset age	4·5 ± 2·1	5·6 ± 2·2*	4·8 ± 2·2	5·5 ± 2·2	4·9 ± 2·2	6·1 ± 1·5
	ESR
Patients' age	22·56 ± 19·6	25·40 ± 19·8	20·41 ± 17·6	28·93 ± 21·3	24·02 ± 20·4	28·87 ± 15·4
Onset age	19·96 ± 16·0	37·46 ± 28·6*	25·11 ± 21·4	37·36 ± 29·7 †	28·20 ± 25·1	37·11 ± 21·7
	Steinbrocker staging score
Patients' age	1·7 ± 1·1	2·7 ± 1·1*	2·0 ± 1·2	2·6 ± 1·1†	2·3 ± 1·2	1·8 ± 0·8
Onset age	2·2 ± 1·3	2·4 ± 1·0	2·5 ± 1·2	1·9 ± 0·8*	2·3 ± 1·1	1·7 ± 0·5
	MTX (mg/week)
Patients' age	11·3 ± 2·3	10·6 ± 2·07	11·12 ± 2·3	10·5 ± 1·87	11·14 ± 2·1	10·7 ± 1·95
Onset age	10·8 ± 2·3	11·09 ± 2·1	10·8 ± 2·3	11·25 ± 2·01	10·93 ± 2·3	10·86 ± 2·2
	Prednisone (mg/day)
Patients' age	8·44 ± 3·9	9·61 ± 3·4	8·11 ± 3·8	10·5 ± 3·5	8·76 ± 4·1	9·15 ± 3·8
Onset age	8·61 ± 2·99	9·20 ± 4·19	8·23 ± 2·94	9·90 ± 4·91	9·01 ± 3·84	9·25 ± 3·92

^*P < 0·05.

^†Borderline statistical significance (0·10 > P > 0·05). Results are expressed as mean ± standard deviation (s.d.). DAS28: disease activity score in 28 joints; TEN28: number of tender joints; SW28: number of swollen joints; VAS: visual analogue scale; ESR: erythrocytes sedimentation rate; MTX: methotrexate.

The radiographic staging score was higher for older patients, divided according to PA with cut-offs at 40 and 50, but not 60 years. Onset age seems to have a much weaker influence on Steinbrocker score, and onset later than 50 years seems to have the opposite influence (Table 1).

Because treatment, particularly the dosages of methotrexate and/or prednisone, could be a confounding factor in our analysis by virtue of changing the disease activity, we determined whether our patients subdivided according to ADO or PA had differed with respect to the dosages of methotrexate and/or prednisone. We found that neither methotrexate nor prednisone dosages differed significantly in any of the groups, regardless of which parameter (ADO or PA) was used for division (Table 1).

Division based on disease activity according to DAS28 value revealed that the patients with severe disease (DAS28 greater than 5·1) had a statistically significantly higher age of onset than patients with moderate and low disease (DAS28 value less than or equal to 5·1), but their biological age did not differ (Fig. 2).

Fig. 2.

Age at disease onset but not patients' age significantly differentiates patients with severe disease activity from those with moderate and low disease activity. Rheumatoid arthritis (RA) patients were divided according to DAS28 (disease activity score in 28 joints) value (>5·1 as a severe disease activity; <5·1 as moderate and low disease activity). Results are presented as median (p25–p75) box-and-whisker plots. *P < 0·05.

There was no difference in DAS28 activity (men: 4·31 ± 1·35 versus women: 4·30 ± 1·27), nor any components of DAS28 related to gender.

The differences between percentages of subpopulations in the groups divided according to age at RA onset and patients' age

Divisions of RA patients based on age of disease onset revealed higher activation of CD4⁺ T cells in later-onset groups for every cut-off age. However, major differences were observed when a cut-off of 40 years was applied for ADO. Almost all examined lymphocyte subpopulations: CD4⁺CD25⁺ (activated), CD4⁺CD95⁺ and CD4⁺HLA-DR⁺, were represented in statistically higher percentages in peripheral blood in later-onset patients; the only exemption was CD4⁺CD69⁺ cells. When division was made based on ADO at 50 years, only CD4⁺CD25⁺ (activated) and CD4⁺CD95⁺ percentages were significantly higher in the later-onset group, and for onset at 60 years only the CD4⁺CD95⁺ subpopulation was increased significantly. In that last division, however, a lower percentage of the CD4⁺CD69⁺ subpopulation was detected in later-onset patients (Fig. 3).

Fig. 3.

Activation status of peripheral blood CD4⁺ T cells in rheumatoid arthritis (RA) patients depends on age at disease onset. RA patients were divided into three groups based on age of onset. The comparison was conducted separately for every division; empty figures represent younger-onset RA patients and filled figures represent older-onset RA patients. Results are expressed as mean and standard deviation. Changes of proportions of CD4⁺ T cells with early activation marker CD69 (a), middle activation markers: CD25 (b), CD95 (c) and late activation marker human leucocyte antigen D-related (HLA-DR) (d) are shown. *Statistically significant differences, P < 0·05.

Divisions of RA patients based on patients' age also revealed higher activation of CD4⁺ T cells in older groups. For 40 and 50 years of age separation subpopulations of CD4⁺CD25⁺ (activated) and CD4⁺CD95⁺ were represented in higher percentages in peripheral blood of older RA patients compared to the younger patients; regulatory CD4⁺ T cells also were higher in those groups. Age above 60 years was associated with a higher percentage of CD4⁺CD95⁺ and CD4⁺CD25⁺ (activated) subpopulations and a lower proportion of CD4⁺CD69⁺ T cells. There were no differences in the mean percentage of analysed subpopulations between men and women.

In order to show which changes of lymphocyte subpopulations increase the probability of disease onset at younger (40 years of age) or older age (50 years), we performed logistic regression analysis for CD4⁺CD69⁺, CD4⁺CD25⁺, CD4^lowCD25^high, CD4⁺CD95⁺ and CD4⁺HLA-DR⁺ (Table 2).

Table 2.

Increased proportions of activated CD4⁺ T cells predict later onset of disease and increased age of patients.

Independent parameters	OR	95% CI	P
Risk of onset at 40 years
CD4+CD69+	0·71	0·51–1·00	0·10
CD4+CD25+ activated	1·10	1·00–1·21	0·04
CD4lowCD25high	1·14	0·64–2·03	0·64
CD4+CD95+	1·14	1·03–1·26	0·01
CD4+HLA-DR+	0·81	0·51–1·28	0·05
Risk of onset at 50 years
CD4+CD69+	0·64	0·37–1·11	0·11
CD4+CD25+ activated	1·08	0·97–1·21	0·15
CD4lowCD25high	0·86	0·45–1·62	0·64
CD4+CD95+	1·17	0·49–1·02	0·01
CD4+HLA-DR+	0·71	0·38–1·30	0·26
Risk of onset at 40 years of age
CD4+CD69+	0·86	0·67–1·07	0·16
CD4+CD25+ activated	1·08	0·99–1·19	0·07
CD4lowCD25high	1·48	0·82–2·62	0·18
CD4+CD95+	1·16	1·05–1·30	0·007
CD4+HLA-DR+	0·64	0·42–0·99	0·07
Risk of onset at 50 years of age
CD4+CD69+	0·63	0·37–1·03	0·07
CD4+CD25+ activated	1·16	1·02–1·33	0·02
CD4lowCD25high	1·07	0·60–1·92	0·81
CD4+CD95+	1·22	1·06–1·40	0·005
CD4+HLA-DR+	0·50	0·25–0·96	0·06

Multiple logistic regression analysis was performed with appropriate age subgroups as dependent variables and proportions of subpopulations of activated CD4⁺ T cells as independent variables. OR: odds ratio, 95% CI: confidence interval; HLA-DR: human leucocyte antigen D-related.

Due to differences between numbers of subjects between subgroups divided based on 60 years as a cut-off, we did not perform modelling in that separation.

In conclusion, we show here that the increased proportions of both CD4⁺CD25⁺ and CD4⁺CD95⁺ cells predict (correspond to higher relative risk of) later onset and characterize older patients (above 40 years).

Discussion

Two forms of RA were defined by age at disease onset (ADO): late-onset RA (LORA), defined as disease beginning at more than 60 or 65 years, and young-onset RA (YORA), beginning in middle age, were described [13]. However, in the literature there is little agreement regarding clinical manifestation, disease activity and laboratory findings between early- and late-onset RA. Higher ESR and CRP values are observed commonly in LORA patients [6,14], while all other analysed parameters give contradictory results, including number of tender and swollen joints and Steinbrocker's erosive scale [6,15]. Relatively higher DAS28 in the group aged more than 65 years was dependent only on ESR or CRP [16], but not on other DAS28 components. In this study we divided patients using 40, 50 and 60 years of age as cut-offs for ADO. For each ADO delimiter, we found major differences in DAS28 activity between those with older and younger ADO, resulting in higher activity in the older group. The increased disease activity for older ADO was dependent on all four DAS28 components used. However, we did not show significant differences in ESR values between patients with ADO after and before 60 years, which contradicts previously published data [6,14]. Our results could be due to the low number of patients with ADO aged more than 60 years in our group, which is a limitation of our study; however, in our experience the numbers reflect a typical age distribution of RA patients in our region [8]. It should be stressed that we did not find any differences in ESR value with regard to patients' age (PA), which may suggest its lack of influence on the degree of inflammation in RA patients.

The Steinbrocker score showed a major difference between patients divided based on ADO at 40 and 50 years. Our results suggest that the former division reflects changes in disease activity, while the erosion process is more dependent on patients' age and related to disease duration. Other authors did not show any correlation of erosive status either with regard to ADO or RA duration, but they used a different scale based on Larsen's grade [17] and 60 years of age as a division criterion, so our results could not be compared directly.

There are no published data about the relation of status of T cells in RA patients and the age of disease onset and patients' age. Conversely, it has been suggested that the T cell system of RA patients undergoes premature ageing [1]. In our study, RA patients with mean age and age onset of less than 60 years had a higher percentage of CD4⁺CD69⁺ subpopulation compared to older (60+) patients, indicating a lower early activation status in older patients. The reason for this observation is not known; however, it was shown that short activation resulted in a reduction of CD69 expression on CD4⁺ cells from healthy subjects aged more than 60 years versus subjects aged less than 35 years [4]. It is not known if the lower percentage of CD4⁺CD69⁺ T cells in older RA patients is due to a lower ability of CD4⁺ T cells to express CD69 upon in vitro activation, but as the CD69 marker is the only activation marker which is lower in older RA patients and only for cut-off at 60 years old, the main reason for different CD69 expression could be the biological patient age. A recent paper indicated that the level of expression of CD69 in murine T cells depends not only on the activity of its gene promoter, but also on the activities of at least four different cis-regulatory elements CSN1-4 [18]. By analogy to our earlier observations of reduced expression of CD28 in the CD4⁺ cells of both healthy elderly and RA patients, related to improper binding of a regulatory element near its promoter [19], we can hypothesize that the expression and activity of these CSN elements may vary between young and old and between healthy and RA individuals. Another possibility for the lower percentage of CD4⁺CD69⁺ T cells in peripheral blood of older RA patients could be due to faster elimination of activated cells from peripheral blood, which seems to be somewhat unlikely because CD4⁺ T cells with other activation markers are increased in those patients.

There are conflicting results concerning proportions of activated and regulatory CD4⁺CD25⁺ T cells in RA patients' peripheral blood, variably reported as increased, not unchanged or decreased [20–22]. Confusion seems to be due mainly to the different definitions used for distinguishing those two T cell subsets, with some authors considering as regulatory all CD4⁺CD25⁺ cells, and others only the CD25^bright lymphocytes [20,21]. Most authors showed that both the percentage and absolute numbers of CD4⁺CD25^high and CD4⁺CD25⁺ increased significantly with age [10,23,24]. Based on our recent paper, we assumed that only CD4^lowCD25^high T cells fulfilled criteria for being regulatory CD4⁺ T cells (T_regs), the remaining CD4⁺CD25⁺ being just activated [10]. Here, analysis of both CD4⁺CD25⁺ T cell subpopulations showed that later disease onset was associated with a higher proportion of activated CD4⁺ T cells for all ADOs, but if only 40 years were used as the ADO delimiter, the percentage of T_regs differed. This suggests an increasing disproportion between activated and regulatory CD4⁺CD25⁺ T cells in later-onset RA patients leading to increased immune reactivity and higher disease activity.

Recent data suggest that CD95 not only induces apoptosis, but also acts also as an activator marker on CD4⁺ T cells [25]. It has been demonstrated that elderly healthy subjects had more peripheral blood CD3⁺CD95⁺ cells than young subjects [26]; after activation, CD95 expression was also significantly higher in all generations of CD4⁺ cells of older people [27]. Our study consistently showed a higher percentage of CD4⁺CD95⁺ cells in older patient groups (regardless of the dividing age or ADO), suggesting that both ADO and PA may affect a comparable proportion of the subpopulation. The proportion of CD4⁺HLA-DR⁺ T cells was dependent both on age of RA onset and patients' age; however, ADO had a stronger influence than patients' age. Regression analysis showed that the decreased CD4⁺HLA-DR⁺ subpopulation was a strong predictor of disease onset even before 40 years of age.

In our opinion, our results support the thesis that major changes in the immune system in RA patients associated with disease activity are already observed in a younger group of patients. Other authors have shown that the number of new CD4⁺ thymic emigrants and the length of telomeric DNA of CD4⁺ T cells decrease in RA patients were observed mainly between 20 and 60 years of age [28]. Therefore, we suggest that the clinically important immunological changes in RA patients should be studied in middle-aged RA patients. In the light of our observations, the cut-off point stratifying the disease activity and T cell activity profile versus age should be placed earlier than at 60 years. We suggest the age of disease onset at 40 years and patients' age at 50 years as the cut-offs distinguishing RA patients most effectively.

Disclosure

All authors have declared that they have no conflicts of interest.