Introduction
Compared to the general population, patients with rheumatoid arthritis (RA) have almost twice the risk of developing cardiovascular disease (CVD)[
1] and their risk for myocardial infarction corresponds to that of non-RA patients who are 10 years older[
2]. CVD mortality accounts for almost half of all-cause mortality in RA[
3,
4]. This is thought to be mediated by accelerated atherosclerosis[
5,
6] and subclinical vasculitis[
7,
8]. The increased burden of established CVD risk factors in RA patients[
9] explains only partly the excess CVD mortality[
10], suggesting that systemic inflammation participates in the development of CVD.
Arterial hypertension is a major modifiable CVD risk factor worldwide[
11]. Previous studies suggest that the prevalence of hypertension is increased in RA[
12,
13], possibly related to clinical status[
14] and the related physical inactivity[
15] as well as to genetic factors[
16,
17]. As recently reviewed, however, the evidence is still conflicting[
18]. This is in part due to the fact that the existing data are based on single-visit office blood pressure (BP) readings, whereas data based on out-of-office BP evaluation (which provides the gold-standard assessment of BP-associated CVD risk[
19]) are lacking. As a result, data on white coat hypertension (WCH) and masked hypertension (MH) phenomena, two BP phenotypes associated with potentially intermediate CVD risk[
19], are also lacking in RA patients.
Elevated BP in RA may derive from increased oxidative stress and systemic inflammation, impaired endothelial function, vasoconstriction and increased total peripheral resistance, as well as arterial stiffening[
5,
9,
12,
13]. RA-related drugs, such as corticosteroids and nonsteroidal anti-inflammatory drugs, and particular genetic polymorphisms, together with environmental factors, may also precipitate BP elevation in RA[
12,
13,
16,
17]. In contrast, anti–tumor necrosis factor and anti–interleukin 6 biologics appear to have beneficial effects on endothelial function and arterial stiffness[
20,
21] and thus may reduce BP independently of Disease Activity Score[
14].
In a contemporary cohort of consecutive RA patients free of established CVD, we evaluated BP and assessed vascular organ damage. Our principal hypotheses were (1) that the prevalence of hypertension in RA is higher than that in the general population and (2) that specific BP diagnostic phenotypes, such as WCH and MH phenomena, have different clinical significance in terms of vascular damage in RA patients. The following were our specific aims: (1) to compare the prevalence of abnormal office BP measurements to those observed in a 1:1 age- and gender-matched general population group free of CVD, (2) to assess the actual prevalence of hypertension in RA by applying, for the first time, out-of-office BP measurements to identify the awareness and effectiveness of BP control as well as the prevalence of resistant hypertension, (3) to identify factors associated with the presence of hypertension in RA and (4) to identify specific BP diagnostic phenotypes, such as WCH and MH phenomena, and their association with the extent of vascular damage.
Discussion
This is the first detailed study of the prevalence and characteristics of hypertension in RA patients without established CVD. It included age- and sex-matched non-CVD controls from the same geographical area and race; it utilized assessment of both office and out-of-office BP, thus enabling estimation of the actual prevalence of hypertension in RA, as well as of the WCH and MH phenomena, which are important in terms of awareness and inadequate therapy; and it incorporated assessment of vascular damage, allowing the exploration of its associations with specific hypertension phenotypes.
The novel important findings of our present study, based on out-of-office BP, are as follows. (1) The prevalence of hypertension in RA was clearly elevated (54%) compared to that in the general population (34%), and it would have been even higher if out-of-office BP assessment were available in the ATTICA study. (2) About 10% of the overall RA cohort (or one of every four RA patients with confirmed hypertension) were not aware that they had hypertension. (3) One of every two patients in the overall RA population (or one in every three RA patients with confirmed hypertension) had uncontrolled BP. (4) The prevalence of the MH phenomenon was negligible. (5) The WCH phenomenon was observed in almost one of every five RA patients overall, which should be regarded as an intermediate to high CVD risk BP phenotype in patients with RA because of the presence of accelerated atherosclerosis, thus raising questions about the optimal BP treatment strategy in this population. (6) We verified that particular RA treatment modalities, but not inflammation per se, are significantly associated with the presence of hypertension and interfere with BP control in addition to classical factors such as age and body mass index.
The observed high prevalence of hypertension (67%) on the basis of office BP is in accordance with previously reported data from studies in which similar office BP measurements were taken (70% in a population with a mean age 63 years[
12] and 57% in a population with a mean age of 59 years[
13]). Most importantly, even the actual prevalence of hypertension in RA on the basis of out-of-office BP assessment was 1.6 times that in the ATTICA study. The data derived from the ATTICA study were based on single-visit BP recordings (at the participants’ workplaces or homes) and are in general agreement with those reported in other studies in the Greek population, which were based on office BP readings[
31‐
35]. It is thus expected that the prevalence of hypertension in the ATTICA study would be lower if based on out-of-office BP methods (either HBPM or ABPM).
A major aim of the present study was to compare the prevalence of hypertension in a typical contemporary cohort of RA patients to the prevalence of hypertension in the general population. Therefore, among all the relevant previously published Greek studies, we decided to use the most contemporary one in the general population and the one which was carried out in exactly the same region as the RA cohort (Attica, Athens). The ATTICA study was performed almost 10 years before the time the data from the present RA cohort were gathered and thus may not depict the updated prevalence of hypertension in the current general population, but the results are in line with all other epidemiological studies published before or since[
31‐
35]. The fact that the precise method of BP measurement differed between the RA and non-RA cohorts may have influenced the results. However, we believe that the magnitude of the difference (twofold) observed in the prevalence of office hypertension between the RA and general populations in this study is extremely unlikely to be due to these limitations. As such, we suggest that the evidence for a higher prevalence of hypertension in RA compared to the general population can be considered conclusive.
The reason for the increased prevalence of hypertension in RA is not clear. In the present study, and in agreement with two previous studies[
12,
13], a direct association between the prevalence of hypertension and inflammation (as assessed by C-reactive protein) was not identified. However, we cannot exclude that systemic inflammation plays a role in the development of hypertension in RA, because (1) RA inflammation fluctuates, and this cross-sectional marker may not capture the long-term cumulative inflammatory burden, and (2) the studied population was well-controlled with conventional and biologic drugs, which affect C-reactive protein, endothelial function, arterial stiffness and BP levels[
14,
20,
21]. Moreover, in the present study, the prevalence of menopause was increased by 15% in RA compared with the matched general female population and was independently associated with office hypertension. Given the fact that early menopause has been associated with increased prevalence of hypertension[
36], as well as with the incidence of RA[
37], these findings may represent a pathogenic link between hypertension and RA. However, biologic drugs exert a beneficial effect on endothelial function and arterial stiffness[
20,
21] and thus may prevent the incidence of hypertension[
14].
The need to improve the awareness of hypertension among hypertensive RA patients is highlighted by the present study, in which the actual percentage of unawareness was 23% based on out-of-office BP measurements. Data based on out-of-office BP measurement are lacking in large general population groups. For purposes of comparison, we report herein the corresponding data from the present RA cohort on the basis of office BP measurements. Using these measurements, 46% of RA patients with hypertension were not aware that they had it, whereas the published data from the ATTICA study in a similar age group (55 to 65 years old) were about 40%[
35]. In a previous study, about 35% of a RA cohort in the same age group (55 to 65 years old) who were unaware of their hypertension[
12].
Moreover, the present data emphasize the need to improve the effectiveness of BP treatment in RA because 29% (on the basis of out-of-office BP) or 35% (on the basis of office BP) of those who were aware of having hypertension had uncontrolled BP. The published data from the ATTICA study in a similar age group (55 to 65 years old) were about 34%[
35]. In a previous RA cohort in the same age group (55 to 65 years old), this figure was about 55%[
12]. Physicians taking care of RA patients should optimize treatment control by taking into consideration the fact that leflunomide, as shown in the present study and as previously described[
38], restricts effective BP control. Although the underlying mechanism is not clear, it is suggested that it might be mediated by the activation of the sympathetic nervous system[
38]. Previous studies[
12,
13] have not provided consistent results regarding other RA-related drugs (for example, corticosteroids, nonsteroidal anti-inflammatory drugs). Similarly, in the present study, we did not identify the well-described effect of corticosteroids[
39] and nonsteroidal anti-inflammatory drugs on BP. This is most probably related to treatment preferences that prevail in each cohort. In the present study, only 5% of patients were being treated with nonsteroidal anti-inflammatory drugs and only low-dose corticosteroids (less than 5 to 7 mg/day) were used. The latter may have undetectable effects on BP level as previously discussed[
12]; however, their effects may be deleterious for the arteries in the long run, and thus their use should be carefully considered[
40]. Notably, in the present study, body mass index was associated with the presence of hypertension as previously observed[
12,
13], as well as with poor BP control. Therefore, the described paradoxical association of obesity with decreased CVD in RA patients[
9] seems to be mediated by other pathways that counterbalance the effect of obesity on BP.
Data regarding the prevalence of resistant hypertension in RA are lacking so far. In this cohort, the prevalence was found to be relatively high (9% in RA vs. 5% in the general population as published in the literature on the basis of office BP readings)[
41]. However, this issue needs further thorough investigation because significant differences exist regarding the prevalence of resistant hypertension from population to population[
41]. Further research is also needed to elucidate whether poor BP control in RA is due to the well-described worldwide physician’s inertia[
42,
43] or whether RA-specific peculiarities, such as treatment modalities, increase the prevalence of resistant hypertension and/or the presence of increased aortic stiffness[
5,
44], also contribute.
The optimal BP diagnostic method and the goals of treatment are unresolved issues in hypertension research[
19]. The results of our present study suggest that there are peculiarities in the RA population because of the negligible prevalence of MH (less than 1% vs. the anticipated 10% to 15% in the general population[
19]), whereas the prevalence of the WCH phenomenon was relatively high (21% vs. the anticipated 15% in the general population[
19]). High prevalence of the WCH phenomenon has also been described in diabetes mellitus type 2[
45], which shares common characteristics with RA[
5], including increased arterial stiffness, which may partly contribute to the WCH phenomenon. The current hypertension guidelines advise clinicians not to treat individuals who exhibit the WCH phenomenon unless there is evidence of target organ damage[
19]. In the present cohort, RA patients with the WCH phenomenon (treated or not treated for hypertension) presented with substantial vascular damage. The prevalence of carotid plaques was 60%, and we observed an intermediate level of carotid hypertrophy, aortic stiffness and reduced ABI compared to those RA patients with sustained normotension and sustained hypertension. These findings are in line with previous findings in non-RA populations[
19]. We suggest that, in similar cases, the decision regarding office BP reduction in RA patients by either drug treatment titration (in those already receiving antihypertensive drugs) or drug treatment initiation (in those currently untreated with drugs) should be carefully weighed on the basis of patient age, the degree of vascular damage, the presence of orthostatic hypotension and, certainly, the out-of-office BP level. This topic requires further research, given the fact that target organ damage has quite a high prevalence in RA and is closely related to hypertension[
46].
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
ADP made substantial contributions to the conception and design of the study, as well as to the acquisition, analysis and interpretation of the data, and was involved in drafting the manuscript and revising it critically. DBP made substantial contributions to the conception and design of the study, acquired data and was involved in revising the manuscript critically. EZ acquired data and was involved in revising the manuscript. KA, AAA and GD Konstantonis acquired data and were involved in critically revising the manuscript. CP made substantial contributions to the conception and design of the study and was involved in critically revising the manuscript. GD Kitas made substantial contributions to the conception and design of the study and to the interpretation of data and was involved in drafting the manuscript and revising it critically. PPS made substantial contributions to the conception and design of the study and to the interpretation of data and was involved in drafting the manuscript and revising it critically. All authors gave their approval of the final version of the manuscript for publication.