Prevalence, incidence, and mortality of Raynaud’s phenomenon, Sjögren’s syndrome and scleroderma: an umbrella review of systematic reviews

Abstract

Objectives

To comprehensively review systematic reviews of prevalence, incidence, and mortality of Raynaud’s, Sjögren’s and Scleroderma, and to identify any research gaps.

Methods

An umbrella review of English language systematic reviews was undertaken using PubMed and Embase (OVID) covering the period 2000–2023 (PROSPERO CRD42023434865). The estimate and its corresponding 95% confidence interval were reported when available from each systematic review. The quality of systematic reviews was assessed using the Scottish Intercollegiate Guidelines Network (SIGN) tool. A narrative synthesis was undertaken.

Results

Seventeen systematic reviews were identified, of which 1 was for RP, 5 for Sjögren’s and 11 for Scleroderma. There were some high-quality systematic reviews for Sjögren’s and mortality of Scleroderma. However, there were only low-quality systematic reviews of prevalence and incidence of RP and Scleroderma. Furthermore, there were no systematic reviews for the mortality of RP. For RP, the pooled prevalence was 4850 per 100 000; pooled annual incidence was 250 per 100 000. For Sjögren’s, prevalence was 60–70 per 100 000; annual incidence was 6.92 per 100 000 and the pooled standardized mortality ratio ranged from 1.38 to 1.48. For Scleroderma, pooled prevalence ranged from 17.6 to 23 per 100 000; annual incidence was 1.4 per 100 000; and the pooled standardized mortality ratio ranged from 2.72 to 3.53.

Conclusion

The outcomes of RP were less well described compared with Sjögren’s and Scleroderma. There was a lack of high-quality systematic reviews for the prevalence and incidence of RP and Scleroderma. Therefore, further studies and systematic reviews with rigorous case definitions, assessing different ethnic groups are warranted in this area.

Introduction

Raynaud’s Phenomenon, Sjögren’s Syndrome and Scleroderma are diseases of unknown aetiology that share some common features and risk factors [1–3]. Over the past two decades, a number of studies were conducted to investigate the prevalence, incidence and mortality of these conditions. These studies also examined the severity and prognosis of conditions which in turn would help plan healthcare resources in the populations, improve healthcare and mean economic costing. There have also been several systematic reviews on the conditions. However, there have not been good-quality, recent, systematic reviews of all aspects of the epidemiology of these conditions. Where there have been several previous systematic reviews, these contained variations in findings relating to prevalence, incidence and mortality [4, 5]. It is unknown why they varied; whether these differences were due to their methodology and study quality, databases searched or criteria of including primary studies. This has created a confusing landscape. There have been no reviews of reviews that have sought to explain these variations or to identify potential gaps in the literature. Therefore, this umbrella review aims to comprehensively review existing systematic reviews on the prevalence, incidence and mortality of Raynaud’s, Sjögren’s and scleroderma and to identify any gaps in the current literature. This will help to identify priorities for future research.

Methods

This umbrella review was prospectively registered in PROSPERO (CRD42023434865) [6] on 15 June 2023. It was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines and recommendations for umbrella reviews. No revisions were made to the protocol registered in PROSPERO during the study period. Ethical approval was not required because this study retrieved and synthesized data from previously published studies.

Search strategy and selection criteria

A comprehensive systematic literature search was undertaken in PubMed (this includes Medline and PubMed Central databases) and Embase (OVID) from 1 January 2000 to 1 June 2023. We searched using MeSH terms and keywords for ‘RP’, ‘SS’ or ‘Scleroderma’ or ‘SSc’ and terms such as ‘prevalence’, ‘incidence’ or ‘mortality’ (see Supplementary Tables S1–S3, available at Rheumatology Advances in Practice online, for full details of the search strategy). For mortality, we included systematic reviews investigating all-cause or cause-specific mortality. We limited our search to systematic reviews published in English and excluded conference abstracts. We included systematic reviews of patients of any age, sex and geographical area, based in the general population. Occupational cohorts, indigenous populations, critically ill patients and cohorts identified from specialist centres were excluded. We excluded systematic reviews which focused on a specific sub-condition (such as scleroderma renal crisis, scleroderma overlap syndrome or scleroderma-associated digital-ulcers; see Supplementary Table S6, available at Rheumatology Advances in Practice online for the list of excluded studies). We searched PROSPERO for any systematic review protocols on topic [7–10]. We emailed the authors of four protocols [7–10] for further information and received one reply [10]. On screening the full text, it was found to be out of scope. Duplicate systematic reviews were removed manually. Two reviewers (AC and SL) independently screened titles and abstracts for relevance using Rayyan [11], and full-text papers for eligibility using AirTable. Discrepancies were resolved through discussion with a third reviewer (FP or PL) where required.

Data extraction and quality assessment

AC independently extracted the following data from each eligible systematic review: first author; publication year and affiliation; number of primary studies included in the systematic review (databases searched and eligibility criteria); study characteristics; main outcome estimates with assessments of heterogeneity (e.g. I²); whether risk of bias was reported (see Tables 1–3) and SIGN quality assessment (SIGN checklist in Supplementary Table S4, available at Rheumatology Advances in Practice online). This was then checked by SL. For each systematic review, we recorded either a narrative statement summarizing the authors’ main findings or extracted a pooled estimate and its corresponding confidence interval (95% unless stated otherwise). Rates were converted to a per 100 000 denominator where required. The methodological quality of each systematic review was assessed by AC and SL using the Scottish Intercollegiate Guidelines Network (SIGN), a 12-question tool appropriate for systematic reviews of observational studies [12]. Any discrepancies in quality assessment were resolved through a discussion between AC and SL. Systematic reviews were reported to be of high quality if most of the SIGN criteria (i.e. had their own quality assessment) were met and had little to no risk of bias. Systematic reviews that met most of the SIGN criteria with only minor flaws (i.e. did not consider publication bias or state conflicts of interests) were deemed to be of acceptable quality. Systematic reviews were rated low quality if they had major flaws such as a lack of quality assessment of their included studies and/or lack of relevant characteristics in the tables. Any discrepancies were resolved through a discussion between the two reviewers (AC and SL). If an agreement could not be reached, a third reviewer (FP or PL) was contacted.

Results

Literature search and selection process

PubMed and Embase yielded 63 studies for Raynaud’s (Fig. 1), 158 for Sjögren’s (Fig. 2) and 178 for scleroderma (Fig. 3). After title and abstract screening, 6 studies remained for Raynaud’s, 15 for Sjögren’s and 23 for scleroderma. After full-text screening, 1 systematic review was included for Raynaud’s [1], 5 for Sjögren’s [4, 5, 13–15] and 11 for scleroderma [14–24]. The earliest included systematic review was published in 2008, with an increasing frequency over time: one systematic review was published in 2005–2009, four in 2010–2014, seven in 2015–2019 and five in 2020–2023. According to the SIGN quality assessment, six systematic reviews were of high quality, four were of acceptable quality and seven were of low quality.

Raynaud’s Phenomenon (RP)

One systematic review by Garner et al. [1] met the inclusion criteria, reporting the prevalence and incidence but not mortality of Raynaud’s. It included studies of the general population from high-income countries. According to the SIGN quality assessment checklist, it was reported to be of low quality due to a lack of quality assessment of their included studies. The authors included 17 studies of prevalence from 4 of the 6 World Health Organisation (WHO) world regions (the Americas, Southeast Asia, Europe, Western Pacific; but not Africa nor the Eastern Mediterranean region). The pooled estimate of prevalence (pooled prevalence from the five studies of definite Raynaud’s in the general adult population) was 4850 per 100 000 (95% CI 2080, 8710), with the pooled prevalence for females being 5740 per 100 000 (2740–9750) and for males 4120 per 100 000 (95% CI 1600, 7740). This review identified only two studies of incidence, which came from Europe and North America. The pooled annual incidence from two studies of incidence was 250 per 100 000 (95% CI 190, 320). There were no systematic reviews of mortality among people with Raynaud’s.

Sjogren’s Syndrome (SS)

Five systematic reviews were included for SS: one systematic review examined both prevalence and incidence [16], one examined prevalence [17], and three examined mortality [4, 5, 15]. Of the five systematic reviews, three were of high quality [4, 5, 13] and two was acceptable [14, 15]. Two systematic reviews reported the prevalence of Sjögren’s. Qin et al. [13] conducted a systematic literature search of PubMed and Embase and included 18 studies conducted worldwide, published in English, between 1995 and 2013. The pooled prevalence of primary Sjögren’s was 60.82 (95% CI 43.69, 77.94) per 100 000. The authors found the prevalence rate (per 100 000) of Sjögren’s in women was higher than in men and a higher rate Sjögren’s prevalence in Europe compared with Taiwan. Albrecht et al. [14] included 20 studies published from 2014 to 2022, conducted in the German population only and including sicca syndrome as well as SS. The authors [14] found a pooled prevalence of primary Sjögren’s including sicca syndrome of 70 per 100 000 people in Germany.

One systematic review of Sjögren’s incidence, by Qin et al. [13] who included six studies of incidence reported a pooled annual incidence rate of 6.92 (4.98–8.86) per 100 000. The authors found a higher incidence rate of primary Sjögren’s in females (12.30, 9.07–15.53) than males (1.47, 0.81–2.12), and there tended to be a higher incidence rate in Asia (Taiwan) than in other regions (EU and North America).

There were three systematic reviews of mortality in Sjögren’s including two studies [4, 5] of all-cause mortality and one of only cardiovascular mortality [15]. Singh et al. [4] reviewed 10 studies of 7888 primary Sjögren’s patients with 682 deaths over a median follow-up of 9 years, which reported a pooled standardized mortality ratio of 1.38 (0.94–2.01), suggesting a possible association of primary Sjögren’s with increased all-cause mortality. Huang et al. [5] performed a systematic review of 14 studies (14 584 primary Sjögren’s patients and 902 deaths observed) and found a 1.46-fold (95% CI 1.10, 1.93) increased risk of death in primary Sjögren’s. The authors further showed the risk of mortality was highest among Sjögren’s individuals with vasculitis. The results for the one study of cardiovascular mortality were similar to the findings for all-cause mortality. Beltai et al. [15] reviewed 14 studies of 67 124 patients with primary Sjögren’s found that the patients had a raised risk of mortality compared with the general population cardiovascular mortality (RR = 1.48, 0.77–2.85), which did not reach statistical significance.

Scleroderma

Eleven systematic reviews were included for scleroderma: three systematic reviews [16–18] reported both prevalence and incidence, one systematic review [19] reported both prevalence and mortality and seven systematic reviews [18–24] reported mortality only. Of the 11 systematic reviews, four were of high quality [20–23], one was acceptable [24] and six were of low quality [16–19, 25, 26]. The papers reported showed wide ranges of estimates of prevalence with Chifflot et al. [16] reviewed 32 studies reporting prevalence ranging from 0.7 in Japan in 1974–6 to 48.9 in Italy in 1992. Morrisroe et al. [19] reviewing 54 studies from Australia reporting that the prevalence of scleroderma increased over time from 0.46 per 100 000 in 1974 to 20.0 per 100 000 in 1993 to 86.0 per 100 000, most recently. The prevalence was found to be similar between the general and the Aboriginal populations in the country. The two meta-analyses by Zhong [18] and Bairkdar [17] reported similar pooled estimates. Zhong et al. [18] pooled the data of 18 studies of 11 574 individuals and estimated a scleroderma prevalence of 23 per 100 000 (16–29 per 100 000). The prevalence of scleroderma varied across countries, ranging from 3.8 per 100 000 in Taiwan to 50 per 100 000 in the USA. Bairkdar et al. [17] reviewed 46 studies published up to October 2020 and reported a pooled prevalence of 17.6 (15.1–20.5) per 100 000, with a prevalence of 28.0 (23.1–33.9) per 100 000 in women, and 6.0 per 100 000 (4.8–7.5) in men. The highest prevalence of scleroderma was found to be in North America 25.9 per 100 000 (21.5–31.2).

Three systematic reviews [16–18] also examined the incidence rates of scleroderma, with two reporting narrative results and only one meta-analysis. In the narrative results, Chifflot et al. [16] reported incidence rates ranging from 0.06 to 12.2 per 100 000 person-years. Zhong et al. [18] reviewed 12 studies and reported incidence rates of scleroderma ranging from 0.77 per 100 000 person-years in the Netherlands to 5.6 per 100 000 person-years in the USA. Bairkdar et al. [17] performed a meta-analysis of the data from 28 studies published up to October 2020, finding the pooled incidence rate of scleroderma in the population was 1.4 (1.1–1.9) per 100 000 person-years and there was a higher incidence in women (2.3, 1.8–2.9) than in men (0.5, 0.4–0.7).

Eight systematic reviews [19–26] reviewed mortality in people with scleroderma, with six presenting results of a meta-analysis and two were systematic reviews [19, 26]. Of the six meta-analyses, three were rated high quality but the latest paper included was published in 2012 and three later systematic reviews including papers up to 2019 were rated acceptable or low quality. The results of the six meta-analyses were similar (irrespective of quality rating) reporting a pooled SMR for scleroderma of around 3.5, which is substantially higher than for Sjögren’s. Three high-quality studies including papers published in 2013 or earlier: Elhai et al. [21] who reviewed nine studies and pooled the data of scleroderma SMR as 3.53 (95% CI 3.03, 4.11). Komosci et al. [22] reviewed 18 studies and found the risk of death was significantly increased among patients with cardiac involvement HR (3.15, 2.33–4.26), pulmonary interstitial disease (HR 2.58, 1.98–3.37), with pulmonary hypertension (HR 3.50, 1.94–6.30) and with renal manifestations (HR 2.76, 1.91–4.00). Toledano et al. [20] reviewed seven studies and reported a pooled SMR of 3.51 (2.74–4.50).

There were five low or acceptable quality scleroderma systematic reviews. Morrisroe et al. [19], a systematic review of 54 studies, found that despite improvements in Australian healthcare, the mortality remained high with an SMR of 3.4 (no confidence interval was given). Reviewing 18 studies of 11 719 patients with scleroderma, Pokeerbux et al. [25] reported a pooled SMR of 3.45 (3.03–3.94). Reviewing 22 studies of 13 214 patients and 4218 deaths, Lee et al. [24] found an overall SMR of 2.82 (2.22–3.59). Reviewing 91 studies including 1884 patients, Erzer et al. [26] reported a scleroderma mortality of 8.49% of all patients (160 out of 1884 total patients from 1945 to 2018). Rubio-Rivas et al. [23] in their systematic review of 17 studies found that scleroderma patients had an increased risk of mortality (SMR =2.72, 1.93–3.83) compared with the general population.

Discussion

In our umbrella review, the outcomes of Raynaud’s were less well-described compared with the other conditions. We identified that there were no systematic reviews for the mortality of Raynaud’s and no high-quality systematic reviews covering the prevalence and incidence of Raynaud’s and scleroderma (prevalence and incidence) (see Supplementary Fig. S1, available at Rheumatology Advances in Practice online). A common theme among the low-quality systematic reviews (n = 7) was a lack of quality assessment or investigation for publication bias among the included studies. Some of these systematic reviews lacked information on whether two authors selected and extracted data and relevant characteristics in their main finding tables. None of the systematic reviews included a list of excluded studies (see Supplementary Table S5 and Supplementary Data S1, available at Rheumatology Advances in Practice online).

There was little evidence about the epidemiology of Raynaud’s, identifying only one systematic review, which was rated low quality. The pooled prevalence was reported as 4850 per 100 000, and incidence was reported as 250 per 100 000 per year. However, the heterogeneity in the prevalence estimate was extremely high (I² = 98.2%) and the incidence findings were based only on two papers, making it difficult to draw firm conclusions. There were no systematic reviews of mortality in Raynaud’s, probably because of a lack of primary research studies to include.

There was more evidence about Sjögren’s. Our umbrella review identified four high or acceptable quality systematic reviews for prevalence, incidence and mortality of Sjögren’s. Most of these systematic reviews reported their own quality assessment or acknowledged publication bias. Overall, the prevalence was found by two systematic reviews [13, 14] to be between 60 and 70 per 100 000; the incidence was reported by one acceptable quality study to be around 7 per 100 000 per year. The incidence of Sjögren’s was eight times higher among females than males; it was higher in studies from Asia (Taiwan) than Europe or North America. The mortality was found in two systematic reviews [4, 5] to be 1.4 times higher than the general population. Two Sjögren’s systematic reviews [4, 5] reported a variation in the significance of Sjögren’s with all-cause mortality. This variations between these findings could be owing to the Huang et al. [5] later publication date including four additional studies that were not in Singh et al. [4] systematic review. Although the largest number of the systematic reviews were about scleroderma, there were no high-quality systematic reviews of prevalence or incidence. Two low-quality systematic reviews [17, 18] reported a pooled prevalence around 20 per 100 000, and the most recent study of incidence reported it around 1.4 per 100 000 per year. Heterogeneity in these studies was very high (>90%) again making it difficult to draw firm conclusions. There were several high-quality studies of mortality, which reported similar pooled estimates of mortality around 3.5 times higher than the general population, which is substantially (about 2.5 times) higher than in Sjögren’s. In scleroderma, the main causes of deaths were highlighted as having cardiovascular, renal and pulmonary involvement [19, 21–23, 25].

Our umbrella review reported a higher proportion of women than men suffering from each condition [5, 17, 18]. This is expected as previous studies demonstrated that high levels of oestrogen were associated with the high prevalence of Raynaud’s in pre- and post-menopausal women [27, 28]. A systematic review [13] found that studies from Europe showed a higher prevalence of primary Sjögren’s compared with Asian counterparts, but the incidence was higher in Taiwan (Asia) compared with European and American studies. However, note that this difference should be interpreted cautiously due to difference in the number of studies from each region. There was a higher prevalence and incidence of scleroderma [18] in America compared with Taiwan and Netherlands, respectively. These geographical differences could be due to differences in genetic make-up, environment, lifestyle and cultural factors, diagnostic criteria and research methods between countries and healthcare resources [13, 22]. Furthermore, methodological differences such as the number of studies examined and differences in publication period of the original study can be contributing factors to the difference in findings.

A recent study [29] published in May 2023 using linked data from the Clinical Practice Research Datalink (CPRD) investigated the prevalence and incidence of autoimmune diseases in a cohort study of 22 million people in the UK. The authors had found that the incidence of Sjögren’s out of 12 292 individuals changed from 6.0 per 100 000 in 2000–02 to 10.7 per 100 000 in 2017–19; SSc 4880 patients 2.6 per 100 000 in 2000–02 to 3.3 per 100 000 in 2017–19. Therefore, the incidence rate ratio (95% CI) for these two auto-immune diseases were 2.09 (1.84–2.37) and 1.23 (1.06–1.43), respectively. However, this study did not include any information regarding on the prevalence, incidence and mortality of Raynaud’s. This study was not included in any of the systematic reviews because their search dates were before November 2022.

Strengths and limitations

To our knowledge, our umbrella review is the first to examine the prevalence, incidence and mortality of Raynaud’s, Sjögren’s and Scleroderma. We captured various forms of measuring outcomes such as trend in time and differences in sex and geography. The umbrella review provides a comprehensive overview of available evidence of the diseases of interests, and the contradictions and consistencies between the systematic reviews. This will help clinicians and researchers to easily understand the depth and breadth of systematic reviews available on the diseases of interest’s outcomes. We assessed the quality of the systematic reviews in our umbrella review using the SIGN guidelines. This will help to identify areas for improvement.

This study has some limitations. First, we limited our search to 2000–2023. Second, only English-language texts were included. Therefore, there remains a potential we have missed studies that are relevant studies to our research question. Third, there were a few systematic reviews available for Raynaud’s. This indicates further research is needed. Finally, due to the number of systematic reviews rated low quality the implications for clinical practise and healthcare planning are limited.

Conclusion

In our umbrella review, gaps in available high-quality systematic reviews of epidemiology were found in Raynaud’s and scleroderma. Furthermore, there were little to no systematic reviews investigating the prevalence, incidence and mortality of Raynaud’s in comparison to Sjogren’s and scleroderma. Therefore, further studies with rigorous case definitions, assessing different ethnic groups and updated systematic reviews are warranted in this area.

Supplementary material

Supplementary material is available at Rheumatology Advances in Practice online.

Data availability

The data and associated materials used in this study will be made available upon request.

Funding

This work was supported by a PhD studentship to Anthony Chen from a benefactor donation to the University of Nottingham. F.P. is funded by an NIHR Advanced Fellowship (NIHR300863). The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.

Disclosure statement: The authors have declared no conflicts of interest.

References