The global magnitude and temporal trend of rheumatoid arthritis burden attributable to smoking from 1990 to 2019

Abstract

Objectives

The relationship between smoking and RA has been confirmed. Most nations have ratified the Framework Convention on Tobacco Control. However, there are considerable regional differences in how effectively tobacco control measures were implemented. This study was carried out to estimate the spatiotemporal trends of smoking-related RA burdens.

Methods

Data were available from the Global Burden of Disease Study 2019 and were analysed by age, sex, year and region. Joinpoint regression analysis was applied to the analysis of temporal trends in the RA burden resulting from smoking over 30 years.

Results

From 1990 to 2019, the number of global RA cases increased each year. The age-standardized prevalence, death and disability-adjusted life-year (DALY) rates also increased. However, there was a wave in the changing trend of the age-standardized death rate, with the lowest point in 2012 and the highest point in 1990. Smoking, in particular, was responsible for 11.9% of total RA deaths and 12.8% of total DALYs in 1990 but only 8.5% of total RA deaths and 9.6% of total DALYs in 2019. A greater burden from smoking exposure was borne by men, older adults and people living in high-middle and high sociodemographic index (SDI) countries and regions. Moreover, the UK demonstrated the highest reduction in age-standardized death and DALY rates over the three decades.

Conclusion

There were reductions in the age-standardized burdens of RA caused by smoking worldwide. Nevertheless, this continues to be an issue in some areas, and efforts to reduce smoking should be made to lessen this growing burden.

Introduction

There is a general belief that the perceived burden of disease attributed to chronic musculoskeletal conditions is lower relative to other diseases that are more closely associated with mortality, such as cancer, kidney disease and cardiovascular disease [1, 2]. However, progressively emerging evidence [3, 4] has unequivocally and consistently identified the profound prevalence, socioeconomic impact and burden of disease imposed by chronic musculoskeletal conditions and their associated pain burden and functional impairments. Obviously, these effects are relevant in conditions such as osteoarthritis, RA, osteoporosis and pain of musculoskeletal origin.

RA is one of the most prevalent chronic autoimmune diseases and is characterized by the destruction of the lining of the synovial joints, bones, muscles, blood vessels, and related soft tissues or connective tissues, which leads to irreversible loss of joint function and, ultimately, considerable disability or premature death [5, 6]. It was reported that 0.1–2.0% of the global population suffers from RA, which is a recognized model of a multifactorial disease and is associated with genetic and environmental factors, and the prevalence is rising [7]. Moreover, 40–70% of RA risk is due to the influence of nongenetic factors, which has a greater effect than the genetic propensity [8]. Therefore, it is important to study the role of nongenetic factors in RA development.

The relationship between smoking, one of the consistently reported nongenetic risk factors, and RA has been corroborated by multiple studies [9, 10]. However, there has been no study on the global epidemiological pattern of the RA burden attributable to smoking. An analysis of RA prevalence from smoking will help determine the disease and economic burden of care, inform health policy to reduce the burden of this disease, and provide information for health care resource allocation [11–13]. Based on the latest data from the Global Burden of Disease (GBD) Study 2019, this study was carried out to estimate the spatiotemporal trends of the smoking-related RA burden and identify highly affected regions to partially aid policy-making and current and future planning processes.

Materials and methods

Data source and definitions

Data on the global burden of RA attributable to smoking were obtained from the GBD Study 2019, which aimed to provide reliable and up-to-date global, regional and national results on the burden of diseases, injuries and risk factors [14–16]. The GBD is a publicly available database and all participants’ data were anonymous. The case definition for RA was based on the 1987 guidelines of the American College of Rheumatology [17]. Smoking cases were those that involved current or past use of any tobacco product. Former smokers were those who had given up smoking for at least six months, whereas current smokers were those who used any smoked tobacco products on a regular or infrequent basis [16].

In addition, the burden of disease was assessed using the disability-adjusted life-year (DALY) rate and the age-standardized death rate (ASDR). The world is geographically divided into 21 regions, and the sociodemographic index (SDI) is a composite indicator used to assess the development status of each location in a year and includes five levels, low SDI, low-middle SDI, middle SDI, high-middle SDI and high SDI [14]. The number of deaths, DALYs and corresponding age-standardized rates (ASRs) attributable to smoking from 1990 to 2019 were extracted by age, sex, location, year and SDI using the Global Health Data Exchange (GHDx) query tool. The general methods for the GBD Study 2019 and the methods for estimations of disease burden have been detailed in previous studies [15].

Analysis procedures

First, the global ASRs of prevalence, incidence, death and DALY changes in RA from 1990 to 2019 were analysed to determine the disease burden trend. Then, the global burden of RA attributable to smoking from 1990 to 2019 was evaluated. Afterwards, we analysed attributable burdens by year, age, sex, location and SDI to determine the impact of smoking on RA in detail. When comparing different populations or the same population in different periods, the ASDRs and DALYs were utilized to eliminate the effects caused by differences in population structures.

Statistical analysis

The ASR is a measure that can eliminate the influence of population age structure differences to the greatest extent [18], which (per 100 000 population) was calculated using the sum of the products of the age-specific rates, i.e. $ASR=∑i=1Aaiwi∑i=1Awi$×100 000, [$ai$⁠, the i^th age class; and the number of individuals (or weight) (⁠$wi$⁠) in the identical age subgroup i of the selected reference standard population]. All cases and their corresponding ASRs per 100 000 people were recorded with 95% uncertainty intervals (UIs) [14]. The 95% UIs for all estimates in the GBD Study were applied using the DisMod-MR 2.1 tool. DALYs were the sum of years lived with disability (YLDs) and years of life lost (YLL) [15]. YLL were equal to deaths multiplied by the standard life expectancy at each age, and YLDs were equal to the prevalence multiplied by disability weights for mutually exclusive sequelae of diseases and injuries. One DALY represents the loss of the equivalent of one year of full health [19]. In addition, the annual percentage change (APC) was used to estimate the rate of change in a given time period. When we assessed the temporal trends in age-standardized death and DALY rates, the joinpoint regression model was employed. The connecting of several line segments on a logarithmic scale at the ‘joinpoints’ delineated the changes in trends, which identified points where the trend line slope significantly changed over time [20]. APCs and their 95% confidence intervals (CIs) were calculated based on the following formulas: $APC=yx1-yxyx×100=eβ1-1×100$⁠, (y, rate; x, year; $β1$⁠, regression coefficient). The significance tests used a Monte Carlo permutation method.

The R software (R Core Team, version 3.5.2, Vienna, Austria) was used to perform all the above-mentioned analyses. Data are presented as values with their 95% CIs or 95% UIs. Pearson correlation tests were employed to assess the link between the burden of RA due to smoking and the SDI to determine the influence of sociodemographic variables on this burden.

Results

1990–2019 Global RA disease burden trends

The global disease burden of RA increased considerably from 1990 to 2019 (Fig. 1 and Supplementary Table S1, available at Rheumatology online). The age-standardized prevalence rate increased from 207.5 (95% UI: 190.0–227.0) per 100 000 people to 224.3 (95% UI: 204.9–246.0) per 100 000 people. Similarly, there was a growing prevalence from 9.0 million (95% UI: 8.2–9.9) in 1990–18.6 million (95% UI: 17.0–20.4) in 2019. Regarding the incidence, a large increase was found from 0.6 million (95% UI: 0.5–0.6) in 1990–1.1 million (95% UI: 1.0–1.2) in 2019. Likewise, the standardized incidence rate rose from 12.2 (95% UI: 11.1–13.4) per 100 000 people to 13.0 (95% UI: 11.8–14.3) per 100 000 people. In contrast, the ASDR decreased from 0.63 (95% UI: 0.5–0.8) per 100 000 people to 0.57 (95% UI: 0.4–0.7) per 100 000 people. However, the number of global RA deaths increased from 22.2 thousand (95% UI: 17.5–27.2) in 1990 to 44.4 thousand (95% UI: 34.0–51.8) in 2019, an increase of 100.1%. There was a very slight increase from 39.1/100 000 people to 39.6/100 000 people in the age-standardized DALY rate. Correspondingly, the number of DALYs increased from 1.7 million (95% UI: 1.3–2.1) in 1990 to 3.3 million (95% UI: 2.5–4.1) in 2019, an increase of 96.4%. Regarding the prevalence, incidence, death and DALY rates, all showed a considerable yearly increase from 1990 to 2019. The age-standardized rates of prevalence, death and DALYs also increased, except in 2016 and 2017. However, there was a wave in the changing trend of the ASDR, with the lowest point in 2012 and the highest point in 1990.

Global burden of RA attributable to smoking in 2019

Except for a few regions, there were increases in the global and regional RA burdens attributable to smoking from 1990 to 2019, which are detailed in Supplementary Table S2, available at Rheumatology online. Specifically, smoking contributed to 312.9 thousand (95% UI: 95.2–563.6) DALYs and 3.8 thousand (95% UI: 1.1–6.7) deaths from RA worldwide in 2019. Smoking was responsible for 11.9% of total deaths and 12.8% of total DALYs of RA in 1990, whereas it accounted for 8.5% of total deaths and 9.6% of total DALYs of RA in 2019. A reduction was observed in the proportion of deaths and DALYs resulting from smoking.

Temporal trends in the RA burden attributable to smoking

RA deaths and DALYs due to smoking exposure have decreased considerably since 1990, but the age-standardized death and DALY rates due to this risk factor presented different trends. As shown in Fig. 2, although the trends of two parameters ascribed to smoking exposure were approximately on a declining curve, the ASDR increased with a 1.0% APC (P = 0.117) between 1999 and 2002. Moreover, in other periods, the smoking-attributable ASDR of RA decreased with different APCs, and the most significant decrease took place between 1990 and 1996 (APC = −2.6%, P < 0.001). The age-standardized DALY rate also dropped markedly with different APCs from 1990 (P < 0.001), reducing the burden by ∼30% by 2019, and the most remarkable decline occurred between 2014 and 2017 (APC = −2.1%, P < 0.001).

Age- and sex-specific RA burden attributable to smoking

As shown in Supplementary Fig. S1, available at Rheumatology online, smoking exposure had the greatest contribution to the RA burden among people aged 50–74 years. From 1990 to 2019, the smoking-attributable DALYs of RA remained stable among older people, and deaths presented slight fluctuations among people aged <49 years. However, there have been substantial increases in the smoking-attributable burden of RA in other age cohorts since 1990. In the subgroup analysis of sex, the total burden of RA due to smoking among males was higher than that among females, and the distributions of DALYs and deaths attributable to this risk factor in different age groups were divergent between females and males (Fig. 3). For males, DALY rates of RA attributable to smoking increased with age, peaking among people aged 70–74 years, then declining yearly, which was in accordance with the rates among females. Correspondingly, the male death rate of RA was exacerbated with age, peaking among people aged 85–89 years and then falling. In contrast, for the female death rate, there was an increase with age, of which the highest was among people aged >95 years.

Regional and national RA burdens attributable to smoking

Regionally, the highest age-standardized DALYs attributed to smoking in both 1990 and 2019 were in high-income North America, while the highest ASDRs attributed to smoking in both 1990 and 2019 were seen in Australasia and East Asia (Supplementary Fig. S2A and B, available at Rheumatology online). In addition, western sub-Saharan Africa had the lowest age-standardized DALYs and deaths in both 1990 and 2019. The largest increase in the age-standardized DALYs occurred in Eastern Europe from 1990 to 2019, with an increase of ∼27.7%, and the largest increase in ASDRs was identified in Central Asia from 1990 to 2019, with an increase of ∼24.3%. Notably, the most considerable reduction in age-standardized DALYs was seen in southern sub-Saharan Africa from 1990 to 2019, with a decrease of ∼55.1%, while the high-income Asia-Pacific region had the highest ASDR of ∼70.3%.

Regarding sex in the 21 GBD regions, the highest age-standardized DALYs were seen among females in high-income North America, and males in East Asia had the highest ASDR of RA ascribed to smoking in 2019 (Fig. 4A and C). However, the lowest male age-standardized DALY rate and female ASDR were noted in western sub-Saharan Africa. In terms of changes in age-standardized DALY rates, the regions with the largest increase in DALY rates among males and females were western sub-Saharan Africa and Eastern Europe during the last 30 years, respectively, and the region with the largest decrease in DALY rates among males and females was southern sub-Saharan Africa (Fig. 4B and D). Regarding the change in ASDR, the uppermost exacerbations were observed among males and females in Oceania and Eastern Europe, respectively, while the greatest improvements were observed among both males and females in the high-income Asia-Pacific region.

Ireland had the highest age-standardized DALY rate in 2019 and ASDR in 1990 among the 204 nations and territories in comparison to other areas. The highest age-standardized DALY rate in 1990 was observed in the United Kingdom (UK), and the highest ASDR in 2019 was in Nepal (Fig. 5A and B). The lowest age-standardized death rates and DALY rates were recorded in Niger and Burkina Faso in 2019, respectively. Bosnia and Herzegovina saw the largest age-standardized DALY rate of RA attributable to smoking exposure increase over the past 30 years, while Paraguay experienced the largest increase in the ASDR among the 204 regions (Supplementary Table S3, available at Rheumatology online). In contrast, the UK demonstrated the highest reduction in age-standardized death and DALY rates.

Association of the RA burden attributable to smoking and the SDI

The age-standardized DALY and death rates of smoking-related RA varied considerably with the SDI (Fig. 6A and B). On the one hand, the age-standardized DALY rates gradually increased when the SDI was below 0.6 and then swiftly increased with an increase in the SDI after a momentary plateau. When the SDI was ∼0.83, a plateau was once again observed. The ASDRs of RA caused by smoking, on the other hand, increased when the SDI was below 0.55 and then dropped with the SDI until ∼0.64 when they increased once more. The age-standardized DALY and death rates both pointed to greater burdens in high SDI countries over the previous 30 years.

Discussion

RA still is a key health issue worldwide. Based on the GBD 2019, Cai et al. [21] reported the most comprehensive and up-to-date information about the burden of RA and its trends in subsequent years but did not focus on RA-related risk factor analysis. Thus, in the present work, we summarized the results for the global prevalence of RA, drew attention to the burden of RA associated with smoking, the vital risk factor, in 204 countries and territories from 1990 to 2019 using data from the GBD Study 2019, and examined secular trends in prevalence for the first time. Age-standardized prevalence and incidence rates of RA increased over the past 30 years worldwide, while the indicators decreased in some years, which may be explained by exposures to environmental risk factors but also by genetic factors, differing demographics and underreporting in other parts of the world [22]. Additionally, the relatively stable trends in the age-standardized death and DALY rates of RA over the past three decades likely reflected modifications in therapy paradigms and generally improved management of the disease [23]. However, in 2019, the age-standardized death and DALY rates for RA attributable to smoking were 0.6 and 39.6 per 100 000 people, respectively, which continued to place a considerable burden on global health. It is critical to understand how the burden of RA is changing due to smoking, which is considered to be the most common environmental risk [24], to enhance national and regional strategies for managing this disease.

The absolute number of deaths and DALYs of RA attributable to smoking increased considerably in 2019 compared with 1990. Fortunately, two age-standardized markers were largely reduced. The results might be ascribed to two possible reasons: on the one hand, the growth and ageing of the population stimulated the aggravation of smoking and increased disease burden [14, 25]; on the other hand, the body of evidence about declining global tobacco use worldwide due to the implementation of health laws around the world was too important to ignore [26]. This was a positive shift, but the challenge is now to identify the shortcomings and then take action to lower the cost of health care worldwide.

In terms of a specific year, there were upward trends in the ASDRs in 2000 and 2001. There was a robust connection between social economic and health outcomes [27, 28]. Therefore, we speculated that the increase in ASDRs in these two years resulted from the global economic crisis in 1998, which also affected the decline rate of age-standardized DALYs. However, there was no statistically significant difference in the impact, which may be because fewer individuals could purchase cigarettes as a result of the monetary crisis, which indirectly reduced the burden of RA brought on by smoking. In contrast, the fiscal crisis in 2008 did not seem to have meaningful effects on the age-standardized death or DALY rate. Obviously, national and regional health laws on smoking have played an encouraging role against the burden of RA [26, 29].

Smoking, which is a social behaviour in some cases, has traditionally been considered a symbol of masculinity and was exclusively reserved for men in China [30]. Additionally, female smokers tend to feel guiltier about smoking and are subjected to greater stigma [31]. Moreover, the sex-based burden differential may be explained by the variation in smoking intensity between men and women [32]. Therefore, our study confirmed that smoking had a stronger negative influence on men across all age groups, similar to the results of previous studies [10, 32]. There were complicated factors that contributed to the age gap between when DALY and death rates were at their maximum levels. The DALY rate distribution of the affected population displayed a mountain pattern, which may be due to the concurrent musculoskeletal diseases of elderly individuals, including osteoarthritis, osteoporosis and low back pain [33, 34]. Similarly, the effects of other systemic disorders on elderly people cannot be disregarded when looking at the death rate.

The proportion of the RA burden attributable to smoking varied considerably across regions and countries. The SDI was a key factor that affected the death rate and loss of health because high SDI countries have better income conditions, which also means better health care services. As a commodity, tobacco use has been linked to household economic circumstances, leading to a very small smoking population in low SDI areas and, consequently, a limited burden [35]. From the preceding, it can be deduced that the DALY and death rates of RA brought on by smoking were influenced by the tendency of the balance between the number of smokers and medical services. This study demonstrated that there were higher DALY rates and mortality rates in 2019 in higher SDI regions. More specifically, the top three regions in terms of age-standardized DALY rates were high-income North America, South America and Western Europe, all of which had a higher SDI. In contrast, the three regions with the highest ASDRs in 2019 were East Asia, South Asia and high-income North America, of which South Asia was not a region with a high SDI. This might be due to the unique circumstance in that bidis were manufactured by a cottage industry that managed to circumvent many of the regional and global tobacco laws and levies imposed on factory-made cigarettes in South Asia [36]. As a result, bidis were offered at low cost, in a variety of price ranges, and in packaging with poorly visible health warnings, and high ASDRs caused by smoking followed by 2019 [37].

Our study showed that as drastic declines in the RA burden attributable to smoking were found during the past 30 years, smoking prevention programs and health care services made substantial developments in regions such as the UK, Netherlands and Ireland, with the UK having made the most progress. Smoking prevalence in the UK has been successfully managed, leading to a decrease in smoking-related DALYs and mortality through smoke-free regulations, the provision of aid to stop using tobacco, and raising the minimum age for purchasing cigarettes [38–40], which suggests that regions where progress has been slower in reducing smoking-related burdens of RA should study and consider implementing policies and strategies that have been applied in countries such as the UK, which show the greatest burden declines.

This study is the first comprehensive analysis of spatiotemporal trends in the burden of RA caused by smoking. Previous research on this subject focused on the connection between smoking and RA but did not reveal the fundamental disease burden caused by smoking and its distribution on a global scale. Our research offers novel insights for countries and local health organizations to take steps to reduce the RA burden and its associated economic burden by reducing modifiable risk factors for the disease.

Nevertheless, this study had several limitations. First, there were acknowledged flaws regarding the methodology of the GBD Study. Second, due to the poor state of health care, data from less developed nations and regions may not have been very thorough or precise, which limited the interpretability of our study and possibly caused an underestimation of the effects of the disease burden. Third, there might have been some unforeseen possible biases as a result of the integration of heterogeneous information from several databases. For example, the smoke from the burning end of a cigarette has more toxins than the smoke inhaled by smokers. The burdens caused by passive smoking were not analysed in our study, which affected the accuracy and validity of the data in the present study. Besides, due to the limitation of the data, we could not conduct in-depth research on tobacco types or exposure doses. In addition, people in some countries may have turned to the use of cannabis and cigarette substitutes, such as e-cigarettes and smokeless tobacco, owing to the recognition of the risks of cigarette smoking, the regulatory restrictions, and the easy access to other tobacco products, which should be regarded as another limit of our results in the present study.

In conclusion, our study demonstrated that the overall burden of RA attributable to smoking decreased worldwide from 1990 to 2019. Most nations ratified the Framework Convention on Tobacco Control; however, there were considerable regional differences in how effectively tobacco control measures were implemented. The implementation of effective tobacco control policies varied greatly between countries. This risk factor increased the RA burden and societal health care costs in a number of regions with limited implementation. In our study, the greatest reduction in the burden of RA attributable to smoking was observed in the UK, which indicated that national efforts aimed at eradicating smoking have played a role in alleviating the disease burden and that this endeavour merits greater attention.

Conclusion

With the exception of a few years, the global burden of RA brought on by smoking decreased yearly from 1990 to 2019. Furthermore, the burden associated with smoking exposure was greater for men, older adults, and people living in high-middle and high SDI countries and regions. In regions where progress in reducing the burden of RA associated with smoking has slowed, it is critical to focus on investigating and taking into account the rules and strategies that have been employed in successful countries.

Supplementary material

Supplementary material is available at Rheumatology online.

Data availability

All data were obtained from the public open database: Global Health Data Exchange (GHDx) query tool (http://ghdx.healthdata.org/gbd-results-tool). The datasets used and analysed by this study are available from the corresponding author on reasonable request.

Contribution statement

Z.W., Y.G., R.W., Y.H., H.G., Z.Y., Z.J., H.J., S.L. and H.Z. all made important contributions to the study. Z.W., Y.G. and R.W. drafted the manuscript. Y.G, Z.Y. and H.J. collected and analysed the data. Y.H. and Z.J. were responsible for data visualization. S.L. and H.Z. made the study design and revised the manuscript. H.Z. supervised the manuscript. All the authors proofread and approved the final manuscript. All authors wrote, reviewed and edited the original draft and approved the final submitted version. Each author agreed to the statement here for their own contributions. Z.W. and Y.G. contributed equally to this work.

Funding

This work was supported by the Shanghai Municipal Health Commission [20MC1920600, shslczdzk03901 and ZY(2021–2023)-0209–02] and the National Natural Science Foundation [82074466].

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

Acknowledgements

We are highly grateful to the work of the Global Burden of Disease Study 2019 collaborators.

References

Author notes