Article Text
PDF
Abstract
Objectives To provide an overview and in-depth analysis of temporal trends in prevalence of musculoskeletal (MSK) disorders in women of childbearing age (WCBA) at global, regional and national levels over the last 30 years, with a special focus on their associations with age, period and birth cohort.
Methods Estimates and 95% uncertainty intervals (UIs) for MSK disorders prevalence in WCBA were extracted from the Global Burden of Diseases, Injuries and Risk Factors Study 2019. An age-period-cohort model was adopted to estimate the overall annual percentage change of prevalence (net drift, % per year), annual percentage change of prevalence within each age group (local drift, % per year), fitted longitudinal age-specific rates adjusted for period deviations (age effects) and period/cohort relative risks (period/cohort effects) from 1990 to 2019.
Results In 2019, the global number of MSK disorders prevalence in WCBA was 354.57 million (95% UI: 322.64 to 387.68). Fifty countries had at least one million prevalence, with India, China, the USA, Indonesia and Brazil being the highest accounting for 51.03% of global prevalence. From 1990 to 2019, a global net drift of MSK disorders prevalence in WCBA was −0.06% (95% CI: −0.07% to −0.05%) per year, ranging from −0.09% (95% CI: −0.10% to −0.07%) in low-middle sociodemographic index (SDI) region to 0.10% (95% CI: 0.08% to 0.12%) in high-middle SDI region, with 138 countries presenting increasing trends, 24 presenting decreasing trends and 42 presenting relatively flat trends. As reflected by local drift, higher SDI regions had more age groups showing rising prevalence whereas lower SDI regions had more declining prevalence. Globally, an increasing occurrence of MSK disorders prevalence in WCBA beyond adolescent and towards the adult stage has been prominent. Age effects illustrated similar patterns across different SDI regions, with risk increasing with age. High SDI region showed generally lower period risks over time, whereas others showed more unfavourable period risks. High, high-middle and middle SDI regions presented unfavourable prevalence deteriorations, whereas others presented favourable prevalence improvements in successively birth cohorts.
Conclusions Although a favourable overall temporal trend (net drift) of MSK disorders prevalence in WCBA was observed over the last 30 years globally, there were 138 countries showing unfavourable rising trends, coupled with deteriorations in period/cohort risks in many countries, collectively raising concerns about timely realisation of the Targets of Sustainable Development Goal. Improvements in the MSK disorders-related prevention, management and treatment programmes in WCBA could decline the relative risk for successively younger birth cohorts and for all age groups over period progressing.
- Epidemiology
- Arthritis, Rheumatoid
- Inflammation
Data availability statement
Data are available in a public, open access repository. Data are available in a public, open access repository. We downloaded data from the Global Health Data Exchange (GHDx) query tool (https://vizhub.healthdata.org/gbd-results/).
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Although several studies have reported secular trends in musculoskeletal (MSK) disorders prevalence in women of childbearing age (WCBA), they used national data or focused on only one certain disease. A variety of countries around the world lack resource to investigate the prevalence trends for MSK disorders in WCBA, therefore, cannot longitudinally track progress in diseases control, and few studies have explored the trends associated with age, period and birth cohort effects in MSK disorders prevalence in WCBA from a global view.
WHAT THIS STUDY ADDS
This is the first study using the age-period-cohort model to comprehensively analyse the temporal trends in prevalence of MSK disorders in WCBA at global, regional and national levels from 1990 to 2019. Although a favourable overall declining trend (net drift) in MSK disorders prevalence in WCBA was detected globally, 138 countries showed unfavourable rising trends, and there was a great heterogeneity in trend within each age group (local drift). Notably, an increasing occurrence of MSK disorders prevalence in WCBA beyond adolescent and towards the adult stage has been prominent. Differentiation of age, period and birth cohort effects enables us to determine the major factor driving changes in prevalence trends for each country, thereby clarify disparities among countries and identify potential gaps in disease control, contributing to an in-depth understanding of the epidemiology and public health making.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Unfavourable rising trends of MSK disorders prevalence in WCBA in more than half of the countries worldwide, coupled with deteriorations in period/cohort risks in manycountries, collectively suggesting the prominent inadequacy of resources invested
into the prevention, management and treatment programmes of these diseases. Of note, the global healthcare of MSK disorders in WCBA needs to be tilted towards the adult stage in which there was increasing prevalence over the last 30 years. To realise the Targets of Sustainable Development Goal 3 of reducing global maternal mortality by 2030, increasing resource investment in healthcare of MSK disorders in WCBA especially at adult stage is urgently needed. Such improvements would decrease the risk for successively younger birth cohorts and for all age groups over period progressing.
Introduction
Musculoskeletal (MSK) disorders are chronic conditions with a remarkable female preference, predominantly occurring at women of childbearing age (WCBA).1 2 In 2015, the United Nations established the Targets of Sustainable Development Goal 3 to reduce the global maternal mortality ratio to less than 70/100 000 live births by 2030.3 Women with MSK disorders are especially relevant as this group is at high risk of developing severe pregnancy complications, leading to striking maternal morbidity and mortality.4 5 Concerns have been prominent that a global overview and analysis of MSK disorders prevalence in WCBA is lacking, suggesting an insufficient attention to this special population, which may limit the realisation of related items in Targets of Sustainable Development Goal 3.
Previous nationwide population-based study in Korea revealed an annual increase in the incidence and prevalence of seropositive rheumatoid arthritis in WCBA.6 Furthermore, WCBA with rheumatic diseases in Korea was found to have higher burden of comorbidities and medication use than those without rheumatic diseases at the same age.7 To provide information for elucidating epidemiology, tracking disease management progress, and determining further interventive priorities, a comprehensive analysis of temporal trends in MSK disorders prevalence for all countries is needed. Notably, the temporal trends are usually reflected by multidimensions to derive an in-depth understanding. For one certain population, prevalence risk of diseases can be further decomposed into age, period and birth cohort effects. The risk of prevalence for MSK disorders in WCBA is not only affected by physiological age but also altered by different periods. It has been reported that the prevalence of seropositive rheumatoid arthritis in WCBA increased with advancing age.6 The distinctions in exposure to environmental risk factors,8 as well as advances in medical technology and public health policy relating to the precaution and management of MSK disorders in different periods impact the prevalence risk significantly. Moreover, the prevalence risk for MSK disorders in WCBA also differs among birth cohorts with the cumulation of various life events. Early-life social, behavioural and health exposures have long-lasting effects on later-life arthritis outcomes.9 Therefore, an in-depth analysis of prevalence trends with a special focus on their associations with age, period and birth cohort effects can broaden our understanding in diseases epidemiology, as well as identify the potential gaps in different aspects of the prevention, management and treatment programmes for diseases.
Although a yearly increase in prevalence of seropositive rheumatoid arthritis in WCBA was observed in Korea,6 the relative effects of age, period and birth cohort contributed to prevalence have not been investigated. Moreover, a variety of countries all over the word, including low-income, middle-income and high-income countries, lack information on prevalence trends for MSK disorders in WCBA, as well as their associations with age, period and birth cohort effects. The Global Burden of Diseases, Injuries and Risk Factors Study (GBD) 2019 generated population-level health metrics using the latest available epidemiological data and advanced statistical methodologies, thus provides an opportunity to analyse the temporal trends of diseases from a global perspective. In this study, we extracted data from GBD 2019 and used age-period-cohort (APC) model to explore the temporal trends in prevalence of MSK disorders in WCBA at global, regional and national levels from 1990 to 2019.
Methods
Overview of GBD
GBD 2019, which was generated by the Institute for Health Metrics and Evaluation, provided a comprehensive up-to-date estimation on the health loss for 369 diseases, injuries and impairments and 87 risk factors by age and sex across 204 countries and territories. Details in the methodologies applied in GBD 2019 have been thoroughly explained in previous literature.10 11
Information on fatal and non-fatal estimates was available at the GBD 2019 websites (https://vizhub.healthdata.org/gbd-compare/; and http://ghdx.healthdata.org/gbd-results-tool). GBD 2019 complied with the Guidelines for Accurate and Transparent Health Estimate Reporting statement.12
GBD data processing and modeling process
GBD 2019 compiled data sources through data identification and extraction. By identifying from systematic reviews, government and international organisation websites, published reports, primary data sources and contributions of datasets by GBD collaborators, the GBD study combined a variety of data sources including censuses, household surveys, civil registration and vital statistics, disease registries, health service use, air pollution monitors, satellite imaging, disease notifications and other sources, among which each of them was given a unique identifier and included in the Global Health Data Exchange (GHDx). The systematic review in GBD 2019 mainly included three approaches: (1) electronic searches in databases, (2) grey literature searches and (3) consultation from experts. Garbage code redistribution and noise reduction data, together with small sample size were excluded. Data sources used in estimating the burden of MSK disorders across different countries and territories worldwide can be found in the GBD 2019 data input source tool (http://ghdx.healthdata.org/gbd-2019/datainput-sources).10
GBD 2019 corrected bias and adjusted data for further modelling process. The collected data were modelled by spatiotemporal Gaussian process regression to allow for smoothing over age, time and location in locations lacking complete datasets. The Meta-Regression with Bayesian priors, Regularisation and Trimming (MR-BRT) programme was adopted to adjust data bias for alternative case definitions and study methods. There are three types of age splitting and sex splitting: (1) Estimates were split by age and sex, where possible. Nevertheless, age-specific estimates would be calculated based on the reported sex ratio and uncertainty bounds when the prevalence data were reported for specific age groups without separating by sex, or by sex for large age intervals; (2) The remaining individual sex estimates were performed with MR-BRT. MR-BRT network meta-analysis estimated the pooled sex ratios and uncertainty bounds, which were then adopted to split combined sex estimates and (3) Researches presenting prevalence estimates for age groups of 25+ years were split into 5-year age groups using the age pattern generated by DisMod-MR 2.1.10 13
GBD 2019 modelled the epidemiology of MSK disorders using a meta-regression tool based on Bayesian model framework, namely, DisMod-MR 2.1. The Bayesian approach serves as an interpretation of statistical probability, where existing data are used to inform the probability of a given hypothesis. A meta-regression can be regarded as an extension of a meta-analysis whereby data from different sources are pooled into a weighted average adjusting for heterogeneities. DisMod-MR 2.1 applied a negative-binomial model of disease incidence, prevalence, remission and case-fatality rates, and fitted models with a randomised Markov-Chain Monte Carlo algorithm. The steps of modelling process in DisMod-MR 2.1 are as follows: (1) It pooled heterogeneous raw data for each parameter and adjusted data for methodological distinctions. If the data were insufficient to indicate an age-pattern variation, DisMod-MR 2.1 may impose a common age pattern according to assessment of age-specific input data for the disease; (2) It checked data on incidence, prevalence, duration, remission and mortality risk for internal consistency; (3) It simultaneously integrated the input data from all parameters plus to the outputs from previous steps to obtain internally consistent epidemiological estimates, carrying forward uncertainty from primary data sources; (4) Even for countries with little or no primary data source, this model could produce estimates based on information from the available data, and this process allowed for estimates of MSK disorders burden in all countries worldwide.10 14 Regarding each specific MSK disorders including rheumatoid arthritis, osteoarthritis, low back pain, neck pain, gout and other MSK disorders, details on flow charts, definitions, input data and modelling strategies are available in online supplemental appendix 1 of the GBD 2019 study (https://www.thelancet.com/cms/10.1016/S0140-6736(20)30925-9/attachment/7709ecbd-5dbc-4da6-93b2-3fd0bedc16cc/mmc1.pdf, Page 1184–1218).
Supplemental material
Data extraction
Metrics in GBD 2019 include estimates and their 95% uncertainty intervals (UIs), which are defined by the 25th and 975th values of the ordered 1000 estimates according to the GBD’s algorithm. All rates are reported per 100 000 population. Data used in this study were downloaded from the GHDx query tool (https://vizhub.healthdata.org/gbd-results/).15
GBD 2019 also produced a sociodemographic index (SDI) for each country or territory serving as a comprehensive indicator of income, education and fertility conditions.16 The SDI values ranged from 0 to 1, with higher values suggesting higher socioeconomic development levels. Based on the GBD 2019 SDI values, each country was categorised into one of the five SDI quintiles including low, low-middle, middle, high-middle and high.
Study population
WCBA denotes the phase of women with reproductive ability and experiences cyclical changes in sex hormones. The WHO defined WCBA as women aged 15–49 years.17 Moreover, adolescence serving as the phase of life between childhood and adulthood was defined as aged 10–19 years by WHO.18 Therefore, WCBA aged 15–49 years in this study can be further divided into adolescent stage (15–19 years) and adult stage (20–49 years).
Analysis of overall temporal trends in MSK disorders prevalence in WCBA
The number and age-standardised rate from 1990 to 2019 were adopted to evaluate the overall temporal trends in MSK disorders prevalence in WCBA. For estimating the age-standardised prevalence rate of MSK disorders in WCBA, age standardisation by direct method, which assumes that the rates are distributed as a weighted sum of independent Poisson random variables,19 20 was used. Furthermore, the relative proportion of MSK disorders prevalence in WCBA stratified by seven age groups (15–19, 20–24, 25–29, 30–34, 35–39, 40–44, 45–49 years) was calculated and the temporal changes in age distribution of prevalence were illustrated.
APC analysis of MSK disorders prevalence in WCBA
In this study, an APC model framework was used to analyse the temporal trends in prevalence by age, period and birth cohort. The APC model is usually recognised to be an advanced method beyond the traditional analyses in health and social science, with a special focus on determining the contributions of age-associated natural history, contemporary time-associated medical technological and social factors, as well as early years of life-associated health behaviours set and social exposure on disease trends.21 It can not only determine the net drift and local drift of disease burden over time, which represent the overall temporal trend and temporal trend within each age group, respectively, but also estimate the age, period and birth cohort effects, which are regarded as three fundamental dimensions of time, therefore, may provide an in-depth understanding in disease epidemiology, and help to identify the potential gaps in different aspects of the prevention, management and treatment programmes for diseases. This method has been applied into epidemiological research for non-communicable diseases such as congenital heart disease.22 However, there is a so-called identification problem, that is, the exact collinearity between these three variables (age=period–birth cohort), leading to the statistical impossibility to estimate the independent effects of age, period and birth cohort.21 To circumvent this problem, we here produced estimable APC parameters and functions without imposing arbitrary constraints on the model.22 23 The details in methodology of APC model have been thoroughly discussed elsewhere.24
When preparing input data for the APC model, prevalence estimates for MSK disorders in WCBA and population data of each country/region from GBD 2019 were adopted. Typically, age intervals must be equal to period intervals, that means, 5-year age groups should be used with 5-year calendar periods. As the childbearing age of women was defined as 15–49 years, seven age groups including 15–19, 20–24, 25–29, 30–34, 35–39, 40–44, 45–49 years were adopted for further analysis. Correspondingly, the whole study period (1990–2019) was divided into six 5-year periods, that is, 1990–1994, 1995–1999, 2000–2004, 2005–2009, 2010–2014, 2015–2019. Therefore, 12 partially overlapping 10-year birth cohorts, namely, 1940–1949, 1945–1954, 1950–1959, 1955–1964, 1960–1969, 1965–1974, 1970–1979, 1975–1984, 1980–1989, 1985–1994, 1990–1999, 1995–2004, were used.
The APC model estimates both overall temporal trend in prevalence and temporal trend of prevalence within each age group. The former is expressed as the annual percentage change of prevalence, that is, the net drift of prevalence (% per year), which is determined based on both components of the trend attributable to calendar time and successive birth cohorts, whereas the latter is expressed as the annual percentage change of age-specific prevalence, that is, the local drift of prevalence (% per year). Even if the value of a drift (% per year) is slight, it may lead to a substantial change in the fitted rate over a period of 30 years. A Wald χ2 test was used for testing the significance of trends in annual percentage change.24 Moreover, in the APC model, the age effects are represented by fitted longitudinal age-specific rates for a given number of birth cohorts adjusted for period deviations. The period/cohort effects are represented by period/cohort relative risks of prevalence, which are calculated as the ratio of age-specific rates in each period/cohort relative to reference period/cohort.24 The choice of reference period/cohort is arbitrary and does not affect results interpretation. A two-sided p value less than 0.05 was considered significant. All analyses and visualisation were performed in R (V.4.2.1).
Results
Trends in MSK disorders prevalence in WCBA, 1990–2019
Global and regional population, number of prevalence, age-standardised prevalence rate, as well as net drift of prevalence were demonstrated in table 1. From 1990 to 2019, along with the increase of global population, the global number of MSK disorders prevalence in WCBA increased by approximately 55.87%, reaching at 354.57 million (95% UI: 322.64 to 387.68) in 2019. The percentage change in the number of prevalence increased in all SDI regions. The global age-standardised prevalence rate for MSK disorders in WCBA was 17 831.22 (95% UI: 15 121.43 to 20 805.61) per 100 000 population in 2019, with a 2.1% decrease from 1990. The relative increase in age-standardised prevalence rate occurred in high, high-middle and middle SDI regions. Furthermore, the APC model estimated a global net drift of MSK disorders prevalence in WCBA at −0.06% (95% CI: −0.07% to −0.05%) per year, ranging from −0.09% (95% CI: −0.10% to −0.07%) in low-middle SDI region to 0.10% (95% CI: 0.08% to 0.12%) in high-middle SDI region.
Trends of MSK disorders prevalence in WCBA from 1990 to 2019 by SDI quintiles
National prevalence number and age-standardised prevalence rate in 2019, as well as net drift of prevalence from 1990 to 2019 for MSK disorders in WCBA were shown in figure 1 and online supplemental table S1. In 2019, among the 204 countries and territories, 50 had at least 1 million prevalence number, with India, China, the USA, Indonesia and Brazil being the top five countries and responsible for 51.03% of MSK disorders prevalence in WCBA globally. Seventy-five had at least the global average age-standardised prevalence rate, and three countries including the USA, Greenland and Canada, had a prevalence rate more than 1.5-fold of the global average level, most of which were high SDI countries. From 1990 to 2019, Saudi Arabia had the highest increase in age-standardised prevalence rate (15.83%), with a net drift of prevalence at 0.50% (95% CI: 0.47% to 0.52%) per year. The relative change in age-standardised prevalence rate decreased in only 28 countries. Although India and China had the highest number of MSK disorders prevalence owing to their large population, their age-standardised prevalence rate declined slightly, with relatively modest net drifts of prevalence. Of these 204 countries and territories, the net drift estimated from APC model suggested 138 with increasing trends, 24 with decreasing trends and 42 with relatively flat trends, suggesting a strong heterogeneity in MSK disorders prevalence trends across all countries.
Supplemental material
The age-standardised prevalence rate in 2019 and net drift of prevalence from 1990 to 2019 for MSK disorders in WCBA in 204 countries and territories. (A) World map of the age-standardised prevalence rate in 2019 for MSK disorders in WCBA. Globally, the age-standardised prevalence rate was 17 831.22 (95% UI 15121.43 to 20805.61) per 100 000 population. (B) World map of the net drift of prevalence from 1990 to 2019 for MSK disorders in WCBA. Globally, the net drift of prevalence was −0.06% (95% CI −0.07% to –0.05%). MSK, musculoskeletal; WCBA, women of childbearing age.
Temporal trends in MSK disorders prevalence in WCBA across different age groups
The annual percentage change in MSK disorders prevalence in WCBA for each age group, that is, the local drift of prevalence calculated from APC model, was presented in figure 2A and online supplemental table S2. Globally, MSK disorders prevalence in WCBA demonstrated decreasing trends from 15–19 to 35–39 years group, flat trend in 40–44 years and increasing trend in 45–49 years. The declining trend from 15–19 to 35–39 years group mitigated with increasing age, being lowest in adolescent stage at 15–19 years (−0.17%, 95% CI: −0.21% to −0.13%) and less evident in adult stage at 20–39 years (from −0.11% (95% CI: −0.14% to −0.09%) in 20–24 years to −0.03% (95% CI: −0.05% to −0.01%) in 35–39 years). In the adolescent stage at 15–19 years, MSK disorders prevalence increased in high SDI and high-middle SDI regions, kept flat in middle SDI region, and decreased in low-middle and low SDI regions. In the adult stage at 20–39 years, along with increasing age, the trends of MSK disorders prevalence have been generally developing towards the opposite direction compared with that in 15–19 years in all SDI regions except in high-middle SDI regions, where there was a constant increasing trend. The local drift of prevalence for each country was demonstrated in online supplemental table S3.
Local drift and age distribution of prevalence from 1990 to 2019 for MSK disorders in WCBA across SDI quintiles. (A) Local drift of prevalence from 1990 to 2019 for MSK disorders in WCBA for seven age groups (15–19, 20–24, 25–29, 30–34, 35–39, 40–44, 45–49 years). The dots and shaded areas denote the local drift (ie, annual percentage change of age-specific prevalence, % per year) and their corresponding 95% CIs. (B) Temporal changes in age distribution of MSK disorders prevalence in WCBA from 1990 to 2019. MSK, musculoskeletal; SDI, sociodemographic index; WCBA, women of childbearing age.
Temporal changes in age distribution of MSK disorders prevalence in WCBA were illustrated in figure 2B. Globally, there was a gradual transition of prevalence from the adolescent stage (15–19 years) to the adult stage (20–49 years), and this trend presented more evident in high-middle, middle and low-middle SDI regions. Moreover, older age groups accounted for higher proportion of prevalence, and over 50% of prevalence were concentrated in women older than 35 years in all SDI regions.
Age, period and birth cohort effects on MSK disorders prevalence in WCBA
The age, period and birth cohort effects derived from APC model were demonstrated in figure 3 and online supplemental tables S4–S6. Generally, there were similar patterns in age effects across different SDI regions that the lowest risk existed in adolescent stage at 15–19 years with risk increasing with age. Moreover, low SDI region showed an overall lower prevalence across all age groups compared with other SDI regions. Overall, period effects presented a declining first and then rising risk of prevalence across different SDI regions except in high SDI region. High SDI region had generally lower period risks over the study period, whereas others had more unfavourable period risks most of the time. Compared with individuals in the reference 2000–2004 period, the relative period risk for individuals in the 2015–2019 period ranged from 1.004 (95% CI: 0.992 to 1.017) in high SDI region to 1.034 (95% CI: 1.028 to 1.040) in high-middle SDI region and 1.034 (95% CI: 1.029 to 1.039) in middle SDI region. Regarding birth cohort effects, there was a rising first and then declining risk of prevalence in successively birth cohorts globally. High, high-middle and middle SDI regions had progressive prevalence deteriorations, whereas low-middle and low SDI regions had favourable prevalence improvements in successively birth cohorts. Compared with individuals born in the reference 1965–1974 cohort, the relative cohort risk for individuals born in the 1995–2004 cohort ranged from 0.917 (95% CI: 0.907 to 0.928) in low-middle SDI region to 1.091 (95% CI: 1.049 to 1.134) in high SDI region.
Age, period and birth cohort effects on MSK disorders prevalence in WCBA across SDI quintiles. (A) Age effects are illustrated by the fitted longitudinal age-specific rates for a given number of birth cohorts adjusted for period deviations. (B) Period effects are illustrated by the period relative risk of prevalence (prevalence rate ratio) and calculated as the ratio of age-specific rates from 1990–1994 period to 2015–2019 period, with the reference period set at 2000–2004. (C) Birth cohort effects are illustrated by the cohort relative risk of prevalence (prevalence rate ratio) and calculated as the ratio of age-specific rates from 1940–1949 cohort to 1995–2004 cohort, with the reference cohort set at 1965–1974. The dots and shaded areas denote the prevalence rates or rate ratios and their corresponding 95% CIs. MSK, musculoskeletal; SDI, sociodemographic index; WCBA, women of childbearing age.
In addition, the age, period and birth cohort effects on MSK disorders prevalence in WCBA for each country were shown in online supplemental tables S7–S9. To better characterise the temporal trends in MSK disorders prevalence worldwide, several exemplary countries across different SDI quintiles with relatively favourable and unfavourable age, period and birth cohort effects were presented in figure 4. The USA is typical of trends in high SDI countries with unfavourable manifestations, where no prevalence decrease was observed across all age groups with both period and cohort risks worsening in recent years. In contrast, the UK had extremely favourable trends in MSK disorders prevalence among high SDI countries, with local drift <0% per year for all age groups and noticeable declining risks over period progressing and in successively birth cohorts. China and India are examples of highly populous middle SDI countries, presenting an emerging transition of prevalence from adolescent stage to adult stage. Older age groups accounted for the rising proportion of prevalence for both China and India. Period effects showed similar patterns in these two populous countries, with declining risk in the early stage and rising risk in the later stage of the study period. Compared with China, there were more improvements in cohort risk in recent birth cohorts in India, especially for those born after 1965–1974. Mozambique stood out for its highest net drift in low SDI countries and showed significant prevalence increase for almost all age groups, with continuous deteriorations in both period and cohort risks. Burundi is another low SDI country with favourable local drift <0% per year in almost all age groups and showed notable gradual attenuation in risk over period progressing and in successively birth cohorts.
Age, period and birth cohort effects on MSK disorders prevalence in WCBA in exemplary countries. Age distribution of prevalence demonstrates the temporal changes in relative proportion of prevalence from 1990 to 2019 across seven age groups (15–19, 20–24, 25–29, 30–34, 35–39, 40–44, 45–49 years). Local drift denotes the annual percentage change of age-specific prevalence (% per year) from 1990 to 2019 for seven age groups (15–19, 20–24, 25–29, 30–34, 35–39, 40–44, 45–49 years). Age effects are illustrated by the fitted longitudinal age-specific rates for a given number of birth cohorts adjusted for period deviations. Period effects are illustrated by the period relative risk of prevalence (prevalence rate ratio) and calculated as the ratio of age-specific rates from 1990 to 1994 period to 2015–2019 period, with the reference period set at 2000–2004. Birth cohort effects are illustrated by the cohort relative risk of prevalence (prevalence rate ratio) and calculated as the ratio of age-specific rates from 1940–1949 cohort to 1995–2004 cohort, with the reference cohort set at 1965–1974. The dots and shaded areas denote the prevalence rates or rate ratios and their corresponding 95% CIs. MSK, musculoskeletal; SDI, sociodemographic index; WCBA, women of childbearing age.
Discussion
In 2015, the United Nations proposed the Targets of Sustainable Development Goal 3, which outlined the aim of decreasing the maternal mortality ratio to less than 70/100 000 live births by 2030 globally.3 Women with MSK disorders, especially within childbearing age, are extremely relevant to this goal as this population has a high risk of developing various pregnancy complications resulting in notable maternal morbidity and mortality.5 Undoubtedly, an in-depth understanding on the trends of prevalence of MSK disorders in WCBA contributes to the assessment of potentiality in meeting related health goals. However, a global overview and analysis of MSK disorders prevalence in WCBA is lacking, suggesting an urgent demand. It is known that using case number to track trends in disease prevalence is likely to be misleading for that it may be altered by the change of total population, and using age-specific rate may also be inaccurate for that it cannot adjust for heterogeneities in the age structure form different populations, resulting in a fail in valid comparisons between regions or countries. Moreover, reliance on overall rate to assess trends may be unable to differentiate the contributions of age, period and birth cohort effects. Therefore, in this study, we presented MSK disorders prevalence in WCBA with both case number and age-standardised rates. Additionally, we further used APC model to derive an in-depth understanding on the temporal trends in prevalence of MSK disorders in WCBA.
From 1990 to 2019, along with the increase of global population by 45%, the total number of MSK disorders prevalence increased by 56%, with the largest increase in low SDI region. Notably, this tremendous increase in low SDI region was mainly mediated by population growth, for that the total population increased by 114% while the age-standardised rate had a slight decrease. Our further analysis estimated the local drift to capture temporal trends in MSK disorders prevalence in WCBA for each age group. Generally, higher SDI regions had more age groups showing rising prevalence whereas lower SDI regions had more age groups showing declining prevalence, suggesting a potential SDI-related distributive inequity. Of note, the MSK disorders prevalence change observed in this study may be not fully commensurate with the general assumption that higher quality in the healthcare system and better medical service correlate well with higher SDI conditions leading to lower disease burden. This is similar with our previous study suggesting that countries with higher sociodemographic development levels shouldered disproportionately higher burden of autoimmune diseases.25 On the one hand, the initiation and exacerbation of MSK disorders are closely associated with harmful environmental exposure such as industrial air pollution emissions,8 26 27 which always exist in higher SDI regions for that these regions usually have accomplished or are experiencing industrialisation and urbanisation. On the other hand, individuals in higher SDI regions are more likely to get into nutrition overload including high fat, high sugar, high protein, etc, which are recognised to be vital risk factors for several MSK disorders such as gout.28 In addition, older age groups especially 45–49 years accounted for the major increase of MSK disorders prevalence in WCBA except in high SDI region, where the younger age groups appeared to be more prominent.
Globally, an increasing occurrence of MSK disorders prevalence in WCBA beyond adolescent and towards the adult stage has been prominent, putting forward new demands to healthcare system construction and policy-making. Pregnancy is an inevitable issue for WCBA at adult stage, and maternal mortality caused by severe pregnancy complications in this population inhibits the realisation of sustainable development goals established by the United Nations. Compared with the general obstetric population, pregnant women with MSK disorders have higher risk in developing adverse maternal, neonatal and fetal outcomes including gestational hypertension, pre-eclampsia, caesarean delivery, preterm delivery, low birth weight, congenital anomalies, stillborn, etc,29 30 which may be associated with medication use and disease activity.30 Moreover, the flare of MSK disorders may coexist with these complications, and difficulties in differential diagnosis are raised.31 Therefore, close monitoring of clinical status, control of disease activity and proper adjustment of medication regimen, are extremely vital for women with MSK disorders before and during pregnancy. The British Society for Rheumatology,32 33 European League Against Rheumatism34 and American College of Rheumatology35 have all provided guidelines on the management of reproductive health and usage of antirheumatic drugs before and during pregnancy and lactation.
The relative effects of age, period and birth cohort contributed to prevalence trends have been explored here. Age effects showed similar patterns across different SDI regions, with risk increasing with age. In fact, ageing has been recognised to be one of the most prominent risk factors for rheumatic degenerative joint diseases, which can be traced to the comprehensive effects of both cumulative exposure during natural history and age-related pathological changes in joint structures.36 Furthermore, the ageing of immune system called ‘immunosenescence’37 is tightly related to decreased adaptive immunity and increased non-specific innate immunity resulting in chronic low-grade sterile inflammation, therefore, participates in the progression of autoimmune inflammatory rheumatic diseases.38 39 Globally, unfavourable period effects and favourable cohort effects were observed. Period risks kept increasing over the last few years, which may be associated with the cumulative risk factors for MSK disorders especially harmful environment exposure derived from industrialisation and urbanisation globally,8 suggesting a striking deterioration in recent years and an emergent need in disease control. As for cohort effects, the later-born individuals showed an overall lower risk than the earlier-born individuals. On the one hand, earlier-born individuals may have more cumulative risk factors during their lifetime. On the other hand, it has been found that exposure to social, behavioural and health factors in the early-life stage exerted effects on MSK disorders outcomes.9 The later-born individuals may experience high-quality management of early-life stage exposure compared with earlier-born individuals, therefore, presented lower risk. Notably, due to the fact that the net drift representing overall temporal trend is calculated based on both components of the trend attributable to calendar time and successive birth cohorts, differentiating period and cohort trends provides opportunity to recognise the major factor driving changes in prevalence. The attenuation of MSK disorders prevalence in WCBA in low-middle and low SDI regions appeared to be mainly driven by favourable cohort effects whereas the deterioration of MSK disorders prevalence in WCBA in high SDI region appeared to be mainly driven by an unfavourable cohort effect. In high-middle and middle SDI regions, period and cohort effects exerted synergistic effects contributing to the deterioration of prevalence. In addition, there is a strong heterogeneity in age, period and birth cohort effects on MSK disorders prevalence in WCBA across 204 countries and territories, indicating distinct disease patterns worldwide, which requires a special focus when making their respective country-level health policy.
To our best knowledge, this is the first study using APC model to comprehensively analyse the temporal trends of MSK disorders prevalence in WCBA at global, regional and national levels. Compared with previous GBD publications regarding MSK disorders,40–43 the major merit of our present study is that we offer a more in-depth understanding on disease temporal trends to produce valuable insights for disease epidemiology and health policy-making. Specifically, the estimation of net drift and local drift enables us to capture overall temporal trend in prevalence and temporal trend in prevalence within each age group adjusting for period effects, respectively. Moreover, the differentiation of period and birth cohort effects enables us to determine the major factor driving changes in prevalence trends by periods and birth cohorts for each country, providing evidence for the evaluation of MSK disorders-related prevention, management and treatment programmes and allowing comparisons among different countries. Additionally, the superiority of data in GBD 2019 appeared to be prominent compared with that in GBD 2017, for instance, the data in GBD 2019 included osteoarthritis of hip, knee, hand and other whereas the data in GBD 2017 included only osteoarthritis of hip and knee. Obviously, the former can better represent the entire burden of osteoarthritis. Despite these advantages, there are several limitations in this study. First, due to the low medical healthcare levels in some underdeveloped countries, there may be potential misdiagnosis and missed diagnosis for diseases, leading to underestimated estimation. Second, the data obtained from GBD relies heavily on the modelled data, as a result of the usage of massive statistical modelling methods by GBD collaborators, particularly at the country level due to the lack of primary data. Estimates from GBD in these countries have wide uncertainty bounds.10 This may affect the accuracy of estimates of age, period and birth cohort effects. Third, the subnational characterisations of prevalence trends cannot be explored due to data unavailability. The incorporation of subnational data could further recognise areas with different trends within one certain country. Additionally, the hysteresis property of GBD data should be noted.
In summary, although a favourable overall temporal trend of MSK disorders prevalence in WCBA was detected globally, there were unfavourable increasing trends arising in over half of the countries worldwide, coupled with deteriorations in period/cohort risks in many countries, collectively suggesting that the resource invested into healthcare of MSK disorders in WCBA is largely insufficient. Notably, the global healthcare of MSK disorders in WCBA needs to be tilted towards the adult stage in which prevalence has increased over the past 30 years. For the demand of meeting Targets of Sustainable Development Goal 3 of reducing global maternal mortality by 2030, increasing resource investment in healthcare of MSK disorders in WCBA especially at adult stage are urgently needed to decrease the risk for successively younger birth cohorts and for all age groups over period progressing. It is noteworthy that gender-specific prevalence rates differ between MSK disorders, targeted policy-making and resources allocation to different conditions should be considered further.
Data availability statement
Data are available in a public, open access repository. Data are available in a public, open access repository. We downloaded data from the Global Health Data Exchange (GHDx) query tool (https://vizhub.healthdata.org/gbd-results/).
Ethics statements
Patient consent for publication
Ethics approval
For the usage of deidentified data in GBD study, a waiver of informed consent has been approved by the University of Washington Institutional Review Board.
Acknowledgments
We appreciate the excellent works by the Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) 2019 collaborators. We also thanks to Xiao Ming (Xiaoming_room@hotmail.com) for his help in our exploration of GBD database.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
FC, D-PL and G-CW are joint first authors.
Handling editor Josef S Smolen
Contributors FC, H-FP, Z-XJ and L-MT conceptualised and designed the study. FC, D-PL and G-CW obtained the raw data, performed the analysis and prepared the illustrations. FC wrote the manuscript. H-FP, Z-XJ and L-MT reviewed and revised the manuscript. Y-SH, Y-CL, J-JH and Q-YN participated in the data collection, analysis and visualisation. FC, H-FP, Z-XJ and L-MT supervised the final version of the manuscript. All authors reviewed the article, read the final manuscript and approved the submission. H-FP is responsible for the overall content as guarantor.
Funding This work was supported by the National Natural Science Foundation of China (82273710), Anhui Provincial Natural Science Foundation (2108085Y26) and Research Fund of Anhui Institute of Translational Medicine (2021zhyx-B04).
Map disclaimer The inclusion of any map (including the depiction of any boundaries therein), or of any geographic or locational reference, does not imply the expression of any opinion whatsoever on the part of BMJ concerning the legal status of any country, territory, jurisdiction or area or of its authorities. Any such expression remains solely that of the relevant source and is not endorsed by BMJ. Maps are provided without any warranty of any kind, either express or implied.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.