Trends in the Global Burden of Glomerulonephritis

The global burden of CKD is high and rising incessantly. Globally, approximately 700 million people have CKD. CKD is projected to rise in the list of cause of death from the 16th place in 2016 to 5th place in 2040.¹ CKD is costly to society, individuals, families, and health care systems. There have been significant mortality reductions in cardiovascular disease, stroke, and respiratory disease over time, but CKD deaths and disability have continued to rise. The Global Burden of Disease (GBD) Study, a worldwide program of research that provides estimates of death and disability due to major diseases, injuries, and risk factors, attributes the CKD burden to four major cause categories—diabetes, hypertension, glomerulonephritis (GN), and “other causes.” Many countries have instituted national programs to detect and treat diabetes and hypertension as they are major cardiovascular disease risk factors. GN has not received similar attention, despite being the second biggest driver of CKD-related disability-adjusted life-years (DALYs) after diabetes in low-resource settings.² Between 1990 and 2016, the age-adjusted incidence of CKD due to diabetes and hypertension has declined, whereas that of GN-related CKD had gone up.²

In this issue of CJASN, Hu et al.³ describe the incidence, prevalence, deaths, and DALYs secondary to GN using the GBD 2019 dataset. They show a rise in GN burden between 1990 and 2019, as reflected in rising age-standardized incidence, prevalence, deaths, and DALYs rates. Countries with the highest burden are in Central America (age-standardized incidence rate of 16.6 [95% uncertainty interval (UI), 15.1 to 18.3] per 100,000 population and prevalence rate of 379.3 [95% UI, 347.1 to 412.5]) per 100,000 population, whereas developed countries exhibit a lower burden. Sweden has the lowest national age-standardized incidence rate of 2.9 (95% UI, 2.3 to 3.4) per 100,000 population, and Spain has the lowest national age-standardized prevalence rate of 97.9 (95% UI, 80.2 to 120.7) per 100,000 population. There was significant difference in the deaths rates globally. The age-standardized deaths rates are more than 10 per 100,000 population in Nicaragua, El Salvador, and Mexico, and Korea had the lowest rate at 0.4 per 100,000 population. Furthermore, the study shows striking global inequities: the age-standardized per 100,000 GN DALY rates being 42.5, 47.7, 89, 117.7 and 142.6 in the declining order of sociodemographic index (SDI) quintile (highest to lowest). The analysis shows a bimodal age distribution of incidence and prevalence, with highest incidence in children aged 1–4 years and in the elderly population.

The study also conducted a frontier analysis, an econometric technique that uses regression modeling to measure efficiency of health care delivery (minimization of DALYs) on the basis of socioeconomic development of countries.⁴ Countries with higher SDI had the lowest effective differences between the observed and potentially achievable age-standardized DALYs rate, suggesting efficient care provision. Most low SDI countries were away from the frontier (Mexico, Nicaragua, and El Salvador being at the greatest distance), indicating substantial potential for improving efficiencies. Interestingly, two countries with low SDI, Somalia and Niger, were close to the frontier. The authors present them as an example of good performance, but undercounting of cases, leading to underestimation of GN DALYs, can also reduce the frontier distance.

The incidence and prevalence data disaggregated by SDI quintiles reveal some interesting insights. Low SDI countries have the lowest rates throughout the period of study, with a steeper increase in the recent years. The rates are highest in the middle SDI countries, followed by low-middle, high-middle, and high SDI countries. The most likely explanation of this trend is missed diagnosis in countries at the lower end of the development spectrum, which picks up as the countries move up the development scale and health care access increases, followed by reduction as the overall health status of the societies goes up further and disease development is prevented. Most developing countries in Africa and Asia do not have access to adequate facilities for diagnosing GN. In a recent survey focused in low-resource settings, 61% of the respondents reported inability to perform a kidney biopsy in >50% of patients with suspected GN.⁵ Only 43% of centers that performed biopsies in Africa had access to immunofluorescence. Similar barriers were identified regarding treatment. It is not surprising therefore that low SDI countries have the highest mortality and DALY rates and highlights the need to plan for increasing capacity for diagnosis and treatment of glomerular diseases as the overall level of development rises.

How confident can we be in these measures of GN burden? The level of confidence in data from countries with universal health coverage and good data collection systems is likely to be high. However, the countries with the highest burden of GN in this study have sparse high-quality data. A large proportion of patients with kidney disease, including GN, do not present to the health care system in early stages (when accurate diagnosis can be made) in most low SDI countries with weak health systems.⁶ Ascribing the cause to GN when patients present for the first time in advanced stages of CKD is problematic.

In such situations, GBD derives disease burden estimates by modeling. Garbage codes are assigned to cases where the information is insufficient to identify the true cause of death.⁷ These garbage codes are then redistributed according to an algorithm to the GBD causes of death. The larger the proportion of garbage codes, the more biased the mortality pattern. In general, CKD estimates have a relatively high frequency of garbage codes, limiting the utility of death statistics. The proportion of deaths assigned garbage codes is one of the parameters used to assess the quality of cause of death data. Hu et al. have not provided this information.

One of the goals of GBD study is to provide visibility over disease trends and evaluate the impact of interventions over time. The previous GBD analysis had showed a higher age-standardized GN incidence and prevalence rate in 2016 compared with the current study.² Given the uptrending incidence and prevalence of GN globally, the lower rates in the current report are counterintuitive. Moreover, in contrast to GBD 2016, which only had data on CKD stages 3–5, the GBD 2019 includes all stages of CKD, which should also have led to an increase, as a majority of GN is not associated with reduced GFR, especially in early stages. Also not fully explained is the downward trend in prevalence rates in higher SDI countries. One would expect that easy access to the health care system will allow accurate diagnosis and appropriate treatment of GN, which would increase the lifespan and ergo the prevalence of GN. Better understanding of data sources and modeling methodologies will help explain the discrepancies.

Notwithstanding these uncertainties, refinement of the GBD approach should allow development of greater confidence in the data, allowing its use for planning public health interventions and measuring their impact on the GN burden. A key task for the global nephrology community is to improve the evidence that goes into the GBD database and engage with the GBD team to standardize modeling approaches. Many countries and groups are setting up GN registries. While their primary purpose is to improve clinical trial efficiency, they will also help with more accurate assessment of disease burden.

The drivers of development of GN are likely to be different worldwide. Most GN cases have an autoimmune basis. The triggers of autoimmunity, however, are likely to be different in different parts of the world. For example, the role of infections is likely to be disproportionately greater in the developing countries. The global GN burden has probably gotten worse since 2019 due to the coronavirus 2019 (COVID-19) pandemic. GN has been associated with the COVID-19 vaccine and COVID-19 infection.⁸ Furthermore, there were major interruptions in the care for patients with GN during the pandemic, which would have worsened their outcomes. Emerging risk factors, such as air pollution, are likely to further increase GN burden.⁹ The increased burden in the elderly deserves further study. This could be related to increasing burden of GN related to cancers or plasma cell dyscrasia, or more people with GN surviving to old age due to access to effective therapies. On the other hand, with a greater understanding of disease pathophysiology, more targeted therapies for treatment of GN are being developed. This is good news for patients but brings the risk of further increase in the global inequity of access.

What can we do to improve our understanding and reduce this rising and unequal burden of GN? Setting up systems for more accurate ascertainment of disease burden is the first step. Countries with high disease burden might want to learn from the example of Japan, Taiwan, and Korea, who have programs for screening of schoolchildren for GN, probably due to the high incidence of IgA nephropathy.¹⁰ There is an urgent need to remove the current barriers in diagnosis and management of GN in low SDI countries. This needs to happen in the context of an overall strengthening of health care systems, including increased funding, and mechanisms for better documentation and reporting.

Disclosures

V. Jha reports grant funding from Baxter Healthcare, Biocon, and GSK; honoraria from AstraZeneca, Baxter Healthcare, Bayer, Boeringer Ingelheim, NephroPlus, and Zydus Cadilla—all paid to the organization; an advisory or leadership role for NephroPlus; and speakers bureau for AstraZeneca and Baxter Healthcare. The remaining author has nothing to disclose.

Funding

None.

Author Contributions

V. Jha conceptualized the paper and provided supervision; B. Bose wrote the original draft; and V. Jha reviewed and edited the manuscript. Both authors approved the final version.