Global Burden of Vitamin A Deficiency in 204 Countries and Territories from 1990–2019

Vitamin A deficiency (VAD) is one of the important public health issues worldwide. However, a detailed understanding of the incidence and disability-adjusted life years (DALYs) due to VAD in recent years is lacking. We aimed to estimate the incidence and DALYs of VAD at global, regional, and national levels in terms of sex, age, and socio-demographic index (SDI). Using data from the 2019 Global Burden of Disease (GBD) study, the estimated annual percentage change (EAPC) was measured to assess trends in the age-standardized incidence and DALY rates from 1990 to 2019. The global age-standardized incidence and DALY rates of VAD decreased with an EAPC of −3.11% (95% confidence interval (CI): −3.24% to −2.94%) and −2.18% (95% CI: −2.38% to −1.93%), respectively. The age-standardized incidence and DALY rates decreased least in low-SDI regions, which had the highest age-standardized incidence and DALY rates of all SDI regions. Sub-Saharan Africa, especially central sub-Saharan Africa, had the highest age-standardized incidence and DALY rates in 2019. At the national level, Somalia and Niger had the highest age-standardized incidence and DALY rates. The age-standardized incidence and DALY rates were higher in males than in females. Younger children, especially those aged < 5 years in low-SDI regions, had a higher VAD burden than other age groups. Although the global burden of VAD has decreased, future work should aim to improve the prevention and treatment strategies for VAD, particularly in children aged < 5 years in countries and territories with low SDI values, such as sub-Saharan Africa.


Introduction
Vitamin A deficiency (VAD), one of the four major nutritional deficiencies worldwide [1], can cause growth and development deficits in children, loss of vision, and increase the risk of infection [2,3]. Furthermore, VAD is a potential risk factor for cognitive impairment and mental illness [4,5]. Therefore, VAD is regarded as the second largest risk factor for the global disease burden, and the life span of populations with disability due to VAD is a crucial quantitative index of burden of malnutrition [6].
According to a report from the World Health Organization (WHO), 190 million preschool children and 19 million pregnant women were exposed to VAD globally [7].
Although the WHO has made great efforts in reducing the epidemic of VAD, due to major differences in the economy, medical care, food carrier, and availability of other nutrients [8], the burden of VAD is unbalanced among population in different regions and is more prevalent in areas with lower socio-economic levels [9,10]. Therefore, an assessment of VAD burden at the regional and national levels and identification of high-risk groups could help to identify the highly endemic areas and populations, which could in turn beneficial to guide effective decision making and resource allocation [11].
To our best knowledge, there are only a limited number of analyses on VAD burden in specific populations and limited regions based on the previous Global Burden of Disease (GBD) study [12,13]. Moreover, the previous GBD analyses have focused more on loss of vision due to VAD rather than VAD itself [2,14]. Currently, a detailed report on the GBD dataset-based analysis of the global VAD burden is lacking. Additionally, previous studies to assess the epidemic of VAD have mainly centered on investigating the prevalence of this condition. However, the annual incidence rate, which is defined as the frequency of new cases per year, expresses more accurately the epidemiological changes of VAD.
Therefore, we used VAD data from the GBD 2019 database to determine temporal trends in the incidence and disability-adjusted life-years (DALYs) of VAD at the global, regional, and national levels and by sex, age, and socio-demographic index (SDI) to systematically summarize the global burden and health development status of VAD.

Overview and Definitions
The methodology used to assess the GBD is continuously improving, and GBD data are updated annually. The 2019 GBD contains annual data of 369 human diseases and 87 risk factors from 204 countries and territories worldwide [15]. We estimated VAD burden every year from 1990 to 2019 worldwide included in 2019 GBD. The study followed the recommendations of the Guidelines for Accurate and Transparent Health Estimates Reporting and was approved by the Medical Ethics Committee of Hwa Mei Hospital.

Data Sources
Data on the annual incidence, DALYs, and age-standardized incidence and DALY rates of VAD were derived from the 2019 GBD dataset. The Global Health Data Exchange query tool, which is a data source tool online (http://ghdx.healthdata.org/gbd-results-tool, accessed on 10 January 2022), was used to assess health loss from VAD using available epidemiological data. The SDI, which is a comprehensive index that includes per capita gross domestic product, average years of education, and fertility rate to reflect social development, is an important covariate for evaluating disease burden and health development in a certain region. The SDI divides 204 countries and territories into five SDI (high, high-middle, middle, low-middle, and low) quintiles [15]. DisMod-MR 2.1, a Bayesian meta-regression modeling tool, was applied to calculate the data and to determine the 95% uncertainty intervals (UIs) [15]. The DALYs due to VAD were calculated as the sum of years of life lost due to disability and years of life lived with disability. The flowchart for VAD analysis in GBD is shown in Supplemental Figure S1.

Statistical Analysis
Assuming that the natural logarithm of the age-standardized rate has a linear relationship with time, the estimated annual percentage changes (EAPCs) in the age-standardized rates of VAD were calculated within a specific time frame using the following formula: Y represents ln (age-standardized rate), X is calendar year, and ε is the error term. β describes the positive or negative age-standardized rate trend. Therefore, Additionally, 95% confidence intervals (CIs) were determined by the linear model. When the EAPC and CI lower boundary are positive, the age-standardized rate will show an upward trend. Conversely, when the EAPC and CI upper boundary are negative, the age-standardized rate will show a downward trend. Otherwise, the age-standardized rate will show a stable trend [16]. We also determined the correlations between the EAPC and age-standardized rates and between the EAPC and SDI by using Gaussian process regression and Pearson's correlation coefficient (ρ).
All statistical analyses were carried out using R software (version 3.5.1, the R foundation for statistical computing, Vienna, Austria). A p-value of <0.05 was regarded as statistically significant.
In 2019, the male-to-female ratio of VAD incidence peaked in the 45-49-year age group globally, in the 0-5-year age group in high-SDI regions, in the 10-14-year age group in high-middle and middle-SDI regions, and in the 45-49-year age group in low-middleand low-SDI regions (Supplemental Figure S4). The incidence of VAD decreased with age globally. The incidence was much higher in children aged < 5 years than in other age groups in all SDI regions (Supplemental Table S5 and Figure S5).
countries and territories are shown in Supplemental Figures S10A-S12A, and age-standardized incidence rates of VAD in 204 countries and territories in 2019 are shown in Supplemental Figures S13A-S15A.   Figure S2B) Figure 1B, and Supplemental Figure S3B). The male-to-female DALY rate of VAD peaked in the 0-5-year age group globally; in the ≥60-year age group in high-and high-middle-SDI regions; and in the 0-5-year age group in middle-, low-middle-, and low-SDI regions (Supplemental Figure S16). Globally, the DALY rate decreased with age. The highest DALY rate was observed among young people aged < 14 years, especially children aged < 5 years, globally, in all SDI regions (Supplemental Table S6 and Figure S17).

Analysis of Changes in DALYs Due to VAD
The age-standardized DALY rate of VAD was much higher in regions with low SDI values than in regions with other SDI values. In 2019, the highest age-standardized DALY rate of VAD was observed in the low-SDI group (38.02 per 100,000; 95% UI: 26.12-52.78), followed by the low-middle-SDI group (19.32 per 100,000; 95% UI: 12.93-27.45). The high-SDI group showed the lowest age-standardized DALY rate (0.29 per 100,000; 95% UI: 0.18-0.44). From 1990 to 2019, the age-standardized DALY rate of VAD declined in all SDI regions. The lowest decrease occurred in low-SDI regions where the EAPC was greatest (−2.30%, 95% CI: −2.51% to −2.07%), whereas the greatest decrease occurred in the high-SDI group with the smallest EAPC (−3.49%, 95% CI: −4.03% to −2.92%) (Table 2, Figure 1B, Supplemental Figure S3B).

Discussion
We systematically analyzed temporal and spatial trends in the incidence and DALY rates of VAD at the global, regional, and national levels by sex, age, and SDI from 1990 to 2019. The global VAD incident number (−44.19%), age-standardized incidence rate (−59.85%), DALY number (−40.15%), and age-standardized DALY rate (−47.07%) declined from 1990 to 2019 due partly to continuous improvement in the global economy and the quality of health care as well as a wide range of food fortification and supplementation strategies.
However, the burden of VAD remained high globally in 2019. The main causes of VAD include an insufficient intake of foods rich in vitamin A, the malabsorption of vitamin A, and a loss of vitamin A due to disease [7]. Although the trend of malnutrition has changed from undernourishment to excess energy intake in recent years, the consumption of micronutrient-rich foods remains far from sufficient [17]. Vitamin A supplementation (VAS) has been implemented in many regions and countries, but the complete implementation of periodic high-dose interventions is difficult in some countries with large populations; thus, the rate of coverage remains low [18]. Additionally, in some countries with poor sanitation, epidemics of infectious diseases have resulted in the loss of vitamin A [19] and thus increased the global burden of VAD.
The global burden of VAD differs hugely among regions with different SDIs, reflecting large discrepancies in medical facilities and health care. Low-SDI regions had high incidence and DALY rates at baseline and the slowest decline in age-standardized rates of VAD from 1990 to 2019, which caused the VAD burden to be much higher than that in regions with low-middle, middle, high-middle, and high SDI values in 2019; in fact, the age-standardized incidence rate was higher by almost a factor of 33, and the age-standardized DALY rate was higher by a factor of 131 in low-SDI regions than in high-SDI regions. Poverty in low-SDI regions is regarded as the main factor underlying the widespread occurrence of VAD [20]. The prevalence of VAD in low-income regions is high although the global mortality rate due to VAD has fallen sharply [21]. Another factor is low literacy [22]. Specifically, people with lower education levels may well suffer from VAD [23]. In regions with limited income and education, increasing the availability and utilization of vitamin A-rich foods through horticultural strategies might provide greater benefits than the large-scale adoption of VAS [24]. Implementing innovative social marketing strategies in addition to traditional nutrition education might be a more effective way to disseminate nutrition information [22]. At the regional level, in 2019, central, eastern, and western sub-Saharan Africa had the highest age-standardized incidence and DALY rates of VAD. These high rates are likely to be primarily due to the contributions of Somalia, Niger, and Mali, among other countries. In central sub-Saharan Africa in particular, the age-standardized rates of VAD declined slowly from 1990 to 2019. Falsified vitamin A (retinol) capsules with insufficient doses, severe degradation, and substandard specifications were widely circulated in Africa, the region with the most serious falsified medical products in the world [25]. It is imperative to crack down on the trafficking of counterfeit medicines. Orange-fleshed sweet potato, a variety of sweet potato rich in plant-based β-carotene, has been widely promoted in Africa for vitamin A fortification. However, the popularity and quality of orange-fleshed sweet potato remain issues that urgently need to be addressed [26]. It is necessary to increase the acceptance in the target population and improve the bioavailability of β-carotene by processing orange-fleshed sweet potato through local public sector programs and large retailers [27,28]. Additionally, research on other fortified staple foods, such as provitamin A carotenoid corn, needs to be promoted [29]. It is noted that properly preserving vitamin A drugs and food supplements, reducing food odors, increasing drug half-lives, and ensuring vitamin A activity are also important problems to be solved in sub-Saharan Africa, which has perennially high temperatures [30,31]. Failure to address these problems might also be related to widespread epidemics of malaria and acquired immunodeficiency syndrome in sub-Saharan Africa [32]. Furthermore, strengthening the coordination of interventions for overlapping epidemic diseases and VAD is an important direction in which to improve public health [19].
The WHO reported that the prevalence of VAD in Southeast Asia was as high as 50% [7]. We found that from 1990 to 2019, the age-standardized incidence and DALY rates of VAD in Southeast Asia decreased significantly, respectively ranking second and fourth among 21 regions, which reflects the great contribution of Southeast Asia in controlling the incident cases of VAD and thus reducing the global burden. Over the past 20 years, the increase times in incident cases and DALYs have been high, and the age-standardized rates have declined most slowly in Oceania, mainly owing to the situations in the Federated States of Micronesia, Kiribati, and the Northern Mariana Islands, which may be related to the dietary shift toward imported processed foods and lifestyle changes [33]. It is necessary to strengthen the awareness of VAD in these countries. Australia, as the most populous country in Oceania and having the lowest VAD burden worldwide, provides an important reference for these countries [34].
Children younger than 5 years of age, whose growth and development will affect the human capital and economic status of society in the future, are the main group susceptible to VAD [35]. Children with VAD are also more susceptible to respiratory infection, diarrhea, and parasitic infection. Repeated infections further reduce the absorption of vitamin A, resulting in a vicious cycle in children [36]. We found that the incidence and DALY rates of VAD declined with age globally, with the highest proportion of VAD cases being in children younger than 5 years of age in all SDI regions. The annual number of incident cases and DALYs of VAD in most parts of the world decreased in adolescents and children (<20 years of age) but increased in adults, which might be attributed to the global emphasis on the growth, development, and nutrition of adolescents and children as well as global aging. Furthermore, nutritional fortification might have a certain age effect [8].
The annual proportion of cases of VAD in low-SDI regions did not differ significantly between the age groups. With an increase in the SDI, the male-to-female ratio of the incidence of VAD was more biased toward the younger age group, whereas the maleto-female ratio of the DALY rate had the opposite bias. This might be because a greater awareness of health and availability of advanced diagnostic techniques in high-SDI regions resulted in a higher VAD detection rate in children and adolescents [37]. Meanwhile, increases in population aging and life expectancy bring increased risks of health problems related to VAD [38]. Regions with low SDI values contributed most to the global maleto-female ratio of the incidence and DALYs of VAD, indicating that the diagnosis of and intervention for VAD are imperative in regions with low SDI values, especially for children younger than 5 years of age. There might be additional children with undiagnosed or marginal VAD [39], which might have resulted in underestimation of the VAD burden in children. The use of VAS has been shown to significantly reduce VAD-related morbidity and mortality in children between 6 months and 5 years of age [18]. It is thus necessary to implement universal policies on VAS to prevent VAD in this age group.
Interestingly, the burden of VAD differed between males and females, with the agestandardized incidence and DALY rates being higher in males than in females. The male sex is considered a risk factor for VAD [40]. Additionally, the serum vitamin A concentration is negatively correlated with the occurrence of some male diseases, such as prostate cancer [41]. VAS has been implemented and combined with immunization programs in many regions and countries, and the interaction between VAS and vaccination is stronger in females [42]. This suggests that females have a greater awareness of micronutrient control, such as VAS and adjustment of the dietary structure, than males. Sex differences in the burden of VAD were more obvious in regions with low SDI levels. Male children might contribute to these figures more than female children because preferences for sons remains widespread in regions with low SDI values [43].
To the best of our information, this is the first report to describe the latest epidemiological patterns of VAD at the global, regional, and national levels and by sex, age, and SDI. The strength of this research is the systematic use of the latest GBD data from 2019 in assessing the incidence and DALY rates of VAD in various regions of the world from 1990 to 2019. The main limitation of this study is that owing to the potentially incomplete registration of VAD incident cases in some low-SDI regions, such as Africa and Latin America, during the GBD study period, the quality and accuracy of GBD data for such regions cannot be guaranteed, which might have affected the regional VAD burden analysis.

Conclusions
Despite the worldwide decrease in the age-standardized incidence and DALY rates of VAD from 1990 to 2019, the burden of VAD remains high in low-SDI regions, including most of sub-Saharan Africa and especially central sub-Saharan Africa. Globally, children younger than 5 years of age are most affected by VAD, and males are more susceptible than females to VAD. Therefore, attention should be paid to improving early prevention and treatment strategies for VAD in low-SDI regions, such as sub-Saharan Africa, especially in children younger than 5 years of age, to reduce the global burden of VAD.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/nu14050950/s1, Figure S1: The flowchart for VAD analysis in GBD; Figure S2: The incidence number and DALY number of VAD between 1990 and 2019 among sexes; Figure S3: The change trends of age-standardized rates of VAD among different SDI quintiles and sexes from 1990 to 2019; Figure S4: The ratio of male to female VAD incidence among different age groups in 2019; Figure S5: Distribution of different ages in VAD incidence; Figure S6: The correlation between EAPC and age-standardized rates of VAD and SDI in 2019; Figure S7: The age-standardized rates of VAD per 100,000 population among regions based on SDI in 2019; Figure S8: The incidence and DALY rates of VAD in different age groups; Figure S9: The proportion of different ages in VAD incidence number for both sexes by years; Figure S10: The global EAPC of VAD for both sexes in 204 countries; Figure S11: The global EAPC of VAD for female in 204 countries; Figure S12: The global EAPC of VAD for male in 204 countries; Figure S13: The global burden of VAD for both sexes in 204 countries; Figure S14: The global burden of VAD for female in 204 countries; Figure S15: The global burden of VAD for male in 204 countries; Figure S16: The ratio of male to female DALY rate among different age groups in 2019; Figure S17: Distribution of different ages in VAD DALY rate; Figure S18: The proportion of different ages in VAD DALY number for both sexes by years; Table S1: The top three and the bottom three regions of VAD incidence or DALY; Table S2: The top three and the bottom three countries of VAD incidence or DALY; Table S3: The incident cases and age-standardized incidence rate of VAD in 1990 and 2019, and its temporal trends from 1990 to 2019; Table S4: The DALY and age-standardized DALY rate of VAD in 1990 and 2019, and its temporal trends from 1990 to 2019; Table S5: Age distribution of incidence rate (per 100,000 people) for VAD in different countries in 2019; Table S6: Age distribution of DALY rate (per 100,000 people) for VAD in different countries in 2019.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki and approved by the Medical Ethics Committee of Hwa Mei Hospital.

Informed Consent Statement:
The institutions that conducted the surveys were responsible for obtaining informed consent from the participants. Data Availability Statement: Data are available in a public, open access repository. All of the data are publicly available. Data are available on request.