Introduction

Calcium is an essential mineral for various physiological activities of cells and organs, particularly the skeletal system [1]. Approximately 99% of body calcium is stored in the bone, contributing to bone rigidity and strength. In the clinic, calcium supplementation is an important therapy for the treatment of osteoporosis, an endocrine disease caused by over-activated osteoclasts [2, 3]. Calcium is also important for the physiological function of many other organs and systems, and inadequate calcium intake is associated with a number of diseases, including osteoporosis and fragile fractures [4], cardiovascular diseases [5], body fat mass [6], and colorectal cancer [7]. Diet intake is the major source of calcium in human body, but the amount of calcium intake may be inadequate due to income status, dietary habits and food diversities [8], leading to low-calcium related disease and health burden.

An early study has estimated the global risk of dietary calcium deficiency in 2011, showing that about 3.5 billion people were at risk of calcium deficiency due to inadequate dietary supply, despite a decreasing trend of exposure to such risk [9]. However, there is a lack of related data for each country, and the most recent prevalence of diet low in calcium (DLC) and the temporal trends in the past decades remains unknown. Besides, despite a well acknowledged association between calcium deficiency and the development of diseases, studies reporting the disease burden associated with DLC are limited and the health consequences of DLC are poorly understood [10]. In addition, there is a lack of data regarding the intake level of dietary calcium in many developing countries, which makes it hard to evaluate the health burden related to DLC in these countries.

Providing a comprehensive understanding about the prevalence and health burden related to DLC will help to gain more public attention to the intake of calcium, and will help to provide references for policy making and nutrition management practice. This will particularly contribute to reducing DLC-related health burden in low-to-middle income countries. In this study, we used data from the Global Burden of Disease study 2019 (GBD 2019) to report the global, regional, and national prevalence and burden of DLC by year, age, sex, and sociodemographic index (SDI), aiming to provide a clearer picture about the epidemiology and health burden of DLC.

Materials and methods

Study design and data source

This is a cross-sectional study about the prevalence and disease burden of DLC, using data from the Global Burden of Disease study 2019 (GBD 2019). Led by the the Institute for Health Metrics and Evaluation (IHME), GBD 2019 is a global disease epidemiology and burden study, estimating the incidence, prevalence, and disease burden of 369 diseases and injuries, and 87 risk factors at the global level and in 204 countries and territories, by location, year, age, sex, and socio-demographic index (SDI) [11]. The GBD study rounds evaluated the global disease burden since 1990 and has been updated to GBD 2019 since several iterations. It is one of the most comprehensive disease burden studies and provides the most up-to-date information for policy making and medical practice.

Definition of DLC and summary exposure value (SEV)

The 87 risk factors in GBD 2019 were categorized into three groups: (1) occupational and environmental risks; (2) behavior risk factors; and (3) metabolic risk factors [12]. In GBD 2019, DLC was considered as a kind of dietary risk factor, defined as the average daily consumption of calcium less than 1.06–1.1 g per day from all sources, including milk, yogurt, and cheese [12]. The raw data about daily calcium intake came from multiple sources, including nationally and sub-nationally representative nutrition surveys, household budget surveys, accounts of national sales from the Euromonitor, and availability data from the United Nations FAO Supply and Utilization Accounts (SUA) [13]. For a certain population, the prevalence of each risk factor was measured as summary exposure value (SEV) [12]. It is a relative risk weighted prevalence of exposure value to each disease-causing risk factors. The relative risk, obtained from published and unpublished primary studies or in secondary summarizing studies, indicates the risk of developing a disease due to exposure to a risk factor, or the risk of mortality or morbidity due to a specific cause. SEV ranges from 0 to 100%, with 0 indicating the population is at no excess risk factor, while 100% is at the highest level of risk factor (DLC in the current study). Increasing value of SEV indicated that the prevalence of a risk factor in the population increased, and vice versa. In this study, the prevalence of DLC was estimated for 204 countries, 21 GBD regions and at the global level.

The disease burden due to exposure to DLC

The exposure to a risk factor will not directly cause health outcomes, such as disabilities and mortalities. The burden associated with a risk factor is caused by the risk-factor-induced diseases. In GBD 2019, each risk factor can cause one or more diseases, and each risk factor to disease association is defined as a risk-outcome pair. The risk factor to a certain disease is determined by convincing or probable evidence, according to the World Cancer Research Fund (WCRF) grade [14]. In the WCRF grades, convincing evidence was determined from substantial evidentiary studies, including randomised controlled trials of sufficient size, duration, and quality that show consistent effects. These studies revealed biologically plausible associations between exposure and disease in different populations to enhance the evidence level. Probable evidence is similarly based on epidemiological studies with consistent associations between exposure and disease, but these studies have shortcomings, such as insufficient available trials or prospective observational studies.

According to the methodology and risk-outcome pairs in GBD 2019, DLC only has colorectal cancer as the outcome, despite that calcium deficiency can lead to a wide range of disease, especially osteoporosis and cardiovascular diseases [15, 16]. Thus, the disease burden associated with DLC was estimated based on its contribution to the development of colorectal cancer. The disease burden was measured as disability adjusted life year (DALY), which is the sum of years lost due to premature death and years lived with disability due to specific disease [17].

Sociodemographic index (SDI)

Disease outcomes may be influenced by many sociodemographic factors, including education and household income [18, 19]. GBD 2019 used to comprehensively represent the influence of sociodemographic factors on disease outcomes. SDI was calculated as the geometric mean of 0 to 1 indices of mean education for those 15 years and old and older (EDU15 + ), total fertility rate (TFR) for those younger than 25 years old (TFU25), and lag−distributed income (LDI) per capita [20]. SDI was further multiplied by 100 and ranged from 1 to 100 in GBD 2019. Country with higher SDI indicates people would have better outcome from a certain disease in that country.

Statistical analysis

In GBD 2019, all the final prevalence (SEV) and disease burden (DALY) were estimates, rather than actual data. These estimated data were produced by using the Bayesian method, which uses prior information of a specific population parameter with evidence from information contained in a sample to guide the statistical inference process. The raw data to produce the Bayesian-based estimates were collected from countries with available data, and the Bayesian method was incorporated into the DisMod-MR 2.1 tool in GBD 2019. The detailed method to produce the final estimates from the raw data has been thoroughly introduced in the parent GBD risk factor study [12], and summarized in recent studies of other GBD risk factor or diseases [13, 20, 21]. We also provided the methodology specific to DLC in the supplemental information file.

We reported the age standardized rates (ASR) of SEV and DALY, along with the 95% uncertainty interval (UI) produced in the Bayesian modeling tool. Spearman’s rank order correlation was additionally used to analyze the association between SEV and SDI. A correlation coefficient (rho) greater than zero with significant P value less than 0.05 indicates positive correlation, and vice versa. On the contrary, we used estimated annual percentage change (EAPC) and its 95% confidence interval (95% CI) to reflect the trends from 1990 to 2019, which has been widely used in other GBD studies [22,23,24,25,26,27]. EAPC was calculated based on a linear regression of the calendar year and the ASR: y = a + bx + e. In this model, y indicates log10 (ASR), while x is the calendar year; b is the regression coefficient, and EAPC was calculated as EAPC = 100 × (10^b −1). The difference between 95% CI and 95% UI can be seen in detail elsewhere [21]. If the EAPC and its lower limit were above zero, the trends were considered to be increasing, and vice versa. All data analysis and visualization were performed in the R software (Version 4.4.3)

Results

The prevalence of DLC

In GBD 2019, SEV, ranging from 15 to 100%, was used to indicated the weighted prevalence of specific risk factor. Higher value of SEV was indicative of higher prevalence. In 2019, the global age standardized SEV was estimated to be 46% (95% UI, 35.9–60.3%), with a decreasing trend from 1990 to 2019 (EAPC, −0.47; 95% CI, −0.5 to −0.43) (Table 1). Central Sub-Saharan Africa, Southeast Asia, and Oceania were the leading three countries in DLC prevalence, reaching an age standardized SEV rate of 88.1 (95% UI, 82.3–94.6), 83.4 (95% UI, 75.3–92.7), and 79.9 (95% UI, 71–90.6), respectively. In all regions and at the global level, the SEV in male people was higher than that in female people (Table 1). Oceania was the only regions that shows increasing trends of SEV for male people (EAPC, 0.14; 95% CI, 0.11–0.18), female people (0.17; 95% CI, 0.13–0.21), and both sexes (EAPC, 0.16; 95% CI, 0.12–0.19).

Table 1 The summary exposure value (%) of diet in low calcium and the 95% UI in the global and 21 GBD regions for males, females, and both sexes in 2019 and the temporal trends from 1990 to 2019.
Full size table

Countries with high prevalence of DLC for both sexes in 2019 were mainly located in Africa and Southeast Asia (Fig. 1A). Among the 204 countries and territories, Zambia leaded other countries in age standardized SEV rate (96; 95% UI, 95–97), followed by Liberia, Zimbabwe, and Democratic Republic of the Congo, mainly African countries (Supplementary Table 1). Most countries demonstrated decreasing trends of DLC prevalence, while only 26 countries showing opposite trends, with the highest increase being seen in United Arab Emirates (EAPC, 2.03; 95% UI, 1.86–2.21), followed by New Zealand, Japan, France (Supplementary Table 1), and many other countries in Europe (Fig. 1B). The geographical distribution of SEV in male and female people showed the same pattern, as indicated in Supplementary Fig. 1. In conclusion, these data suggested that the prevalence of DLC has decreased worldwide from 1990 to 2019, but increased prevalence was also seen in many countries, such as United Arab Emirates, New Zealand, France, Japan.

Fig. 1: The prevalence of diet low in calcium.
figure 1

A Age standardized SEV of diet low in calcium for both sexes in 204 countries and territories in 2019. SEV indicates the prevalence of diet low in calcium. Higher SEV value indicates higher prevalence in a country. B The temporal trends of the prevalence from 1990 to 2019 in each country, represented by EAPC. EAPC greater than zero indicates increasing trends, and vice versa. SEV summary exposure value, EAPC estimated annual percentage change.

Full size image

The health burden attributed to DLC

Due to the organization of risk-outcome pairs in GBD 2019, the disease burden associated with DLC was estimated only for DLC-induced colorectal cancer. Globally, the DALYs associated with LDC-colorectal cancer were estimated to be 3.14 million (95% UI, 2.25–4.26 million) in 2019, with an age standardized rate of 38.2 (95% UI, 27.2–51.8) per 100,000 (Table 2). At the global level, the burden of DLC-induced colorectal cancer accounted for a percentage of 12.91% (95% UI, 9.31%–17.46%) of all colorectal cancer related burden for both sexes. This percentage was higher in male people (13.55; 95% UI, 9.83–18.23) than that in female people (12.05%; 95% UI, 8.68–16.38). The global age standardized DALY rate (ASDR) due to DLC has remained unchanged during this period (EAPC, −0.03; 95% CI, −0.12 to 0.07), but have increased in many regions, including Sub-Saharan Africa regions, Latin America, South and Southeast Asia, and Oceania (Table 2). The ASDR in males was higher in all regions than that in females (Table 2).

Table 2 the number and age standardized DALY rate of diet low in calcium in the global and 21 GBD regions for males, females, and both sexes in 2019 and the estimated annual percentage change from 1990 to 2019.
Full size table

At the country level, Viet Nam and Palestine had an ASDR exceeding 100 per 100,000 (Supplementary table 1), being the top two countries in ASDR worldwide. Other countries with high ASDR included Philippines, Seychelles, Malaysia, Indonesia, Cambodia, Zambia, and Laos, all located in Southeast Asia or Southern Africa (Fig. 2A). The ASDR in male and female people demonstrated similar geographical pattern in different countries (Supplementary Fig. 2). Notably, increasing trends of ASDR from 1990 to 2019 have been observed in over 80 of the 204 countries, mainly distributed in Asia, Africa, and South America. The highest increase was seen in Lesotho (EAPC, 2.62; 95% CI, 2.38–2.87), followed by Viet Nam, Mozambique, and Bulgaria. A geographical feature was that these countries with high increase of ASDR tended to be located at tropical and subtropical regions. In conclusion, the disease burden associated with DLC has remained stable in the past thirty years worldwide, but many countries in Asia, Africa, and South America have showed increased burden due to DLC.

Fig. 2: The disease burden associated with diet low in calcium.
figure 2

A The age standardized DALY rate per 100,000 of diet low in calcium for both sexes in 204 countries and territories in 2019. DALY indicates the disease burden associated with diet in low calcium; Due to the GBD study risk-outcome pairs, the ASDR here is shown only for those due to low calcium induced colorectal cancer. B The temporal trends of the age standardized DALY rate from 1990 to 2019 in each country, represented by EAPC. EAPC greater than zero indicates increasing trends, and vice versa. DALY disease adjusted life year, ASDR age standardized DALY rate, EAPC estimated annual percentage change.

Full size image

The prevalence and burden of DLC by SDI, age, and sex

For all SDI regions, the prevalence gradually decreased, revealed by the age standardized SEV rates from 1990 to 2019 (Fig. 3A and Supplementary Fig. 3). However, the ASDR for low-SDI and low-middle-SDI regions showed increasing trends during this period, indicating increased health burden due to DLC (Fig. 3B and Supplementary Fig. 4). In middle-SDI regions, the ASDR increased in male people, decreased in female people, and the total rate for both sexes increased. At the country level, Spearman analysis revealed a strong positive correlation between country SDI and the age standardized SEV value (rho = −0.823; P < 0.001), as well as a mild correlation between country SDI and ASDR of DLC (rho = −0.433; P < 0.001). These correlations remained significant when separately analyzed in male and female people (Supplementary Figs 5, 6), and the correlation between SDI and ASDR was stronger in female people than that in male people (Supplementary Fig. 6). These results indicated that low-SDI countries tended to have high prevalence and disease burden due to DLC.

Fig. 3: The prevalence and disease burden of diet low in calcium by SDI.
figure 3

A The age standardized SEV rate in each from 1990 to 2019 in different SDI regions. SEV indicates the prevalence of diet low in calcium. Higher SEV value indicates higher prevalence in a country. B the age standardized DALY rate for different SDI regions from 1990 to 2019. C The correlation of age standardized SEV in 2019 to SDI by Spearman rank order correlation analysis; rho indicates of the coefficient of Spearman analysis. D The correlation between age standardized DALY rate in 2019 to SDI. SEV summary exposure value, DALY disease adjusted life year, ASDR age standardized DALY rate, SDI sociodemographic index.

Full size image

The prevalence and burden of DLC by sex and age at the global level from 1990 to 2019 are shown in Fig. 4. Both male and female people showed decreasing prevalence during this period (Fig. 4A). However, a sex difference was observed for the ASDR of DLC. Male people had increasing trends of ASDR from 1990 to 2019 (Fig. 4B). The EAPC for the ASDR in male people was 0.27 (95% CI, 0.15–0.40) (Table 2), indicating an increasing burden by year for male people. On the contrary, the health burden decreased in female people in the past thirty decades (Fig. 4B).

Fig. 4: The prevalence and disease burden of diet low in calcium by sex and age.
figure 4

A The global age standardized SEV rate for male and female people from 1990 to 2019. SEV indicates the prevalence of diet low in calcium. Higher SEV value indicates higher prevalence in a country. B The global age standardized DALY rate for male and female people from 1990 to 2019. C Global SEV of diet low in calcium in 2019 in different age groups. D Global DALY rates in 2019 in different age groups. SEV summary exposure value, DALY disease adjusted life year.

Full size image

Data of the SEV and DALY rates were available for people older than 30 years. At all ages, the SEV was higher in male people than that in female people. SEV decreased with age until 70 years old, then slightly increased until 85 years, and decreased again in older age (Fig. 4C). Unlike SEV, the DALY rates per 100,000 increased with age and peaked at about 90 years for male, female, and both sexes, without an obvious decrease in older age (Fig. 4D), indicating increasing health burden due to DLC with age.

Discussion

In this study, we reported the prevalence and health burden due to DLC in 2019 and the temporal trends from 1990 to 2019 at the global, regional, and country level, by age, sex, and SDI. Globally, the prevalence of DLC has decreased year by year, while the health burden has remained unchanged. However, at the national level, many countries in Southeast Asia and Africa have increasing burden during this period. Sociodemographic levels have essential influence on the prevalence and burden of DLC, with low-SDI countries having significantly higher prevalence and burden than those high-SDI countries. Male people have higher SEV and DALY rates in all years and at all ages than male people, and the burden due to DLC in male people have increased in the past thirty years.

The results in the current study indicated substantial variability in the prevalence and burden of low dietary calcium intake between high- and low-income countries, which are consistent with previous studies revealing higher calcium intake in high-income countries and lower intake in and low-income countries [28]. In countries in Asia, Africa, and South America, the average dietary intake of calcium ranges from 400 to 700 mg/day [29]. Another study reported that average dietary calcium in Southeast Asia and some countries in Africa was even as low as 200 mg/day, but was about 600–800 mg/day in high-income countries [30]. Such discrepancies between high- and low-income countries are explained by many factors, including but not limited to the access to calcium-rich foods and dietary habits [31]. Low availability to dairy products is a major risk for low dietary calcium intake, especially in low-income countries in Africa [8]. In Southeast Asia, the local complementary foods are of low micronutrient density and the lack of food diversity renders it hard to supply the adequate amount of calcium as required [32, 33]. Low diary consumption and inappropriate marketing of ultra-processed, complementary foods, and beverages that are nutritionally inadequate also underlies calcium deficiency in such region [8, 32].

On the contrary, in high-income countries, many commercial food products, such as fruit juices and wheat flour, are usually fortified with calcium [34], which provides more calcium sources to their residents. They can also benefit from the diversity of food source regarding calcium intake. As a result, people in high-come countries are less likely to experience the risk of low calcium intake from diet. However, the results showed that some countries from developed regions, including France, Spain, and many other countries in Europe, shoed increasing trends in the prevalence of DLC. The risk of calcium deficiency in high-income countries are seen in many subgroups of people, including adolescents, post-menopause women, athletic women and those maintaining a vegan diet [8, 35]. Compared to common people, calcium intake status of these people at higher risk should be more accurately monitored to prevent related health burden.

In all regions and years, the health burden due to low diet intake of calcium was consistently lower in female people than that in male people. This may be explained by higher calcium intake per day in women in some countries [36, 37]. For example, a study from Spain reported that the dietary intake of calcium was 872.5 mg/day for female people and 764.5 mg/day for men [37]. When calcium supplements were included, the intake reached 1057.6 mg/day for female people, which was significantly higher than that for male people (787.5 mg/day). Besides, another study from Lebanon indicated that dairy products, including milk and milk derivatives, accounted for a higher percentage in the total diary intake in female people [38]. However, some countries have also reported higher calcium intake per day in male people [39, 40], suggesting that divergent daily calcium intake is not the only determinant to underlie the differences.

In this study, we found that the prevalence of DLC decreased with age. This decrease may be explained by the fact that young people are more likely to have unbalanced dietary habits, and old people tend to pay more attention to maintaining a balanced diet due to decreased health status. In contrast to the prevalence, the DALY rates of DLC increased with age, indicating that old people have higher disease burden due to DLC. The discrepancy between the prevalence and burden with age may be explained by the measurements of the disease burden. In GBD 2019, the burden associated with DLC was estimated only for those caused by DLC-related colon-rectal cancer, which is an age-related disease and showed higher incidence rate and disease burden in old people [41]. In addition to colorectal cancer, DLC predisposes people to a wide range of diseases, including cardiovascular diseases [42], low bone mineral density and osteoporosis [43]. It can be inferred that when all the DLC-related diseases were included for the calculation of the disease burden, the results wiould differ from that in the current study.

Recommended daily calcium intake differs by age groups throughout life. Due to the tight connection between calcium and bone health, the requirements for calcium have been commonly evaluated according to bone mass status, which is planned to support the increasing bone mass during adolescent and prevent bone loss and related fracture outcomes in post-menopause women [44, 45]. Another stage that requires additional calcium intake is pregnancy or lactation in women, which can help to prevent pregnancy disorders [46] and may be associated with offspring bone and cardiovascular health [47, 48]. In adolescents aged 14–18 years, the Institute of Medicine (IOM) recommends that the daily calcium should reach 1300 mg, and the intake ranges from 700 to 1200 mg per day for people aged 19+ years [46]. For pregnant women, the intake increases again to 1300 mg per day. However, in many countries, the daily intake is below the recommended value, especially in low-middle income countries and in pregnant women [49, 50]. Calcium supplementary is an effective measure to fill the gap in calcium intake. Foods rich in calcium are mainly dietary products, such as yogurt, cheese, and milk, especially hard cheese that contains high levels of calcium. Calcium-fortified commercial products and calcium tablets are another important and cost-effective sources for calcium supplementation [51], particularly in low-income-countries where people have limited availability to diverse kinds of diary-derived calcium. The epidemiology and burden due to DLC differs with counties. It’s important for low-diet-calcium-intake countries to recognize their situations and take related measures. Public education and government-advocated nutrition program will help to reduce the disease burden due to low diet-intake of calcium.

This study has some limitations. Firstly, DLC is a risk factor for a number of diseases. However, in GBD 2019, the disease burden associated with DLC was only predicted for its related colorectal cancer. Thus, the health burden due to a wide septum of diseases associated with DLC is unclear, resulting in a lack of total disease burden attributed to such dietary risk. However, the prevalence was based on an entire estimation, regardless of diseases. Thus, the prevalence may be of more reference value to estimate the entire disease burden. Secondly, as data used in this study were collected from people aged 30 years and older, the prevalence and burden in people younger than 30 years, especially adolescents, are not clear and require further study. The third limitation results in the inherent GBD limitations, especially the quality of raw date [11]. Due to the differences in the availability to raw data, the estimates may be of different accuracy among countries, reflected by the wide range of 95% UI of some estimates. For countries with sparse raw data, the estimates were produced largely depending on Bayes-based statistical models, leading to bias from the actual prevalence and burden. This limitation has been improved round by round for the GBD studies. The methodology and the raw data of GBD 2019 have improved greatly compared to previous rounds. However, future efforts are still in need to improve the data quality to provide more accurate and more up-to-data estimation of the disease burden.

Conclusion

The prevalence of DLC has decreased worldwide and in most countries, but the disease burden of DLC induced colorectal cancer has increased in over 80 countries worldwide. Countries with low sociodemographic level and male people are more prone to the risk of DLC and are more likely to experience calcium related disease burden. Male and old people have higher prevalence and disability rate due to DLC. Effective measures are in need to address the health challenges from DLC.