Long-term dynamic burden and attributable risk analysis of indirect maternal mortality over the past 30 years: an observational study and insights for perioperative management in the next decade

Abstract

Objective:

Reducing indirect maternal mortality (IMM) and promoting maternal health remain complex global challenges. This study evaluates 30-year trends in IMM burden and provides evidence for policy formulation.

Materials and methods:

Data on IMM disease burden in 195 countries/regions were extracted from the Global Burden of Disease (GBD) database. Trends in age-standardized death rate (ASDR) and disability-adjusted life year rate (AS-DALY) were analyzed, with subgroup analyses by age, region, and socioeconomic demographic index (SDI). Risk factor associations with IMM were also explored.

Results:

Over 30 years, ASDR and AS-DALY showed downward trends with estimated annual percentage changes (EAPCs) of −0.64 (95% CI: −0.87 to −0.41) and −3.04 (95% CI: −3.33 to −2.74), respectively. However, high-SDI regions exhibited a significant upward trend in ASDR (EAPC = 3.88, 95% CI: 3.31–4.45), while middle-SDI regions showed the steepest decline (EAPC = −3.04). Iron deficiency was an independent risk factor for IMM-related DALYs, with marked regional heterogeneity.

Conclusion:

Despite the reduction in the burden of IMM, low- and middle-income regions, as well as women of childbearing age between 25 and 34, still face high risks. Iron supplementation interventions and systematic optimization of medical resources are key focuses for improvement.

Introduction

Worldwide, indirect maternal mortality rate (IMM) is one of the most commonly used indicators for evaluating maternal mortality, and it is also an important indicator for measuring the level of economic development, cultural education, healthcare, and national health of a country or region^[1–3]. According to estimated data from the World Bank, the global maternal mortality rate (deaths caused by pregnancy or childbirth complications) has decreased by 40% from 328 per 100 000 in 2000 to 197 per 100 000 in 2023^[4,5]. However, maternal mortality lifetime risks exhibit significant cross-country variation.

According to recent research, it is estimated that by 2023, the IMM in low-income countries will reach 1/66, while in high-income countries it will be around 1/8000^[6–8]. However, among all regions, especially women in sub Saharan Africa face the highest lifetime risk (1/55), which is about 250 times higher than in Western Europe (1/14 337), leading to a sustained tightening of global aid funds and forcing many countries to cut critical services for maternal, newborn, and child health^[6,9,10]. Furthermore, despite notable progress in perinatal medicine, the 2023 World Health Organization (WHO) report reveals that around 295 000 women die annually during the perinatal period[11]. Among contributing factors, rising obstetric pathologies are critical. Preeclampsia, affecting 2 to 8% of pregnancies, can lead to severe complications, while the increasing incidence of the placenta accreta spectrum often causes life threatening hemorrhages^[12,13]. Addressing these pathologies is key to reducing maternal mortality and improving perinatal care. Given these disparities, reducing IMM and promoting maternal health and well-being remain complex healthcare burden issues that urgently need to be addressed in countries and regions around the world.

The Global Burden of Disease (GBD) database, led by the Institute for Health Metrics and Evaluation (IHME) at the University of Washington, is a public database for the study of global burden of disease, injury, and risk factors[14]. It covers health burden data from 459 diseases and injuries in 195 countries and regions since 1990. Fortunately, the database provides comprehensive data on IMM incidence rate, morbidity, mortality and disability adjusted life years (DALYs) for 195 countries and regions over the past 30 years^[14,15]. In view of this, we used the database to analyze the incidence rate, morbidity, mortality and DALYs related to IMM dynamics in the past 30 years, and studied the trend, and also discussed the relationship between incidence rate or mortality and social demographic index (SDI), in order to provide valuable reference for the prevention and management strategies of IMM.

Materials and methods

Observational study design

This investigation adopted a retrospective observational study approach. We scrutinized historical data spanning the past three decades (1990–2019) to discern the trends and patterns of IMM. The observational nature of this study permitted the collection of real world data without the confounding effects of experimental interventions, thereby providing an authentic portrayal of the disease burden.

Ethical considerations

Given that this study utilized publicly available data and did not involve direct patient interaction or the collection of new personal data, ethical approval from an institutional review board was not mandated. Nevertheless, all data were utilized in strict accordance with the terms and conditions set by the data providers, ensuring the confidentiality and proper handling of the information.

Data extraction

Data extraction was executed by a team of trained researchers utilizing standardized data collection forms. For the GBD data, we leveraged the online data visualization and extraction utility provided by the GBD research consortium. This tool enabled the extraction of data related to specific causes of death, age groupings (in 5-year intervals within the 15–49-year reproductive age bracket), and gender (female). In the case of national health statistics, we devised customized data extraction forms tailored to the available data fields in each country’s records. These forms incorporated fields for maternal death identification, diagnostic codes (standardized using the International Classification of Diseases, 10th Revision [ICD-10]), date of death, age, and other relevant demographic details. Our work has been reported in line with the STROCSS criteria^[16,17].

Data cleaning

Upon extraction, the data underwent a meticulous cleaning process. Duplicate records were identified and eliminated using unique identifiers, such as patient identification numbers (in the case of national data) or a combination of demographic and death related attributes. Missing values were addressed through multiple imputation methodologies. For categorical variables with missing entries, mode imputation was applied, whereas for continuous variables, predictive mean matching imputation was employed.

In the predictive mean matching imputation process for a variable X with missing values, an initial regression model was formulated:

X= β0 + β1X1 +⋯ + βpXp + ϵ

HIGHLIGHTS

• The study uses the 2019 Global Burden of Disease (GBD) study to analyze the global burden of indirect maternal mortality diseases, linking it with the sociodemographic index (SDI).

• Findings show higher disease burdens in low-SDI countries, as analyzed through health inequality indices.

• Iron deficiency and perioperative management are identified as the main factors increasing the disease burden across nations.

• Future projections using Bayesian age-period-cohort (BAPC) model indicate persistent geographic and temporal disparities in disease burden related to SDI levels through 2030.

• The study highlights the need for targeted prevention and treatment strategies for high-risk groups to reduce the global disease burden.

where X1,⋯, Xp are other variables within the dataset, βi are the regression coefficients, and ϵ represents the error term. Subsequently, for each missing value, the observed value in the dataset with the closest predicted value from this model was selected for imputation. Incorrect or inconsistent diagnostic codes were cross referenced against the ICD-10 classification manual, and corrections were made based on the available clinical information.

Temporal and geographical analysis

The data for this research were primarily sourced from the 2019 iteration of the GBD initiative. This extensive database offers systematic quantification of mortality, morbidity, and risk factors across 195 countries and territories, spanning the period from 1990 to 2019. The GBD study deploys a hierarchical Bayesian meta-regression framework known as DisMod-MR for estimating disease burden. The fundamental equation utilized by the DisMod-MR model to estimate the incidence rate (λ) is articulated as: ln(λijk) = αi + βj + γk + δijk

In this formula, αi denotes the influence of the i-th geographical location, βj represents the effect of the j-th age cohort, γk signifies the impact of the k-th time period, and δijk constitutes the random error component. The model assimilates data from diverse sources including vital registration systems, sample registration systems, verbal autopsy investigations, and surveillance programs.

For the evaluation of indirect maternal mortality, we extracted specific data on mortality frequencies, years of life prematurely lost (YLLs), years lived with disability (YLDs), and DALYs associated with indirect maternal causes. YLLs were computed via the formula:

YLLs = N× L

Here, N indicates the count of deaths, while L represents the standardized life expectancy at the age of demise. YLDs were determined using the equation:

YLDs = I× DW × D

where I is the number of incident cases, DW is the assigned disability weight, and D denotes the average duration of the disability state. DALYs were subsequently derived by aggregating YLLs and YLDs, i.e., DALYs = YLLs + YLDs. Indirect maternal causes were delineated according to WHO standards, encompassing conditions like anemia, infectious diseases, cardiovascular disorders, and mental health issues that are not directly linked to the pregnancy process but are aggravated during pregnancy or its management phases.

Statistical analysis

To analyze whether there are differences in the burden of IMM in countries/regions with different SDI, and the impact of SDI on the burden of disease. The study compared the differences in age-standardized death rate(ASDR) among different SDI groups. The formula for calculating the age-standardized rate (ASR) per 100 000 people is as follows:

$$\mathrm{A}\mathrm{S}\mathrm{R}=\frac{{\mathrm{\Sigma }}_{\mathrm{i}=1}^{A}aiWi}{{\mathrm{\Sigma }}_{\mathrm{i}=1}^A}\times 100,000$$

Among them, letter A represents the total number of age groups, i next to letter a represents the specific age rate of the i-th age group, and w is the proportion of the standard population corresponding to the i-th age group.

We use methods such as t-test or analysis of variance to compare, which divides countries/regions into five groups (high, high, medium, low, and low) based on SDI. In the study of regression analysis, the linear relationship between SDI and age-standardized rate was analyzed using least squares regression. For nonlinear relationships, a generalized additive model (GAM) is used to analyze the nonlinear relationship between SDI and age normalization rate. Additionally, we also evaluated the DALYs related to IMM, which refers to the total healthy life years lost from onset to death. The data processing and visualization involved in this study were completed using R software (version 4.3.2).

Results

Age-standardized death rate trends of IMM

From a global perspective, over the past 30 years, as shown in Table 1, Figure 1, and Supplementary Digital Content Figure 1 (available at: http://links.lww.com/JS9/E743), we observed that the ASDR associated EAPC was −0.64 (95% CI: −0.87 to −0.41), indicating a significant downward trend year by year. Starting from 1990 and 2019, we found that ASDR decreased from 0.77/100 000 people (95% UI, 0.68–0.87) in 1990 to 0.65/100 000 people (95% UI, 0.56–0.74) in 2019. Among 195 countries and regions worldwide, we found that the top three countries in terms of ASDR related EAPC decline were Romania (28.69), Mongolia (27.2), and Kazakhstan (25.22). However, as shown in Supplementary Digital Content Table 1 (available at: http://links.lww.com/JS9/E743) to Supplementary Digital Content Table 3 (available at: http://links.lww.com/JS9/E743), the top three countries in terms of ASDR related EAPC decline are Estonia (0.05), Poland (0.06), and Finland (0.11). In addition, the top three regions in terms of ASDR related EAPC decline are high-income North America (5.9), Andean Latin America (5.15), and Central Sub Saharan Africa (3.56). On the contrary, the top three regions with the lowest decline in ASDR related EAPC are East Asia (0.22), Eastern Europe (0.29), and high-income Asia Pacific (0.34). In summary, the global IMM related ASDR has shown a dynamic downward trend over the past 30 years, and the downward trend is more significant in developing countries and regions than in developed countries and regions.

Table 1.

The death cases and ASDR in 1990 and 2019 and its temporal trends

	1990			2019		1990–2019
	Death cases No. 102 (95% UI)	ASDR per 100 000 No. (95% UI)		Death cases No. 102 (95% UI)	ASDR per 100 000 No. (95% UI)	EAPC No. (95% CI)
Global	213.39 (189.2–242.03)	0.77 (0.68–0.87)		252.14 (216.76–289.52)	0.65 (0.56–0.74)	−0.64 (−0.87 to −0.41)
Socio-demographic index
High SDI	1.24 (1.12–1.37)	0.03 (0.03–0.03)		2.69 (2.38–3.08)	0.06 (0.05–0.06)	3.88 (3.31–4.45)
High-middle SDI	18.88 (16.13–22.53)	0.3 (0.26–0.36)		9.42 (8.23–10.92)	0.14 (0.12–0.16)	−3.04 (−3.33 to −2.74)
Middle SDI	51.81 (45.45–59.08)	0.56 (0.49–0.63)		40.55 (35.05–47.32)	0.33 (0.28–0.38)	−1.99 (−2.17 to −1.81)
Low-middle SDI	103.16 (87.81–119.86)	1.86 (1.58–2.16)		110.44 (92.27–129.64)	1.17 (0.97–1.37)	−1.66 (−2.05 to −1.26)
Low SDI	38.22 (32.42–44.06)	1.64 (1.39–1.88)		88.87 (73.98–105.53)	1.68 (1.39–1.99)	0.32 (0.12–0.51)
Region
Andean Latin America	0.23 (0.19–0.27)	0.12 (0.1–0.14)		1.18 (0.88–1.53)	0.35 (0.26–0.46)	4.77 (4.12–5.42)
Australasia	0.02 (0.01–0.02)	0.02 (0.01–0.02)		0.03 (0.02–0.03)	0.02 (0.02–0.02)	0.81 (−0.05–1.67)
Caribbean	0.66 (0.58–0.75)	0.33 (0.29–0.38)		1.24 (0.95–1.58)	0.51 (0.39–0.65)	2.27 (1.89–2.64)
Central Asia	0.27 (0.24–0.31)	0.07 (0.06–0.08)		0.43 (0.37–0.51)	0.08 (0.07–0.1)	1.97 (1.3–2.65)
Central Europe	0.08 (0.07–0.09)	0.01 (0.01–0.01)		0.05 (0.04–0.06)	0.01 (0.01–0.01)	0.78 (−0.65–2.23)
Central Latin America	2.52 (2.26–2.8)	0.29 (0.26–0.32)		6.33 (5.09–7.91)	0.47 (0.38–0.58)	1.92 (1.52–2.32)
Central Sub-Saharan Africa	2.32 (1.78–2.98)	0.98 (0.74–1.26)		8.26 (6.05–10.67)	1.41 (1.02–1.84)	2.12 (1.73–2.5)
East Asia	25.22 (19.86–31.38)	0.36 (0.28–0.44)		5.58 (4.24–7.08)	0.08 (0.06–0.1)	−5.42 (−6.2 to −4.63)
Eastern Europe	2 (1.73–2.29)	0.18 (0.16–0.21)		0.57 (0.45–0.74)	0.06 (0.05–0.08)	−3.87 (−4.17 to −3.56)
Eastern Sub-Saharan Africa	12.22 (10.22–14.33)	1.46 (1.22–1.7)		24.97 (19.67–31.2)	1.27 (1–1.58)	−0.03 (−0.39–0.33)
High-income Asia Pacific	0.21 (0.17–0.24)	0.02 (0.02–0.03)		0.07 (0.06–0.08)	0.01 (0.01–0.01)	−3.12 (−3.37 to −2.86)
High-income North America	0.37 (0.31–0.44)	0.02 (0.02–0.03)		2.18 (1.88–2.53)	0.12 (0.11–0.14)	8.18 (7.11–9.26)
North Africa and Middle East	13.29 (11.28–15.57)	0.86 (0.73–1.01)		22.31 (16.26–29.71)	0.69 (0.5–0.92)	−0.71 (−0.78 to −0.64)
Oceania	0.56 (0.38–0.75)	1.76 (1.21–2.37)		1.65 (1.17–2.3)	2.41 (1.7–3.36)	0.72 (0.39–1.05)
South Asia	127.02 (106.09–151.86)	2.36 (1.99–2.83)		114.55 (92.48–140.99)	1.17 (0.95–1.44)	−2.7 (−3.17 to −2.22)
Southeast Asia	5.43 (4.6–6.54)	0.22 (0.19–0.27)		6.85 (5.7–8.04)	0.19 (0.16–0.22)	−0.64 (−0.74 to −0.55)
Southern Latin America	0.36 (0.31–0.41)	0.14 (0.12–0.16)		0.95 (0.81–1.11)	0.28 (0.24–0.32)	3.06 (2.41–3.71)
Southern Sub-Saharan Africa	3.28 (2.77–3.82)	1.24 (1.05–1.44)		5.38 (3.79–7.19)	1.22 (0.87–1.62)	2.06 (0.44–3.7)
Tropical Latin America	3.11 (2.68–3.64)	0.38 (0.33–0.45)		5 (4.31–5.74)	0.41 (0.36–0.48)	0.49 (0.3–0.69)
Western Europe	0.38 (0.35–0.42)	0.02 (0.02–0.02)		0.19 (0.17–0.21)	0.01 (0.01–0.01)	−2.17 (−2.53 to −1.81)
Western Sub-Saharan Africa	13.86 (11.29–16.83)	1.68 (1.37–2.03)		44.37 (34.65–56.28)	2.07 (1.62–2.62)	1.1 (0.93–1.27)

ASDR age-standardized death rate.

Figure 1.

The worldwide ASRs (per 100 000 population) of indirect maternal death and DALY in 194 countries in 2019. (A) ASDR. (B) Age-standardized DALY rate. ASDR: age-standardized death rate; DALY: disability-adjusted life year.

Age-standardized DALY rate trends of IMM

In the past 30 years, the EAPC of AS-DALY worldwide has also shown a decreasing trend year by year, as shown in Table 2, Figure 1, and Supplementary Digital Content Figure 1 (available at: http://links.lww.com/JS9/E743). The EAPC of AS-DALY is −0.69 (95% CI: −0.92 to −0.45). Starting from 1990 and 2019, we found that AS-DALY decreased from 45.88/100 000 people (95% UI, 40.65–52.05) in 1990 to 38.37/100 000 people (95% UI, 32.89–44.35) in 2019. As shown in Supplementary Digital Content Table 1 (available at: http://links.lww.com/JS9/E743), Supplementary Digital Content Table 2 (available at: http://links.lww.com/JS9/E743), and Supplementary Digital Content Table 4 (available at: http://links.lww.com/JS9/E743), among 195 countries and regions worldwide, we found that the top three countries in terms of AS-DALY related EAPC decline were Romania (27.19), Mongolia (25.71), and Kazakhstan (24.1). However, the top three countries with the lowest decline in AS-DALY related EAPC are Poland (0.05), Estonia (0.05), and Italy (0.11). In addition, the top three regions in terms of AS-DALY related EAPC decline are high-income North America (5.47), Andean Latin America (5.06), and Central Sub Saharan Africa (3.49). On the contrary, the top three regions with the lowest decline in AS-DALY related EAPC are East Asia (0.21), Eastern Europe (0.28), and high-income Asia Pacific (0.33). Similarly, in the past 30 years, the global IMM related ASDR has shown a dynamic downward trend, and the downward trend is more significant in developing countries and regions than in developed countries and regions.

Table 2.

The DALY and age-standardized DALY rate in 1990 and 2019 and its temporal trends

	1990		2019		1990–2019
	DALY No. 102 (95% UI)	Age-standardized DALY Rate per 100 000 No. (95% UI)	DALY No. 102 (95% UI)	Age-standardized DALY Rate per 100 000 No. (95% UI)	EAPC No. (95% CI)
Global	12 954.68 (11 465.18–14 723.71)	45.88 (40.65–52.05)	14 910.59 (12 793.71–17 208.9)	38.37 (32.89–44.35)	−0.69 (−0.92 to −0.45)
Socio-demographic index
High SDI	72.07 (65.46–79.59)	1.67 (1.52–1.84)	145.36 (128.04–166.72)	3.12 (2.74–3.58)	3.63 (3.08–4.19)
High-middle SDI	1141.48 (972.07–1362.51)	18.25 (15.56–21.76)	552.41 (479.59–643.25)	8.21 (7.1–9.58)	−3.08 (−3.38 to −2.78)
Middle SDI	3152.49 (2745.2–3601.54)	33.19 (29.21–37.71)	2408.58 (2074.74–2819)	19.5 (16.8–22.81)	−2 (−2.19 to −1.82)
Low-middle SDI	6310 (5374.1–7336.14)	110.97 (94.44–129.42)	6594.46 (5527.08–7788.46)	69.01 (57.73–81.44)	−1.7 (−2.09 to −1.3)
Low SDI	2273.89 (1919.89–2620.79)	94.62 (80.29–108.98)	5199.56 (4316.31–6175.75)	95.2 (79.13–113.03)	0.25 (0.05–0.45)
Region	72.07 (65.46–79.59)	1.67 (1.52–1.84)	145.36 (128.04–166.72)	3.12 (2.74–3.58)	3.63 (3.08–4.19)
Andean Latin America	13.88 (11.58–16.4)	6.94 (5.8–8.18)	70.18 (52.31–90.97)	20.85 (15.55–27.01)	4.8 (4.14–5.46)
Australasia	1.01 (0.85–1.19)	0.92 (0.78–1.09)	1.55 (1.32–1.81)	1.15 (0.98–1.35)	0.8 (−0.06–1.66)
Caribbean	41.16 (36–46.65)	20.49 (17.96–23.24)	74.47 (57.56–94.74)	30.78 (23.77–39.11)	2.2 (1.81–2.58)
Central Asia	16.53 (14.54–18.77)	4.39 (3.88–4.97)	25.94 (22.24–30.61)	5.13 (4.39–6.03)	1.92 (1.25–2.58)
Central Europe	4.52 (3.94–5.2)	0.76 (0.66–0.89)	2.87 (2.29–3.54)	0.61 (0.48–0.75)	0.76 (−0.68–2.22)
Central Latin America	153.62 (138.04–171)	17.19 (15.43–19.07)	382.54 (307.96–473.18)	28.2 (22.72–34.88)	2.01 (1.6–2.41)
Central Sub-Saharan Africa	135.01 (103.13–173.51)	55.02 (42.22–70.79)	471.1 (346.81–604.48)	77.71 (56.91–100.63)	2.05 (1.67–2.43)
East Asia	1527.05 (1203.08–1900.27)	21.33 (16.8–26.47)	323.87 (246.84–410.15)	4.68 (3.58–5.88)	−5.44 (−6.24 to −4.65)
Eastern Europe	121.65 (104.96–139.78)	11.12 (9.64–12.79)	33.51 (26.09–42.66)	3.62 (2.83–4.48)	−4 (−4.31 to −3.69)
Eastern Sub-Saharan Africa	724.97 (604.24–851.31)	83.36 (69.63–97.89)	1462.19 (1146.27–1827.63)	71.94 (56.82–89.45)	−0.06 (−0.43–0.3)
High-income Asia Pacific	12.11 (9.99–14.24)	1.38 (1.14–1.63)	4.01 (3.26–4.78)	0.55 (0.45–0.66)	−3.19 (−3.44 to −2.93)
High-income North America	21.44 (17.68–25.54)	1.42 (1.17–1.7)	117.23 (100.6–136.55)	6.84 (5.85–8)	7.84 (6.78–8.91)
North Africa and Middle East	781.38 (665.02–916.37)	48.93 (41.49–57.34)	1285.47 (938.46–1709)	39.77 (29.07–52.9)	−0.7 (−0.78 to −0.62)
Oceania	33.42 (22.5–45.2)	103.3 (70.87–138.52)	98.23 (68.97–136.38)	141.79 (99.49–197.98)	0.73 (0.4–1.06)
South Asia	7816.83 (6520.88–9328.48)	142.99 (119.57–170.98)	6919 (5562.17–8552.67)	70.12 (56.48–86.52)	−2.73 (−3.21 to −2.26)
Southeast Asia	320.5 (271.51–386.07)	12.88 (10.94–15.51)	396.17 (328.6–465.53)	11.02 (9.16–12.96)	−0.62 (−0.71 to −0.53)
Southern Latin America	21.41 (18.59–24.47)	8.49 (7.37–9.71)	55.4 (47.25–64.64)	16.2 (13.8–18.87)	3.05 (2.41–3.69)
Southern Sub-Saharan Africa	191.9 (161.56–224.61)	70.61 (59.6–82.19)	310.09 (215.9–418)	69.98 (49.4–93.18)	2.07 (0.44–3.74)
Tropical Latin America	183.86 (157.65–215.23)	22.27 (19.18–26.05)	295.88 (255.1–340.57)	24.82 (21.34–28.69)	0.6 (0.39–0.81)
Western Europe	22.46 (20.22–24.7)	1.16 (1.04–1.27)	11.03 (9.85–12.21)	0.61 (0.54–0.67)	−2.18 (−2.55 to −1.81)
Western Sub-Saharan Africa	809.95 (657.68–986.38)	94.48 (77.16–114.46)	2569.86 (1994.68–3267.44)	116.08 (90.82–147.13)	1.1 (0.92–1.27)

DALY, disability adjusted life-years.

Correlation analysis of IMM related ASDR, AS-DALY and different SDI

In the past 30 years, as shown in Table 1, Table 2, and Figure 2, we have found that the EAPC of high-medium SDI has shown a significant downward trend year by year, with an EAPC of −3.04 (95% CI: −3.33 to −2.74), while the EAPC of high SDI has shown a significant upward trend year by year, with an EAPC of 3.88 (95% CI: 3.31–4.45). In addition, the EAPC for high-middle SDI is −3.08 (95% CI: −3.38 to −2.78), followed closely by middle SDI with an EAPC of −2.00 (95% CI: −2.19 to −1.82). From the global SDI classification level, as shown in Figure 3 and Supplementary Digital Content Figure 2 (available at: http://links.lww.com/JS9/E743), we observed a significant correlation between ASDR and SDI, with EAPC showing a negative correlation with ASDR (R = −0.27, P < 0.01). Similarly, there was a significant correlation between AS-DALY and SDI, with EAPC showing a negative correlation with AS-DALY (R = −0.27, P < 0.01). This indicates that under the strong support of economic and health policies in underdeveloped countries and regions, the adverse outcomes of IMM have gradually improved.

Figure 2.

The trends of age-standardized death (A) and DALY (B) rate from 1990 to 2019 among different SDI quintiles.

Figure 3.

The correlations between EAPC and age-standardized death (ASDR; A), DALY (C) rate in 1990, and correlations between EAPC of death (B), DALY (D), and HDI in 2019. Each circle represents a country and the size represents number of indirect maternal death patients. The r value is the correlation coefficient of Pearson’s correlation. EAPC, estimated annual percentage change; HDI, human development index.

Correlation analysis of IMM related ASDR, AS-DALY and different age periods

Previous literature has reported a significant correlation between age and IMM. In this study, we conducted segmented comparisons of age and analyzed the correlation between ASDR, AS-DALY, and age, as shown in Figure 4 and Supplementary Digital Content Figure 3 (available at: http://links.lww.com/JS9/E743). Even when analyzing the proportion of ASDR and AS-DALY in different regions and age groups, the results showed that in the maternal population age groups of 1990 and 2019, those aged 25–29 maintained the highest proportion, and there was a consistent high proportion trend regardless of country and region. Moreover, as shown in Supplementary Digital Content Table 5 (available at: http://links.lww.com/JS9/E743), Supplementary Digital Content Table 6 (available at: http://links.lww.com/JS9/E743), and Supplementary Digital Content Figure 4 (available at: http://links.lww.com/JS9/E743), the top three regions in 1990 were mainly Australasia (34.03%), high-income Asia Pacific (32.22%), and Western Europe (27.74%). In 2019, the top three regions were mainly Australasia (31.65%), Central Asia (28.37%), and high-income Asia Pacific (24.02%). The following high-risk age group is 30–34 years old, which also indicates that in various countries and regions around the world, the trend of age-related trends has led to IMM related ASDR and AS-DALY becoming a serious disease burden at the optimal reproductive age for women. Therefore, it is necessary to provide sufficient prenatal and pregnancy monitoring for women of childbearing age.

Figure 4.

The composition of indirect maternal death of different ages by region. (A) Death rate in 1990. (B) Death rate in 2019.

Risk factors for the burden of IMM

In this study, we analyzed the disease burden risk of IMM based on the risk factors entered into the GBD database. The results showed that iron deficiency exhibited significant regional heterogeneity in IMM among pregnant women worldwide. As shown in Figure 5 and Figure 6, iron deficiency is an independent risk factor for IMM related DALYs, particularly evident in South Asia, low-middle SDI, and low SDI. However, in high-SDI regions, the risk factors for iron deficiency over the past 30 years have consistently stabilized and shown a gradual decline. Fortunately, in the low-SDI and low-middle-SDI regions, the association between iron deficiency and IMM has shown a significant weakening trend over the past 30 years, reflecting the significant achievements of underdeveloped areas in maternal nutrition supplementation.

Figure 5.

The indirect maternal death DALYs attributed to iron deficiency in 1990 (A) and 2019 (B). DALY: disability-adjusted life year.

Figure 6.

The trends of age-standardized DALYs (A) and deaths (B) attributable to iron deficiency from 1990 to 2019.

Discussion

In recent years, multiple crises including public health emergencies, climate change, economic and political turmoil, and armed conflicts have intensified the challenges faced by the public in accessing medical services and increased the health risks faced by pregnant women^[18,19]. In fact, maternal health is the result of a combination of various factors, especially individual and family factors, medical institutions and health personnel factors, health system factors, social and natural factors, and these factors often interact with each other^[20–22]. The 2023 WHO data shows that approximately 295 000 women die each year during the perinatal continuum, with stark regional disparities exacerbating this crisis. Sub-Saharan Africa bears the greatest burden, accounting for 67% of these fatalities[11]. Although extensive public health interventions have been implemented globally, maternal mortality rates in low-income regions remain 14 times higher than in high-income areas, a gap that highlights systemic inefficiencies in healthcare delivery[23].

Notably, despite decades of investment in maternal health initiatives, the persistent disparity between resource-rich and resource-constrained settings underscores the need for targeted, context-specific strategies. While high-income nations have achieved progress through integrated care models, many low-income countries continue to face structural challenges, including insufficient obstetric staffing and fragmented healthcare systems, thus highlighting the urgency of systemic reforms to address global health inequity. Given these disparities, reducing the disease burden of IMM urgently requires new strategies to address it, as it involves changes in the healthcare system. To our knowledge, this is the first time tracking the dynamic disease burden trends of IMM in nearly 200 countries and regions worldwide over the past 30 years. Based on objective indicators such as ASDR, and AS-DALY, we provide complete and detailed data support for accurately depicting the current status of IMM in different SDI regions, as well as future policy formulation and medical allocation. We believe that this updated IMM disease burden report will contribute to the development of healthcare policies.

Recent studies have shown that although the global maternal mortality rate has decreased by 40% since 2000, with an average annual decline rate of 2.2%, it is only about one-third of the 6.4% annual decline rate required to achieve the sustainable development goals with a maternal mortality rate of less than 70/100 000 by 2030^[24–27]. According to current trends, 4/5 of countries will struggle to achieve the global goal of reducing maternal mortality by 2030, and 1/3 of countries will be unable to achieve the goal of reducing neonatal mortality. In view of this, in order to ensure that every woman and newborn has access to necessary healthcare services, United Nations agencies are calling on governments and the international community to increase investment in maternal and infant health, particularly in strengthening healthcare infrastructure and training and equipping midwives, nurses, and community health workers. However, the allocation of medical resources must be reasonably based on the local disease burden to ensure the maximum reduction of maternal mortality rate without wasting medical assistance resources.

The results of this study show that in the past 10 years, the maternal mortality rate in sub Saharan Africa has decreased by about 40%, achieving significant results, especially in one of the three United Nations regions where maternal mortality rates have significantly decreased since 2015, the other two regions being Australia and New Zealand, and Central and South Asia. This is consistent with previous research reports, and it is worth mentioning that as of 2024, the maternal mortality rate in low-income countries is 346/100 000, nearly 35 times higher than in high-income countries^[24,28,29]. Among them, due to high poverty rates and multiple conflicts, nearly 70% of maternal deaths occur in sub Saharan Africa. In addition, in high-income and upper middle income countries, 99% of deliveries are attended by professional medical personnel, while in low-income countries, this proportion is only 73%. This indicates that the healthcare security system still exists due to persistent inequality and uneven economic development between regions and countries. Therefore, with 4/5 of countries projected to miss the 2030 maternal mortality target, urgent scaling of interventions is critical.

In fact, advanced pregnancy can pose significant health risks to both pregnant women and fetuses. The age of the mother emerges as a pivotal factor in the landscape of indirect maternal mortality. As per the WHO’s 2023 report, adolescent mothers, those under 20 years of age, endure a 35–50% greater peril of maternal mortality compared to their adult counterparts[30]. In this study, we also observed a significant correlation between IMM and age. For example, women aged 20–30 have the highest IMM burden, especially in this childbearing age group, we observed that incidence rate, mortality and disability adjusted life year rate all reached peak values. According to the United Nations’ plan and combined with global future population projections, the global birth rate will slowly increase from 133 million in 2023 to 136 million in 1951, and then slowly decline to 109 million by 2100^[31,32]. With the prevalence of low birth rates and phenomena such as delayed marriage and childbirth, we have to pay attention to the correlation between the optimal childbearing age and IMM.

Previous studies have shown that compared with young women, women over 35 years old have an increased risk of maternal diseases, such as pregnancy diabetes, eclampsia, pregnancy induced hypertension, placenta previa, cesarean section, etc^[33–36]. In addition, maternal age is also an independent risk factor for placenta previa[37]. Placenta previa, including placental abruption, placenta previa, and placental hyperplasia, is related to factors such as parity, hypertension, and the number of fetuses[38]. The risk of placenta previa is 10 times higher for first-time mothers aged 40 or older compared to first-time mothers aged 20–29^[39,40]. Consistent with the results of this study, in developed countries, the maternal mortality rate ranges from 3.1 to 8.6 per 100 000. As the proportion of elderly women continues to increase, an increase in the “elderly specific age” maternal mortality rate can be observed. In summary, elderly pregnant women have an increased risk of adverse pregnancy outcomes compared to eligible pregnant women, and the risk of adverse perinatal outcomes increases with age. The relationship between the two is a continuous effect rather than a threshold effect. Moreover, there is a potential correlation between the disease burden brought by this age and the level of economic development and social security. Collectively, the elevated risk can be primarily attributed to two intertwined aspects. Physiologically, younger mothers’ bodies may not be fully matured, making them more prone to non-obstetric complications like hypertensive disorders. Simultaneously, socioeconomic barriers often impede their access to adequate healthcare, thereby increasing the likelihood of succumbing to infectious or other indirect causes. Therefore, it is necessary to strengthen the science popularization and health policy support for the optimal childbearing age.

Previous studies have also shown that advanced age, comorbidities, obesity, adverse pregnancy and childbirth history are independent risk factors for IMM^[41,42]. In addition, the professional skill level of medical technicians, the gap in health services provided by medical institutions, and the unequal distribution of medical resources may also lead to pregnant women being unable to obtain high-quality medical services in a timely manner, thereby increasing the risk of maternal mortality^[42,43]. In this study, based on the risk factor analysis included in the GBD database, we found that iron deficiency is an important factor leading to IMM, especially in women aged 25–34 in low-SDI areas. Anemia is usually the result of iron deficiency, and previous studies have shown that it significantly increases the risk of miscarriage and other pregnancy complications. According to the latest estimates from the WHO, the incidence of anemia in pregnant and non-pregnant women aged 15–49 has increased in most countries[44]. In 2016, 40.05% of pregnant women worldwide were diagnosed with anemia^[44,45]. Moreover, anemia is listed as an important risk factor for postpartum hemorrhage in the WHO Global Health Guidelines and is used as a predictive indicator in multiple postpartum hemorrhage risk prediction models[46]. Although our study once again confirmed the close association between anemia and IMM. However, the impact of anemia severity and occurrence period on more adverse outcomes is still unclear, so clarifying its specific association is of great significance for clinical intervention of anemia in pregnant women and ensuring the health of both mother and child. Future research should strengthen monitoring of anemia in pregnant women, comprehensively evaluate the relationship between hemoglobin levels and changes during different stages of pregnancy and pregnancy outcomes, further clarify the optimal range of hemoglobin during pregnancy, and provide a basis for revising the definition of anemia during pregnancy and improving clinical management strategies for anemia in pregnant women.

Our study still inevitably has the following limitations. Firstly, this study is a retrospective observational study, relying on IMM data entered into the GBD database. It is inevitable that there may be potential biases related to data integrity and quality in some countries and regions, which are particularly evident in economically underdeveloped countries. Therefore, the health supervision system for low-SDI areas still needs to be improved in the future. Second, the fluctuation of incidence rate and DALYs annualized rate partly describes the dynamic changes of female IMM in the past 30 years. However, for the age distribution and childbearing age in local areas, as well as in different countries and regions, in view of the lack of GBD database, it is still necessary to focus on population cohort research in specific regions in the future, so as to achieve targeted health policy formulation. Thirdly, the risk factors for maternal mortality are complex and closely related. Although this study explored anemia as an independent risk factor for IMM, it cannot comprehensively and systematically summarize the attribution risk of IMM. Therefore, large-scale, cross regional, and multi center epidemiological studies are still needed in the future to clarify the adverse effects caused by various attribution risk factors.

Conclusion

In conclusion, the global burden report of IMM for pregnant and lying in women shows that although the world has made some progress in reducing maternal mortality, especially in high-SDI regions, the incidence rate, mortality and DALYs rate continue to decline, reflecting the improvement of maternal health. However, significant differences still exist, with low-SDI areas facing a heavier burden due to limited access to healthcare, socio-economic factors, and iron deficiency. Therefore, more efforts are still needed, especially in areas with limited resources and frequent conflicts. At the same time, based on the dynamic analysis of age stratification, it was determined that women aged 25–34 had the highest incidence rate of IMM, which emphasized the necessity of scientific advocacy for people of the best childbearing age. In addition, according to current trends, 4/5 of countries will be unable to achieve the global goal of reducing maternal mortality by 2030, and 1/3 of countries will not be able to achieve the goal of reducing neonatal mortality. Therefore, addressing socio-economic factors, improving access to healthcare, and targeted nutrition interventions are crucial for further reducing the global burden of IMM.

Ethical approval and consent

Due to the retrospective nature of this study and the fact that the privacy information of IMM has been encrypted on the GBD official website, we have granted an informed consent waiver for ethical applications in this study.

Sources of funding

Not applicable.

Author contributions

B.W. and J.J. took major responsibility for the design and conduct of the study, contributed to data analysis, and drafted the manuscript. L.Q. assisted in data curation and validation. All authors reviewed, edited, and approved the final manuscript.

Conflicts of interest disclosure

The authors declare no competing interests.

Guarantor

Wang Bo accepts full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish.

Research registration unique identifying number (UIN)

Not applicable.

Provenance and peer review

Not commissioned, externally peer-reviewed

Data availability

Data from GBD database are accessible through the Institute for Health Metrics and Evaluation (IHME)’s online data tool (https://ghdx.healthdata.org/gbd).