Spatiotemporal disparity of breast cancer incidence in Iranian female populations at the district level from 2000 to 2021: Bayesian disease mapping

Shadi Rahimzadeh; James Bentham; Beata Burczynska; Farshad Farzadfar; Mariachiara Di Cesare

doi:10.1371/journal.pone.0330017

Abstract

Introduction

While trends in breast cancer incidence in Iran are generally monitored at the national level, little is known about subnational variations in these trends. This study aimed to assess the levels and trends (2000–2021) of the relative risk (RR) of breast cancer incidence at the district level in Iran and its relation to key socioeconomic dimensions to understand the full extent of geographical and social inequalities in the country associated with breast cancer morbidity.

Methods

District-level incidence data by age and sex from the National Cancer Registry System of the Iranian Ministry of Health were used. Related covariates were extracted from the Census and Household Expenditure and Income Survey (HEIS) datasets. The RR of breast cancer incidence was estimated in women above the age of 30 for all 316 districts in Iran from 2000 to 2010 using a Bayesian spatiotemporal model. Finally, predictions were estimated for the period 2011–2021.

Results

The national RR of breast cancer incidence in Iran increased from 0.21 (95% credible interval (CrI): 0.19, 0.22) in 2000 to 0.66 (0.63, 0.68) in 2010 and 1.23 (1.18, 1.28) in 2021. The RR of breast cancer incidence was highest in Yazd (1.96 [1.63, 2.33]), Shiraz (1.90 [1.72, 2.09]) and Shemiranat (1.90 [1.12, 2.91]) in 2010 and in Tehran (3.99 [3.86, 4.33]), Bushehr (3.89 [3.07, 4.77]) and Abadan (3.67 [2.99, 4.39]) in 2021. In contrast, Savojbolagh, Saravan and Nikshahr had the lowest RRs in both 2010 (0.11 [0.05, 0.20], 0.17 [0.08, 0.30] and 0.20 [0.09, 0.36], respectively) and 2021 (0.19 [0.10, 0.33], 0.34 [0.18, 0.54] and 0.35 [0.17, 0.62], respectively). The RR of breast cancer incidence was 60% greater across districts in the highest YOS quintile (average years of schooling: 3.9) than in those in the lowest YOS quintile (average years of schooling: 2.2; relative index of inequality: 1.6).

Conclusions

The results show that the RR of breast cancer incidence has increased over time (2000–2021) at the national and subnational levels in Iran. Breast cancer is one of the few diseases with a positive education gradient, with a greater RR of breast cancer incidence among higher-educated women than among lower-educated women. However, this is likely due to better awareness of diagnostic approaches and access to those approaches rather than reflecting patterns in the true incidence of breast cancer. While social inequalities are a major barrier to reducing the prevalence and incidence of breast cancer, it is important to track the progress made at the district level based on the characteristics of specific policies aimed at reducing health inequalities. A scaling-up in the quality of healthcare services, national and subnational policies addressing prevention and treatment, and more specialised training programmes for women’s health are needed.

Citation: Rahimzadeh S, Bentham J, Burczynska B, Farzadfar F, Di Cesare M (2025) Spatiotemporal disparity of breast cancer incidence in Iranian female populations at the district level from 2000 to 2021: Bayesian disease mapping. PLoS One 20(9): e0330017. https://doi.org/10.1371/journal.pone.0330017

Editor: Muhammad Farooq Umer, King Faisal University, SAUDI ARABIA

Received: March 5, 2025; Accepted: July 24, 2025; Published: September 11, 2025

Copyright: © 2025 Rahimzadeh et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting information files.

Funding: The author(s) received no specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Currently, accelerating the pace of progress in cancer diagnosis and treatment is crucial, given the COVID-19 pandemic, which has led to health care provision interruptions worldwide [1–3]. In 2020, female breast cancer was the leading cause of global cancer incidence, with an estimated 2.3 million new cases, representing 11.7% of all cancer diagnoses [4], making it one of the most severely burdensome cancers globally [4,5]. Among women, breast cancer accounts for 1 in 4 cancer cases and 1 in 6 cancer deaths, ranking first in incidence (159 of 185 countries) and mortality (110 of 185 countries) [4].

To address the increasing global burden of cancer, in 2021, the World Health Organization (WHO) established the Global Breast Cancer Initiative (GBCI), with the aim of reducing global breast cancer mortality by 2.5% per year until 2040 via early detection, timely diagnosis, and comprehensive breast cancer management [6,7].

While breast cancer incidence and mortality trends are generally monitored at the national level [8,9], little is known about the distribution and variation of breast cancer at the subnational level. Considering geographical variability in breast cancer incidence across small areas and over time is essential to facilitate appropriate subnational allocation of resources to increase breast cancer screening and achieve better outcomes [10,11]. Improved geographic resource allocation, especially in low- and middle-income settings where resources are scarce, could maximise the impact of screening interventions and possibly reduce geographic variability in mortality [10].

In Iran, breast cancer is the most common cancer and the fifth most common cause of death among females [12]. In 2019, the age-standardised breast cancer incidence rate was 34.1 per 100,000 women (30.7, 37.9), while the age-standardised death rate was 11.9 deaths per 100,000 women (10.8, 13.1), with percentage changes of 71.2 (27.9, 118.3) and 15.5 (−3.7, 31.7), respectively, between 1990 and 2019 [13]. Current trends highlight the need for robust estimates of geographical and temporal patterns to support effective public health policies in a country characterised by insufficient financial resources and a lack of continuity of financing [14]. Here, we estimated the relative risk (RR) of breast cancer incidence in women aged 30 years or older in all 316 districts (S1 Table, S1 Fig) in Iran from 2000 to 2021 and assessed the correlation between breast cancer incidence and socioeconomic status at the district level.

Methods

Data

The data used in the present study primarily came from two sources: the National Cancer Registry System of Iran (CRS) (from 2000 to 2010, excluding 2006), which has been used to retrieve breast cancer data, and the Census and Household Expenditure and Income Survey (HEIS) (from 2000 to 2010), which has been used to retrieve population data and covariates. The selection of covariates, including female mean years of schooling, urbanisation, wealth index, and cancer registry completeness, was based on their availability, relevance to breast cancer risk factors, and contribution to model fit as assessed by the Deviance Information Criterion (DIC), ensuring balance between data-driven estimation and model stability. Detailed information on these data sources is given elsewhere [15]. All data were provided at the aggregated district-year level without missing values. Any data cleaning or imputation for missingness was conducted by the original data custodians before aggregation. Several health system components, including the number of health centres/units, number of specialist physicians (consultants), number of general physicians (GPs), number of pharmacies, number of beds in hospitals, number of intensive care units (ICUs), number of nurses and number of community health workers (Behvarzes) per 100,000 people at the province level, have also been used to assess their correlation with the predicted incidence of breast cancer in 2020.

Statistical analysis

When some populations at risk are small or the disease is rather rare, RRs may be unreliable [16]. Spatiotemporal modelling allows the borrowing of information from neighbouring areas and years, providing estimates for locations and years with little or no data [17]. In this study, a Bayesian Poisson spatiotemporal model was fitted to model the log-risk estimates using a combination of spatially structured and unstructured random effects [18] and covariates. The female mean years of schooling (YOS), female urbanisation, wealth index (WI) and completeness percentage of the cancer registry were used as covariates. We assumed that the number of incident breast cancer cases in each district-year follows a Poisson distribution. The spatiotemporal random effects included both spatially structured and unstructured components. The spatially structured random effect was modelled using a conditional autoregressive (CAR) prior to account for spatial correlation between neighbouring districts, whereas the unstructured component was assigned an independent and identically distributed normal prior. Temporal random effects were modelled assuming exchangeability across years without explicit temporal structure. Independence between the spatially structured and unstructured effects was assumed. It was assumed that the covariates had fixed effects. We examined the relationship between the standardised morbidity ratio (SMR) against the RR of breast cancer incidence for 316 districts. SMRs represent the observed number of cases divided by the expected number of cases [19], which were calculated using an indirect age-standardisation technique. The expected number of incidence cases was estimated by multiplying the mean national rate for Iran in 2010 by the age-specific population for each district-age group.

To assess the probability that each district’s RR exceeded the national RR, we compared the posterior samples of each district to those of the national estimate. For each district we obtained 20,000 posterior samples using Markov Chain Monte Carlo (MCMC) methods. After discarding the initial 5,000 iterations as a burn-in period we used the remaining 10,000 iterations from 2 chains for the analysis (combined to form 20,000 posterior samples). At each iteration, we assessed whether the district’s RR was higher or lower than the national RR. We repeated this process for the 20,000 iterations, and calculated the probability as the proportion of iterations where the district RR exceeded the national RR. Additionally, we used 20,000 posterior samples to estimate the national rates for each year and to project district-level rates from 2011 onward.

Given the absence of data beyond 2011, projecting future trends was necessary. To address this issue, we employed projection methods that leverage the power of Bayesian spatiotemporal modelling while accounting for uncertainty. Our approach involved incorporating the uncertainty from the Bayesian spatiotemporal model to make robust and reliable predictions. The first step in our projection methodology was to derive the RR estimates for each district from the Bayesian spatiotemporal model. From this model, we generated a posterior distribution of RR using two chains. To capture the full range of uncertainty, we drew a large sample of 20,000 values from the posterior distribution for each district. Next, we apply the projection model, which takes each sampled RR value as input. This allowed us to fit multiple projection models, each based on a different sampled value, and capture the variations in the projected outcomes. By running the projection model for each sampled value, we accounted appropriately for the uncertainties in the RR estimates. To obtain a comprehensive understanding of the projected outcomes, we computed summary statistics for the combined projection models. These summary statistics provided a concise representation of the projected trends while encapsulating the variability arising from the uncertainty in the RR estimates.

To examine the association between the predicted RR of breast cancer incidence and several health system indicators in 2020, a Pearson correlation test was used. Specifically, the total number of GPs, total number of pharmacies, total number of ICUs, total number of nurses, total number of specialists, total number of beds in hospitals of medical universities, and total number of health centres and community health workers (Behvarzes) were available for only one year (2020) at the province level. P values less than 0.05 were considered significant [20].

Model specification

A spatiotemporal model was applied to the district-level data, which incorporated both spatial fitting and temporal considerations (S1 Appendix). The model is specified as follows:

where i represents districts, t corresponds to years, Y(i,t) represents breast cancer incidence by district and year, E(i,t) represents the expected number of cases by district and year, λ(i,t) represents the RR parameters by district and year, α is the intercept, YOS is the mean female years of schooling, URB is the female urbanisation percentage, COMP is the cancer registry completeness percentage, U_i is a spatially structured random term, V_i is an unstructured random term, X_t is a temporal random term, ε(i,t) represents residuals by district and year and the β vector comprises the spatiotemporal coefficients corresponding to YOS, URB and COMP indicating the change in λ associated with change in YOS, URB and COMP by district and year.

The projection model applied in this study is specified as follows:

where i represents districts (i = 1,2,..., 316), j corresponds to the number of posterior samples (j = 1,2,...,20000), R_{i j} is each posterior sample of RR of breast cancer incidence for each district, X is the year, and b is a vector of regression coefficients.

The model was fitted using OpenBUGS version 3.2.3 with the Markov chain Monte Carlo (MCMC) algorithm, accessed through the R2OpenBUGS package in R for Windows version 4.2.1 (http://www.r-project.org/) and STATA software (version 11.0). The objective was to estimate the RR of breast cancer incidence by district and year, including the 2.5th and 97.5th percentiles of the posterior distribution, which represent the lower and upper bounds of the 95% credible intervals (CrIs).

Cross-validation

The validity of the model estimates was assessed using cross-validation. Ten percent of the data was withheld, and the model was fitted using the remaining 90% of the district-level data. Predictions were then generated for the held-out district-years and compared to the actual observed values. Model performance was evaluated using the median relative error, median absolute relative error, median error and median absolute error. In addition, the Bayesian Wilcoxon rank-sum test was used to assess whether the differences between predicted and observed values were statistically significant [21]. The coverage of the 95% CrIs was examined to evaluate the uncertainty in the estimates. A model with good external predictive validity, would have 95% of the held-out value falling within the 95% CrIs of the predicted estimates. In addition, formal diagnostic methods including the Brooks-Gelman-Rubin (R-hat) statistic and the Deviance Information Criterion (DIC) were used to assess model convergence and goodness-of-fit, as described in detail in S2 Appendix (S7 Fig, S2 Table).

Ethics

For this study, aggregated (by age, sex, province/district) anonymised data were accessed on 11/05/2020 and used for this analysis. Ethical approval was obtained through the Ethics Committee of Middlesex University in the UK (14142.2020) on 11/05/2020 as part of a PhD project. Approval was based on the ethical approval obtained through the Ethics Committee of the National Institute for Medical Research Development in Iran (IR.NIMAD.REC.1396.192) on 10/12/2017 by FF in their role as Director of the Non-Communicable Disease Research Centre (NCDRC), Tehran University of Medical Science. No form of consent was requested because access was limited to the aggregated anonymised data. All authors had access to the data. Consent to participate is “Not applicable”.

Results

Bayesian spatiotemporal findings for relative risk of breast cancer incidence

Among the 316 districts, there were 1095 new breast cancer cases in 2000, with the most in Isfahan (16.8% of all new cases) and 7960 in 2010, with the most in Tehran (25.4%). The national average RR of breast cancer increased from 0.21 (95% CrI: 0.19, 0.22) in 2000 to 0.52 (0.49, 0.54) in 2005 and 0.66 (0.63, 0.68) in 2010 (Fig 1). The analysis at the district level implied having more sparse data and geographical areas with no available data, especially in 2000. In S2 Fig, the estimated RR of breast cancer incidence resulting from the spatiotemporal model (left side) was compared with the raw data (right side) for all years. The comparison shows how the model can smoothly interpolate for districts with missing data. In addition, since there were no data available for 2006, the model could make estimates for all districts for 2006 consistent with earlier and later years.

Download:

Fig 1. National average for RR of breast cancer incidence by year out of all posterior samples (red diamonds show the mean value).

The absence of empirical data in 2006 leads to greater uncertainty in the model estimates for that year, resulting in wider credible intervals compared to years with available registry data.

https://doi.org/10.1371/journal.pone.0330017.g001

The three districts with the highest RR of breast cancer incidence in 2000 were Isfahan (1.14 [0.98, 1.31]), Kashan (1.02 [0.71, 1.40]) and Ahvaz (0.74 [0.57, 0.92]) while Savojbolagh (0.05 [0.02, 0.11]), Tehran (0.06 [0.04, 0.08]) and Saravan (0.08 [0.03, 0.16]) had the lowest RRs in 2000. The RR of breast cancer incidence was highest in Yazd (1.96 [1.63, 2.33]), Shiraz (1.90 [1.72, 2.09]) and Shemiranat (1.90 [1.12, 2.91]) in 2010. In contrast, Savojbolagh (0.11 [0.05, 0.20]), Saravan (0.17 [0.08, 0.30]) and Nikshahr (0.20 [0.09, 0.36]) had the lowest RRs in 2010 (S3 Table).

In addition, the posterior probability of RR for each district being greater than the national RR with values between 0 and 1 were plotted on the map of districts by year (Fig 2). In 2000, with a national mean of 0.21 (0.95% CrI: 0.19, 0.22), there were 94 districts out of 316 in which the posterior probability was more than 50% compared to 121 districts in 2010 with a national mean of 0.66 (0.63, 0.68).

Download:

Fig 2. Maps of the posterior probability for each district in which the RR exceeded the national RR by year from 2000 to 2010.

(National RR by year: 2000 = 0.21, 2001 = 0.23, 2002 = 0.31, 2003 = 0.38, 2004 = 0.41, 2005 = 0.52, 2006 = 0.55, 2007 = 0.58, 2008 = 0.61, 2009 = 0.65, 2010 = 0.66. Full national RR values with 95% credible intervals are reported in S4 Table).

https://doi.org/10.1371/journal.pone.0330017.g002

Breast cancer estimations in relation to years of schooling and wealth index

The average RR of breast cancer incidence increased with the YOS quintile, ranging from 0.15 in the lowest quintile to 0.24 in the highest quintile for 2000 and from 0.48 to 0.79 for 2010. The RR of breast cancer incidence was 60% greater across districts in the highest YOS quintile (average years of schooling: 3.9) than in those in the lowest YOS quintile (average years of schooling: 2.2; relative index of inequality: 1.6). Fig 3 shows that while the RR of breast cancer increased over time, it also increased according to the YOS quintile. The districts in the highest quintile had a greater RR than districts in the lower quintiles. In addition, S3 Fig confirms that the RR of breast cancer incidence increased more quickly according to the mean YOS in recent years than in earlier years. However, the RR slightly changed with increasing mean YOS over Q4 (S3 Fig). The same results were also observed for the WI quintiles; districts in the highest WI quintiles had a greater RR of breast cancer incidence over time (S4 Fig).

Download:

Fig 3. RR of breast cancer incidence categorised by YOS quintiles over time (each point represents a district).

Light blue indicates the lowest YOS quintile, and pink indicates the highest YOS quintile.

https://doi.org/10.1371/journal.pone.0330017.g003

Projections of the relative risk of breast cancer incidence between 2011 and 2021

If post-2010 trends continue, the RR of breast cancer incidence will reach a value of 1.23 in 2021, almost two times greater than the 2010 level (0.66) (Fig 4, S4 Table). S5 Fig shows how the RR of breast cancer changed from 2011 to 2021 based on the posterior values and uncertainty resulting from the spatiotemporal model for earlier years. The trend results for each district are also presented in S6 Fig.

Download:

Fig 4. Trend in national RR of breast cancer incidence by year from 2000 to 2021 (grey shading shows the 95% credible interval).

The absence of empirical data in 2006 leads to greater uncertainty in the model estimates for that year, resulting in wider credible intervals compared to years with available registry data. The three leading districts with the highest RRs were Tehran (3.99 [95% CrI: 3.86, 4.33]), Bushehr (3.89 [3.07, 4.77]) and Abadan (3.67 [2.99, 4.39]) in 2021. In contrast, Savojbolagh (0.19 [0.10, 0.33]), Saravan (0.34 [0.18, 0.54]) and Nikshahr (0.35 [0.17, 0.62]) had the lowest RRs in 2021 (S5 Table).

https://doi.org/10.1371/journal.pone.0330017.g004

Correlation of the RR of breast cancer incidence with health system components in 2020

The RR of breast cancer incidence was associated with multiple components of the health system. The 2020 projected RR of breast cancer was significantly positively correlated with the total number of GPs (r = 0.43, p = 0.01), the total number of pharmacies (r = 0.42, p = 0.01), the total number of ICUs (r = 0.44, p = 0.01) and the total number of nurses (r = 0.43, p = 0.01). While there was also a significant positive correlation between the RR of breast cancer and the total number of specialists (r = 0.50, p = 0.001) and the number of beds in medical university hospitals (r = 0.48, p = 0.001), no significant correlation was detected between the number of health centres (r = 0.03, p = 0.83) and the number of community health workers (Behvarzes) (r = 0.06, p = 0.69). The obtained results confirm that the presence of more health system components may contribute to more effective detection and increased incidence of breast cancer.

Cross-validation results for the spatiotemporal model at the district level

The results of the cross-validation are presented in S6 Table. The estimates of the RR of breast cancer incidence were unbiased, as evidenced by the median relative errors being very close to zero ranged from −0.10 to 0.24, and the median absolute relative error ranged from 0.22 to 0.51 across the years 2000–2010, indicating good predictive performance of the model. The median errors for the estimated RRs were also small and ranged between −0.24 (2001) and 0.22 (2010). The mean difference between the held-out values and estimates was assessed using the Bayesian Wilcoxon rank test, and the corresponding credible intervals were all nonsignificant (the distributions included zero). The cross-validation therefore confirmed that the statistical model used to estimate the RR of breast cancer incidence at the district level fit well. The 95% CrIs of the estimated RRs covered more than 86% to 100% of the observed values from 2000–2010.

Discussion

The current study investigated the RR of breast cancer incidence across 316 districts in Iran from 2000 to 2021. The results showed a substantial increase in the incidence of breast cancer nationally and sub nationally and high levels of heterogeneity at the subnational level in Iran.

When attempting to compare the results with those of previous studies, no studies were found to directly compare the RR of breast cancer incidence at the district level per year. This work revealed that all districts had increasing trends in the RR of breast cancer incidence from 2000 to 2021. The corresponding estimates from this study were in line with those of studies on province-level estimations [15,22]. However, here, we present for the first time the RR of breast cancer at the district level over the years, which can be instrumental in obtaining localised insights, optimising healthcare planning, identifying high-risk areas, investigating environmental factors, addressing treatment disparities, and promoting public health awareness [23]. The increase in incidence could be attributable to improvements in the national cancer registry system, resulting in an increase in the number of patients registered [22]. Upgrading diagnostic tools, more extensive healthcare coverage, awareness among the general population of breast cancer symptoms, and greater readiness to undergo screening despite cultural barriers were all possible factors that contributed to the increase in breast cancer incidence [22]. On the other hand, potential lifestyle risk factors, including increased fat intake, smoking and low physical activity (especially in postmenopausal women), could explain the increase in breast cancer incidence observed among Iranian women [24–26]. In addition, changes in cultural habits, such as increased age at childbearing and, in particular, older age at first pregnancy, can partly explain the observed trends [27–29]. However, the specific effect of the factors mentioned above on breast cancer incidence is still controversial and needs further investigation [22].

Our results suggest high levels of geographical heterogeneity in breast cancer incidence across Iranian districts. Previous research shows that cancer incidence in Asian countries has a positive correlation with the country’s level of development, as measured by the Human Development Index (HDI) [30]. In Iran, a direct and substantial association was also found between the incidence of breast cancer and HDI [31]. These findings align with our results in which higher levels of breast cancer incidence are observed across districts with higher levels of female education and wealth. This could be explained by increasing life expectancy, urbanisation, greater exposure to risk factors, delayed childbearing, and a greater rate of screening resulting from higher socioeconomic status [32].

Considering the significance of the healthcare system, extensive research has demonstrated a compelling association between delays in the health system and an elevated likelihood of advanced-stage breast cancer diagnosis in patients [33,34]. Addressing this issue requires the implementation of potential solutions aimed at improving diagnosis timing and, consequently, the clinical stages at which breast cancer is detected. These solutions should encompass strategies targeting the general population, healthcare professionals, and the healthcare system itself [33].

Disparities in breast cancer outcomes can result from modifiable social and health system determinants, such as poor access to care, a lack of health education, a lack of financial resources, challenging patient‐provider interactions, and structural barriers within the health system itself [35], which can guide policymakers in identifying strategies to more equally distribute clinical expertise and health infrastructure across populations [35]. Similarly, our results showed a significant positive association between breast cancer incidence and health workforce size (the number of nurses, GPs, and consultants) and between breast cancer incidence and service delivery (the number of beds in hospitals, the number of pharmacies and the number of ICUs) as components of the health system. The higher quality of the health system may allow earlier detection, leading to an increased number of cases, earlier treatment, and better overall patient outcomes [36].

Our results emphasise the high heterogeneity across districts, confirming the need for a comprehensive and practical plan to control breast cancer that considers subnational variability and a call to improve data collection for breast cancer surveillance in the country. These differences emphasise the crucial need to improve access to diagnosis and treatment facilities in the most deprived areas and reduce inequalities through a stratified resource allocation approach. The observed heterogeneity can reflect lower access, knowledge and acceptance of screening in underprivileged and smaller areas compared to large cities such as Tehran, which have multiple diagnostic facilities, cancer specialists, and healthcare coverage, in addition to higher educational levels of patients, leading to increased screening participation [37].

The Iranian health system has invested in policies focused on the domains of primary and secondary care, training healthcare professionals, and research to reach universal health coverage [38]; however, international sanctions against Iran from 2011 forced some restrictions on all these efforts [39] affecting all aspects of the health system in monitoring morbidity and mortality including access to essential cancer medications and radiotherapy services, which may influence breast cancer incidence patterns seen in this study. Sanctions have led to shortages of modern and effective cancer treatments, forcing oncologists to rely on older, less effective drugs [40]. In addition, difficulties in importing and maintaining radiotherapy equipment have disrupted timely cancer care [41]. These challenges could reduce survival rates and affect registry-based incidence estimates, as survival-dependent prevalence may vary across regions. Therefore, the observed geographic differences in breast cancer incidence may partly reflect the healthcare system limitations caused by sanctions. Policymakers should consider how sanctions indirectly had critical adverse effects on the Iranian population’s health by diminishing access to diagnostic and treatment facilities, specifically for those with cancers [42]. Eventually, the COVID-19 pandemic and its convergence with socioeconomic challenges, economic burdens, and the heavy burden of noncommunicable disease in Iran could directly and indirectly affect the health system for major diseases, especially in the most deprived areas [43].

Our results are based on a model that has accommodated spatial and temporal dependencies that could account for similarities in unobservable factors in neighbouring districts and years. Our specification of space-time interaction allowed each district to have its own temporal pattern. Therefore, this specification allowed us to provide the most comprehensive analysis of breast cancer incidence in Iran at the subnational level to date. The research covers a substantial period (2000–2021), with estimates available for 316 districts in Iran that have not been analysed and published before.

Limitations

First, we presented estimates at the district level for breast cancer incidence only since no mortality data were available at the district level. Modelling mortality data at a smaller spatial level could yield detailed information regarding mortality trends in the country and critical information for prioritising resource allocation. We also lacked access to tumor stage data at the district level, which limited our ability to assess potential surveillance bias versus etiological effects in the observed education gradient. Second, a Pearson correlation test was used to examine the association of breast cancer incidence with several health system components. During this project, these components were only available for one year (2020) and limited to provincial level. A multivariate analysis using more comprehensive data for the health system over the years and possibly for smaller areas (i.e., district level) is warranted to provide further insight into these associations. In addition, most of the cancer registry data in Iran are pathology-based, which regularly provides beneficial information to improve cancer care systems, especially in low- and middle-income settings where there are no comprehensive population-based cancer registries [44]. However, pathology-based registries may underestimate the true incidence due to underdiagnosis in specific areas or subgroups of the population (e.g., urban areas). Access to alternative data sources or linkage to hospital records was not possible due to data availability. While this limitation may introduce some bias in the RR estimates, cancer registries remain the most consistent and comparable data source at subnational level. In addition, despite the use of national cancer registry data, the availability of data remains limited and sparse leading to potential bias in the estimates. In this regard, we developed a Bayesian Poisson model that allowed us to impute the gaps in the data and improve the estimate where the data were available.

Despite these limitations, our study fills a critical gap in subnational cancer surveillance literature, especially in low- and middle-income settings, where the evidence remains limited. Additionally, it highlights the geographical distribution of underserved populations where enhanced screening programs are essential to support the World Health Organization’s Global Breast Cancer Initiative (GBCI) goal of reducing age-standardized breast cancer mortality rates by 2.5% annually through 2040 [6].

Conclusions

To the best of our knowledge, this study is the first subnational-level analysis of the RR of breast cancer incidence in Iran that focuses on the district level over time and simultaneously uses several administrative datasets and Bayesian spatiotemporal modelling to obtain district-level estimations between 2000 and 2021 and addresses the incompleteness of the cancer registry. At the same time, with increasing incidence, a lower incidence was observed in the most deprived districts, possibly due to underdiagnosis or late-stage diagnosis. On the other hand, the incidence of breast cancer was much greater in districts with higher levels of female education than in those with the lowest level of female education.

Supporting information

S1 Fig. Map of Iran by district.

https://doi.org/10.1371/journal.pone.0330017.s001

(PNG)

S2 Fig. Estimated RR of breast cancer incidence (left side) versus SMR (right side) at the district level from 2000 to 2010.

Cancer registry data were not available in 2006, so the SMR map for 2006 is unavailable.

https://doi.org/10.1371/journal.pone.0330017.s002

(PDF)

S3 Fig. National average RR of breast cancer incidence categorised by year and YOS quintiles (each point represents a year).

https://doi.org/10.1371/journal.pone.0330017.s003

(PNG)

S4 Fig. RR of breast cancer incidence categorised by wealth index quintiles from 2000 to 2010 (each point represents a district).

Light blue shows the lowest WI quintile, and pink shows the highest WI quintile).

https://doi.org/10.1371/journal.pone.0330017.s004

(PNG)

S5 Fig. Projected RR of breast cancer incidence at the district level from 2011 to 2021.

https://doi.org/10.1371/journal.pone.0330017.s005

(PDF)

S6 Fig. Trend of RR of breast cancer incidence for each district from 2000–2021 in Iran (spatiotemporal model for 2000–2010 and projection for 2011–2021).

The grey shading shows the 95% credible intervals.

https://doi.org/10.1371/journal.pone.0330017.s006

(PDF)

S7 Fig. BGR diagnostics.

https://doi.org/10.1371/journal.pone.0330017.s007

(PDF)

S1 Table. List of provinces and districts in Iran.

https://doi.org/10.1371/journal.pone.0330017.s008

(DOCX)

S1 Appendix. Spatiotemporal model in OpenBUGS.

https://doi.org/10.1371/journal.pone.0330017.s009

(DOCX)

S2 Appendix. Model convergence and goodness of fit.

https://doi.org/10.1371/journal.pone.0330017.s010

(DOCX)

S2 Table. R-hat index for the model-specific parameters to check model convergence.

https://doi.org/10.1371/journal.pone.0330017.s011

(DOCX)

S3 Table. Posterior mean and 95% credible intervals for the RR of breast cancer incidence by district and year (sorted by mean values in 2010).

https://doi.org/10.1371/journal.pone.0330017.s012

(DOCX)

S4 Table. National average RR of breast cancer incidence from 2000 to 2021 (based on all posterior samples).

https://doi.org/10.1371/journal.pone.0330017.s013

(DOCX)

S5 Table. Posterior mean and 95% credible intervals for the RR of breast cancer incidence by district and year (sorted by mean values in 2021).

https://doi.org/10.1371/journal.pone.0330017.s014

(DOCX)

S6 Table. Results of cross-validation for the spatiotemporal model at the district level for RR of breast cancer incidence.

https://doi.org/10.1371/journal.pone.0330017.s015

(DOCX)

S1 Data. Aggregated data.

https://doi.org/10.1371/journal.pone.0330017.s016

(PDF)

Acknowledgments

We acknowledge OpenStreetMap https://www.openstreetmap.org/ which is an open-source website to extract Iran shapefiles at the district level.

References

1. Kuderer NM, Choueiri TK, Shah DP, Shyr Y, Rubinstein SM, Rivera DR, et al. Clinical impact of COVID-19 on patients with cancer (CCC19): a cohort study. Lancet. 2020;395(10241):1907–18. pmid:32473681
- View Article
- PubMed/NCBI
- Google Scholar
2. Maringe C, Spicer J, Morris M, Purushotham A, Nolte E, Sullivan R, et al. The impact of the COVID-19 pandemic on cancer deaths due to delays in diagnosis in England, UK: a national, population-based, modelling study. Lancet Oncol. 2020;21(8):1023–34. pmid:32702310
- View Article
- PubMed/NCBI
- Google Scholar
3. Sato Y, Fujiwara Y, Fukuda N, Hayama B, Ito Y, Ohno S, et al. Changes in treatment behavior during the COVID-19 pandemic among patients at a cancer hospital. Cancer Cell. 2021;39(2):130–1. pmid:33465338
- View Article
- PubMed/NCBI
- Google Scholar
4. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209–49. pmid:33538338
- View Article
- PubMed/NCBI
- Google Scholar
5. GBD 2017 Disease and Injury Incidence and Prevalence Collaborators. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2018;392:1789–858.
- View Article
- Google Scholar
6. Anderson BO, Ilbawi AM, Fidarova E, Weiderpass E, Stevens L, Abdel-Wahab M, et al. The Global Breast Cancer Initiative: a strategic collaboration to strengthen health care for non-communicable diseases. Lancet Oncol. 2021;22(5):578–81. pmid:33691141
- View Article
- PubMed/NCBI
- Google Scholar
7. The Global Breast Cancer Initiative. [Accessed 2023 July 22]. https://www.who.int/initiatives/global-breast-cancer-initiative
8. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394–424. pmid:30207593
- View Article
- PubMed/NCBI
- Google Scholar
9. Diez-Roux AV, Kiefe CI, Jacobs DR Jr, Haan M, Jackson SA, Nieto FJ, et al. Area characteristics and individual-level socioeconomic position indicators in three population-based epidemiologic studies. Ann Epidemiol. 2001;11(6):395–405. pmid:11454499
- View Article
- PubMed/NCBI
- Google Scholar
10. Schootman M, Sun D. Small-area incidence trends in breast cancer. Epidemiology. 2004;15(3):300–7. pmid:15097010
- View Article
- PubMed/NCBI
- Google Scholar
11. Ji P, Gong Y, Jin M-L, Hu X, Di G-H, Shao Z-M. The burden and trends of breast cancer from 1990 to 2017 at the global, regional, and national levels: results from the global burden of disease study 2017. Front Oncol. 2020;10:650. pmid:32528870
- View Article
- PubMed/NCBI
- Google Scholar
12. Farhood B, Geraily G, Alizadeh A. Incidence and mortality of various cancers in Iran and compare to other countries: a review article. Iran J Public Health. 2018;47:309–16.
- View Article
- Google Scholar
13. Global Burden of Disease Collaborative Network. Global Burden of Disease Study 2019 (GBD 2019) results. Seattle, United States: Institute for Health Metrics and Evaluation (IHME); 2020. https://vizhub.healthdata.org/gbd-results/
14. Tabrizi JS, Pourasghar F, Gholamzadeh Nikjoo R. Status of Iran’s primary health care system in terms of health systems control knobs: a review article. Iran J Public Health. 2017;46(9):1156–66. pmid:29026780
- View Article
- PubMed/NCBI
- Google Scholar
15. Rahimzadeh S, Burczynska B, Ahmadvand A, Sheidaei A, Khademioureh S, Pazhuheian F, et al. Geographical and socioeconomic inequalities in female breast cancer incidence and mortality in Iran: a Bayesian spatial analysis of registry data. PLoS One. 2021;16:e0248723.
- View Article
- Google Scholar
16. Knorr-Held L, Besag J. Modelling risk from a disease in time and space. Stat Med. 1998;17(18):2045–60. pmid:9789913
- View Article
- PubMed/NCBI
- Google Scholar
17. Blangiardo M, Cameletti M, Baio G, Rue H. Spatial and spatio-temporal models with R-INLA. Spat Spatiotemporal Epidemiol. 2013;7:39–55. pmid:24377114
- View Article
- PubMed/NCBI
- Google Scholar
18. Lawson AB. Bayesian disease mapping: hierarchical modeling in spatial epidemiology. 3rd ed. CRC Press; 2018.
19. Yu N, Donnan PT, Flynn RWV, Murphy MJ, Smith D, Rudman A. Increased mortality and morbidity in mild primary hyperparathyroid patients: the Parathyroid Epidemiology and Audit Research Study (PEARS). Clin Endocrinol (Oxf). 2010;73:30–4.
- View Article
- Google Scholar
20. Saccenti E, Hendriks MHWB, Smilde AK. Corruption of the Pearson correlation coefficient by measurement error and its estimation, bias, and correction under different error models. Sci Rep. 2020;10(1):438. pmid:31949233
- View Article
- PubMed/NCBI
- Google Scholar
21. Chechile RA. A Bayesian analysis for the Wilcoxon signed-rank statistic. Commun Statist Theory Methods. 2017;47(21):5241–54.
- View Article
- Google Scholar
22. Ataeinia B, Saeedi Moghaddam S, Shabani M, Gohari K, Sheidaei A, Rezaei N, et al. National and subnational incidence, mortality, and years of life lost due to breast cancer in Iran: trends and age-period-cohort analysis since 1990. Front Oncol. 2021;11:561376.
- View Article
- Google Scholar
23. Ng B, Puspitaningtyas H, Wiranata JA, Hutajulu SH, Widodo I, Anggorowati N, et al. Breast cancer incidence in Yogyakarta, Indonesia from 2008-2019: a cross-sectional study using trend analysis and geographical information system. PLoS One. 2023;18(7):e0288073. pmid:37406000
- View Article
- PubMed/NCBI
- Google Scholar
24. Fararouei M, Iqbal A, Rezaian S, Gheibi Z, Dianatinasab A, Shakarami S, et al. Dietary habits and physical activity are associated with the risk of Breast Cancer Among Young Iranian Women: a case-control study on 1010 premenopausal women. Clin Breast Cancer. 2019;19(1):e127–34. pmid:30503310
- View Article
- PubMed/NCBI
- Google Scholar
25. Vahid F, Hatami M, Sadeghi M, Ameri F, Faghfoori Z, Davoodi SH. The association between the Index of Nutritional Quality (INQ) and breast cancer and the evaluation of nutrient intake of breast cancer patients: a case-control study. Nutrition. 2018;45:11–6. pmid:29129231
- View Article
- PubMed/NCBI
- Google Scholar
26. Mobarakeh ZS, Mirzaei K, Hatmi N, Ebrahimi M, Dabiran S, Sotoudeh G. Dietary habits contributing to breast cancer risk among Iranian women. Asian Pac J Cancer Prev. 2014;15(21):9543–7. pmid:25422253
- View Article
- PubMed/NCBI
- Google Scholar
27. Namiranian N, Moradi-Lakeh M, Razavi-Ratki SK, Doayie M, Nojomi M. Risk factors of breast cancer in the Eastern Mediterranean Region: a systematic review and meta-analysis. Asian Pac J Cancer Prev. 2014;15(21):9535–41. pmid:25422252
- View Article
- PubMed/NCBI
- Google Scholar
28. Hosseinzadeh M, Eivazi Ziaei J, Mahdavi N, Aghajari P, Vahidi M, Fateh A, et al. Risk factors for breast cancer in Iranian women: a hospital-based case-control study in tabriz, iran. J Breast Cancer. 2014;17(3):236–43. pmid:25320621
- View Article
- PubMed/NCBI
- Google Scholar
29. Akbari A, Razzaghi Z, Homaee F, Khayamzadeh M, Movahedi M, Akbari ME. Parity and breastfeeding are preventive measures against breast cancer in Iranian women. Breast Cancer. 2011;18(1):51–5. pmid:20217489
- View Article
- PubMed/NCBI
- Google Scholar
30. Ghoncheh M, Mohammadian-Hafshejani A, Salehiniya H. Incidence and mortality of breast cancer and their relationship to development in Asia. Asian Pac J Cancer Prev. 2015;16(14):6081–7. pmid:26320499
- View Article
- PubMed/NCBI
- Google Scholar
31. Rezaei M, Jalilian A, Mahaki B, Veismoradi M. Spatial structure of breast cancer using Poisson generalized linear mixed model in Iran. Tehran Univ Med J. 2019;77:152–9.
- View Article
- Google Scholar
32. Coccia M. The effect of country wealth on incidence of breast cancer. Breast Cancer Res Treat. 2013;141(2):225–9. pmid:24036692
- View Article
- PubMed/NCBI
- Google Scholar
33. Unger-Saldaña K, Miranda A, Zarco-Espinosa G, Mainero-Ratchelous F, Bargalló-Rocha E, Miguel Lázaro-León J. Health system delay and its effect on clinical stage of breast cancer: multicenter study. Cancer. 2015;121(13):2198–206. pmid:25809536
- View Article
- PubMed/NCBI
- Google Scholar
34. Poum A, Promthet S, Duffy SW, Parkin DM. Factors associated with delayed diagnosis of breast cancer in northeast Thailand. J Epidemiol. 2014;24(2):102–8. pmid:24335087
- View Article
- PubMed/NCBI
- Google Scholar
35. Wheeler SB, Reeder-Hayes KE, Carey LA. Disparities in breast cancer treatment and outcomes: biological, social, and health system determinants and opportunities for research. Oncologist. 2013;18(9):986–93. pmid:23939284
- View Article
- PubMed/NCBI
- Google Scholar
36. Hu K, Lou L, Tian W, Pan T, Ye J, Zhang S. The outcome of breast cancer is associated with national human development index and health system attainment. PLoS One. 2016;11(7):e0158951. pmid:27391077
- View Article
- PubMed/NCBI
- Google Scholar
37. Trewin CB, Strand BH, Weedon-Fekjær H, Ursin G. Changing patterns of breast cancer incidence and mortality by education level over four decades in Norway, 1971–2009. Eur J Public Health. 2017;27:160–6.
- View Article
- Google Scholar
38. Sajadi HS, Ehsani-Chimeh E, Majdzadeh R. Universal health coverage in Iran: where we stand and how we can move forward. Med J Islam Repub Iran. 2019;33:9. pmid:31086788
- View Article
- PubMed/NCBI
- Google Scholar
39. Kokabisaghi F. Assessment of the effects of economic sanctions on Iranians’ right to health by using human rights impact assessment tool: a systematic review. Int J Health Policy Manag. 2018;7(5):374–93. pmid:29764102
- View Article
- PubMed/NCBI
- Google Scholar
40. Nazari M, Haghjou L, Rashidian H, Zendehdel K, Bahmanziari N, Haghjoo A, et al. Effects of economic sanctions on cancer care for patients in Iran: a qualitative study. BMC Health Serv Res. 2025;25(1):361. pmid:40065348
- View Article
- PubMed/NCBI
- Google Scholar
41. Hassani M. Impact of sanctions on cancer care in Iran. Arch Bone Jt Surg. 2018;6(4):248–9. pmid:30175170
- View Article
- PubMed/NCBI
- Google Scholar
42. GBD 2019 Iran Collaborators. Health system performance in Iran: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2022;399:1625–45.
- View Article
- Google Scholar
43. Maani N, Abdalla SM, Galea S. Avoiding a legacy of unequal non-communicable disease burden after the COVID-19 pandemic. Lancet Diabetes Endocrinol. 2021;9(3):133–5. pmid:33549161
- View Article
- PubMed/NCBI
- Google Scholar
44. Jedy-Agba EE, Curado M-P, Oga E, Samaila MO, Ezeome ER, Obiorah C, et al. The role of hospital-based cancer registries in low and middle income countries-The Nigerian Case Study. Cancer Epidemiol. 2012;36(5):430–5. pmid:22704971
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Kuderer NM, Choueiri TK, Shah DP, Shyr Y, Rubinstein SM, Rivera DR, et al. Clinical impact of COVID-19 on patients with cancer (CCC19): a cohort study. Lancet. 2020;395(10241):1907–18. pmid:32473681
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Maringe C, Spicer J, Morris M, Purushotham A, Nolte E, Sullivan R, et al. The impact of the COVID-19 pandemic on cancer deaths due to delays in diagnosis in England, UK: a national, population-based, modelling study. Lancet Oncol. 2020;21(8):1023–34. pmid:32702310
View Article
PubMed/NCBI
Google Scholar

[6] View Article

[7] PubMed/NCBI

[8] Google Scholar

[ref3] 3. Sato Y, Fujiwara Y, Fukuda N, Hayama B, Ito Y, Ohno S, et al. Changes in treatment behavior during the COVID-19 pandemic among patients at a cancer hospital. Cancer Cell. 2021;39(2):130–1. pmid:33465338
View Article
PubMed/NCBI
Google Scholar

[10] View Article

[11] PubMed/NCBI

[12] Google Scholar

[ref4] 4. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209–49. pmid:33538338
View Article
PubMed/NCBI
Google Scholar

[14] View Article

[15] PubMed/NCBI

[16] Google Scholar

[ref5] 5. GBD 2017 Disease and Injury Incidence and Prevalence Collaborators. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2018;392:1789–858.
View Article
Google Scholar

[18] View Article

[19] Google Scholar

[ref6] 6. Anderson BO, Ilbawi AM, Fidarova E, Weiderpass E, Stevens L, Abdel-Wahab M, et al. The Global Breast Cancer Initiative: a strategic collaboration to strengthen health care for non-communicable diseases. Lancet Oncol. 2021;22(5):578–81. pmid:33691141
View Article
PubMed/NCBI
Google Scholar

[21] View Article

[22] PubMed/NCBI

[23] Google Scholar

[ref7] 7. The Global Breast Cancer Initiative. [Accessed 2023 July 22]. https://www.who.int/initiatives/global-breast-cancer-initiative

[ref8] 8. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394–424. pmid:30207593
View Article
PubMed/NCBI
Google Scholar

[26] View Article

[27] PubMed/NCBI

[28] Google Scholar

[ref9] 9. Diez-Roux AV, Kiefe CI, Jacobs DR Jr, Haan M, Jackson SA, Nieto FJ, et al. Area characteristics and individual-level socioeconomic position indicators in three population-based epidemiologic studies. Ann Epidemiol. 2001;11(6):395–405. pmid:11454499
View Article
PubMed/NCBI
Google Scholar

[30] View Article

[31] PubMed/NCBI

[32] Google Scholar

[ref10] 10. Schootman M, Sun D. Small-area incidence trends in breast cancer. Epidemiology. 2004;15(3):300–7. pmid:15097010
View Article
PubMed/NCBI
Google Scholar

[34] View Article

[35] PubMed/NCBI

[36] Google Scholar

[ref11] 11. Ji P, Gong Y, Jin M-L, Hu X, Di G-H, Shao Z-M. The burden and trends of breast cancer from 1990 to 2017 at the global, regional, and national levels: results from the global burden of disease study 2017. Front Oncol. 2020;10:650. pmid:32528870
View Article
PubMed/NCBI
Google Scholar

[38] View Article

[39] PubMed/NCBI

[40] Google Scholar

[ref12] 12. Farhood B, Geraily G, Alizadeh A. Incidence and mortality of various cancers in Iran and compare to other countries: a review article. Iran J Public Health. 2018;47:309–16.
View Article
Google Scholar

[42] View Article

[43] Google Scholar

[ref13] 13. Global Burden of Disease Collaborative Network. Global Burden of Disease Study 2019 (GBD 2019) results. Seattle, United States: Institute for Health Metrics and Evaluation (IHME); 2020. https://vizhub.healthdata.org/gbd-results/

[ref14] 14. Tabrizi JS, Pourasghar F, Gholamzadeh Nikjoo R. Status of Iran’s primary health care system in terms of health systems control knobs: a review article. Iran J Public Health. 2017;46(9):1156–66. pmid:29026780
View Article
PubMed/NCBI
Google Scholar

[46] View Article

[47] PubMed/NCBI

[48] Google Scholar

[ref15] 15. Rahimzadeh S, Burczynska B, Ahmadvand A, Sheidaei A, Khademioureh S, Pazhuheian F, et al. Geographical and socioeconomic inequalities in female breast cancer incidence and mortality in Iran: a Bayesian spatial analysis of registry data. PLoS One. 2021;16:e0248723.
View Article
Google Scholar

[50] View Article

[51] Google Scholar

[ref16] 16. Knorr-Held L, Besag J. Modelling risk from a disease in time and space. Stat Med. 1998;17(18):2045–60. pmid:9789913
View Article
PubMed/NCBI
Google Scholar

[53] View Article

[54] PubMed/NCBI

[55] Google Scholar

[ref17] 17. Blangiardo M, Cameletti M, Baio G, Rue H. Spatial and spatio-temporal models with R-INLA. Spat Spatiotemporal Epidemiol. 2013;7:39–55. pmid:24377114
View Article
PubMed/NCBI
Google Scholar

[57] View Article

[58] PubMed/NCBI

[59] Google Scholar

[ref18] 18. Lawson AB. Bayesian disease mapping: hierarchical modeling in spatial epidemiology. 3rd ed. CRC Press; 2018.

[ref19] 19. Yu N, Donnan PT, Flynn RWV, Murphy MJ, Smith D, Rudman A. Increased mortality and morbidity in mild primary hyperparathyroid patients: the Parathyroid Epidemiology and Audit Research Study (PEARS). Clin Endocrinol (Oxf). 2010;73:30–4.
View Article
Google Scholar

[62] View Article

[63] Google Scholar

[ref20] 20. Saccenti E, Hendriks MHWB, Smilde AK. Corruption of the Pearson correlation coefficient by measurement error and its estimation, bias, and correction under different error models. Sci Rep. 2020;10(1):438. pmid:31949233
View Article
PubMed/NCBI
Google Scholar

[65] View Article

[66] PubMed/NCBI

[67] Google Scholar

[ref21] 21. Chechile RA. A Bayesian analysis for the Wilcoxon signed-rank statistic. Commun Statist Theory Methods. 2017;47(21):5241–54.
View Article
Google Scholar

[69] View Article

[70] Google Scholar

[ref22] 22. Ataeinia B, Saeedi Moghaddam S, Shabani M, Gohari K, Sheidaei A, Rezaei N, et al. National and subnational incidence, mortality, and years of life lost due to breast cancer in Iran: trends and age-period-cohort analysis since 1990. Front Oncol. 2021;11:561376.
View Article
Google Scholar

[72] View Article

[73] Google Scholar

[ref23] 23. Ng B, Puspitaningtyas H, Wiranata JA, Hutajulu SH, Widodo I, Anggorowati N, et al. Breast cancer incidence in Yogyakarta, Indonesia from 2008-2019: a cross-sectional study using trend analysis and geographical information system. PLoS One. 2023;18(7):e0288073. pmid:37406000
View Article
PubMed/NCBI
Google Scholar

[75] View Article

[76] PubMed/NCBI

[77] Google Scholar

[ref24] 24. Fararouei M, Iqbal A, Rezaian S, Gheibi Z, Dianatinasab A, Shakarami S, et al. Dietary habits and physical activity are associated with the risk of Breast Cancer Among Young Iranian Women: a case-control study on 1010 premenopausal women. Clin Breast Cancer. 2019;19(1):e127–34. pmid:30503310
View Article
PubMed/NCBI
Google Scholar

[79] View Article

[80] PubMed/NCBI

[81] Google Scholar

[ref25] 25. Vahid F, Hatami M, Sadeghi M, Ameri F, Faghfoori Z, Davoodi SH. The association between the Index of Nutritional Quality (INQ) and breast cancer and the evaluation of nutrient intake of breast cancer patients: a case-control study. Nutrition. 2018;45:11–6. pmid:29129231
View Article
PubMed/NCBI
Google Scholar

[83] View Article

[84] PubMed/NCBI

[85] Google Scholar

[ref26] 26. Mobarakeh ZS, Mirzaei K, Hatmi N, Ebrahimi M, Dabiran S, Sotoudeh G. Dietary habits contributing to breast cancer risk among Iranian women. Asian Pac J Cancer Prev. 2014;15(21):9543–7. pmid:25422253
View Article
PubMed/NCBI
Google Scholar

[87] View Article

[88] PubMed/NCBI

[89] Google Scholar

[ref27] 27. Namiranian N, Moradi-Lakeh M, Razavi-Ratki SK, Doayie M, Nojomi M. Risk factors of breast cancer in the Eastern Mediterranean Region: a systematic review and meta-analysis. Asian Pac J Cancer Prev. 2014;15(21):9535–41. pmid:25422252
View Article
PubMed/NCBI
Google Scholar

[91] View Article

[92] PubMed/NCBI

[93] Google Scholar

[ref28] 28. Hosseinzadeh M, Eivazi Ziaei J, Mahdavi N, Aghajari P, Vahidi M, Fateh A, et al. Risk factors for breast cancer in Iranian women: a hospital-based case-control study in tabriz, iran. J Breast Cancer. 2014;17(3):236–43. pmid:25320621
View Article
PubMed/NCBI
Google Scholar

[95] View Article

[96] PubMed/NCBI

[97] Google Scholar

[ref29] 29. Akbari A, Razzaghi Z, Homaee F, Khayamzadeh M, Movahedi M, Akbari ME. Parity and breastfeeding are preventive measures against breast cancer in Iranian women. Breast Cancer. 2011;18(1):51–5. pmid:20217489
View Article
PubMed/NCBI
Google Scholar

[99] View Article

[100] PubMed/NCBI

[101] Google Scholar

[ref30] 30. Ghoncheh M, Mohammadian-Hafshejani A, Salehiniya H. Incidence and mortality of breast cancer and their relationship to development in Asia. Asian Pac J Cancer Prev. 2015;16(14):6081–7. pmid:26320499
View Article
PubMed/NCBI
Google Scholar

[103] View Article

[104] PubMed/NCBI

[105] Google Scholar

[ref31] 31. Rezaei M, Jalilian A, Mahaki B, Veismoradi M. Spatial structure of breast cancer using Poisson generalized linear mixed model in Iran. Tehran Univ Med J. 2019;77:152–9.
View Article
Google Scholar

[107] View Article

[108] Google Scholar

[ref32] 32. Coccia M. The effect of country wealth on incidence of breast cancer. Breast Cancer Res Treat. 2013;141(2):225–9. pmid:24036692
View Article
PubMed/NCBI
Google Scholar

[110] View Article

[111] PubMed/NCBI

[112] Google Scholar

[ref33] 33. Unger-Saldaña K, Miranda A, Zarco-Espinosa G, Mainero-Ratchelous F, Bargalló-Rocha E, Miguel Lázaro-León J. Health system delay and its effect on clinical stage of breast cancer: multicenter study. Cancer. 2015;121(13):2198–206. pmid:25809536
View Article
PubMed/NCBI
Google Scholar

[114] View Article

[115] PubMed/NCBI

[116] Google Scholar

[ref34] 34. Poum A, Promthet S, Duffy SW, Parkin DM. Factors associated with delayed diagnosis of breast cancer in northeast Thailand. J Epidemiol. 2014;24(2):102–8. pmid:24335087
View Article
PubMed/NCBI
Google Scholar

[118] View Article

[119] PubMed/NCBI

[120] Google Scholar

[ref35] 35. Wheeler SB, Reeder-Hayes KE, Carey LA. Disparities in breast cancer treatment and outcomes: biological, social, and health system determinants and opportunities for research. Oncologist. 2013;18(9):986–93. pmid:23939284
View Article
PubMed/NCBI
Google Scholar

[122] View Article

[123] PubMed/NCBI

[124] Google Scholar

[ref36] 36. Hu K, Lou L, Tian W, Pan T, Ye J, Zhang S. The outcome of breast cancer is associated with national human development index and health system attainment. PLoS One. 2016;11(7):e0158951. pmid:27391077
View Article
PubMed/NCBI
Google Scholar

[126] View Article

[127] PubMed/NCBI

[128] Google Scholar

[ref37] 37. Trewin CB, Strand BH, Weedon-Fekjær H, Ursin G. Changing patterns of breast cancer incidence and mortality by education level over four decades in Norway, 1971–2009. Eur J Public Health. 2017;27:160–6.
View Article
Google Scholar

[130] View Article

[131] Google Scholar

[ref38] 38. Sajadi HS, Ehsani-Chimeh E, Majdzadeh R. Universal health coverage in Iran: where we stand and how we can move forward. Med J Islam Repub Iran. 2019;33:9. pmid:31086788
View Article
PubMed/NCBI
Google Scholar

[133] View Article

[134] PubMed/NCBI

[135] Google Scholar

[ref39] 39. Kokabisaghi F. Assessment of the effects of economic sanctions on Iranians’ right to health by using human rights impact assessment tool: a systematic review. Int J Health Policy Manag. 2018;7(5):374–93. pmid:29764102
View Article
PubMed/NCBI
Google Scholar

[137] View Article

[138] PubMed/NCBI

[139] Google Scholar

[ref40] 40. Nazari M, Haghjou L, Rashidian H, Zendehdel K, Bahmanziari N, Haghjoo A, et al. Effects of economic sanctions on cancer care for patients in Iran: a qualitative study. BMC Health Serv Res. 2025;25(1):361. pmid:40065348
View Article
PubMed/NCBI
Google Scholar

[141] View Article

[142] PubMed/NCBI

[143] Google Scholar

[ref41] 41. Hassani M. Impact of sanctions on cancer care in Iran. Arch Bone Jt Surg. 2018;6(4):248–9. pmid:30175170
View Article
PubMed/NCBI
Google Scholar

[145] View Article

[146] PubMed/NCBI

[147] Google Scholar

[ref42] 42. GBD 2019 Iran Collaborators. Health system performance in Iran: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2022;399:1625–45.
View Article
Google Scholar

[149] View Article

[150] Google Scholar

[ref43] 43. Maani N, Abdalla SM, Galea S. Avoiding a legacy of unequal non-communicable disease burden after the COVID-19 pandemic. Lancet Diabetes Endocrinol. 2021;9(3):133–5. pmid:33549161
View Article
PubMed/NCBI
Google Scholar

[152] View Article

[153] PubMed/NCBI

[154] Google Scholar

[ref44] 44. Jedy-Agba EE, Curado M-P, Oga E, Samaila MO, Ezeome ER, Obiorah C, et al. The role of hospital-based cancer registries in low and middle income countries-The Nigerian Case Study. Cancer Epidemiol. 2012;36(5):430–5. pmid:22704971
View Article
PubMed/NCBI
Google Scholar

[156] View Article

[157] PubMed/NCBI

[158] Google Scholar

Figures

Abstract

Introduction

Methods

Results

Conclusions

Introduction

Methods

Data

Statistical analysis

Model specification

Cross-validation

Ethics

Results

Bayesian spatiotemporal findings for relative risk of breast cancer incidence

Breast cancer estimations in relation to years of schooling and wealth index

Projections of the relative risk of breast cancer incidence between 2011 and 2021

Correlation of the RR of breast cancer incidence with health system components in 2020

Cross-validation results for the spatiotemporal model at the district level

Discussion

Limitations

Conclusions

Supporting information

S1 Fig. Map of Iran by district.

S2 Fig. Estimated RR of breast cancer incidence (left side) versus SMR (right side) at the district level from 2000 to 2010.

S3 Fig. National average RR of breast cancer incidence categorised by year and YOS quintiles (each point represents a year).

S4 Fig. RR of breast cancer incidence categorised by wealth index quintiles from 2000 to 2010 (each point represents a district).

S5 Fig. Projected RR of breast cancer incidence at the district level from 2011 to 2021.

S6 Fig. Trend of RR of breast cancer incidence for each district from 2000–2021 in Iran (spatiotemporal model for 2000–2010 and projection for 2011–2021).

S7 Fig. BGR diagnostics.

S1 Table. List of provinces and districts in Iran.

S1 Appendix. Spatiotemporal model in OpenBUGS.

S2 Appendix. Model convergence and goodness of fit.

S2 Table. R-hat index for the model-specific parameters to check model convergence.

S3 Table. Posterior mean and 95% credible intervals for the RR of breast cancer incidence by district and year (sorted by mean values in 2010).

S4 Table. National average RR of breast cancer incidence from 2000 to 2021 (based on all posterior samples).

S5 Table. Posterior mean and 95% credible intervals for the RR of breast cancer incidence by district and year (sorted by mean values in 2021).

S6 Table. Results of cross-validation for the spatiotemporal model at the district level for RR of breast cancer incidence.

S1 Data. Aggregated data.

Acknowledgments

References