Skip to main content
Download PDF

Dietary iron intake is nonlinearly associated with the risk of diabetic retinopathy in adults with type 2 diabetes

BMC Endocrine Disorders volume 25, Article number: 102 (2025) Cite this article

Abstract

Objective

To elucidate the association between dietary iron intake and diabetic retinopathy (DR) in type 2 diabetes (T2D) patients.

Methods

Participants from the National Health and Nutrition Examination Survey (NHANES) 2005–2008 aged over 40 years with T2D were included. Dietary iron intake was estimated from standardised questionnaires. The presence of DR and vision-threatening DR (VTDR) was determined through retinal imaging. We used logistic regression to assess the relationship between iron intake and DR, and restricted cubic splines to reveal nonlinear links.

Results

The study enrolled 1172 T2D adults. We found significant nonlinear associations between dietary iron intake and DR among females (P = 0.023), but not in males (P = 0.490). Compared with the lowest quartile of iron intake, the third quartile (13.2–18.1 mg/d) yielded significantly lower odds of developing DR (odds ratio [OR], 0.59; 95% CI, 0.39–0.90) and VTDR (OR, 0.42; 95% CI, 0.19–0.94). Stratified logistic analyses showed that medium-high iron intake was associated with lower risks of DR in females (OR, 0.44; 95% CI, 0.24–0.81), non-Hispanic Blacks (OR, 0.38; 95% CI, 0.17–0.85), and individuals with obesity (OR, 0.45; 95% CI, 0.25–0.82), high HbA1c (OR, 0.56; 95% CI, 0.34–0.93), long diabetes duration (OR, 0.40; 95% CI, 0.21–0.76) or low blood haemoglobin (OR, 0.17; 95% CI, 0.05–0.60).

Conclusion

Dietary iron intake was nonlinearly negatively associated with the prevalence of DR and VTDR, showing protective effect against retinopathy of medium-high iron intake in T2D patients. Such associations significantly vary by multiple factors such as age, ethnicity, obesity and glycaemic control.

Peer Review reports

Introduction

Diabetic retinopathy (DR) is one of the major causes of vision impairment globally [1]. According to the Global Burden of Disease Study 2019, DR has been the fifth leading cause of blindness and moderate-to-worse visual impairment among individuals aged 50 years and greater [2]. Of note, it was the only cause of blindness with a globally increasing trend in age-standardised prevalence from 1990 to 2020 [2]. Among individuals with diabetes, approximately a third have any sign of DR, and a third of them might have vision-threatening diabetic retinopathy (VTDR) [3]. With the estimated global diabetes prevalence to be rising from 9.3% (463 million) in 2019 to 10.9% (700 million) by 2045 [4], along with longer life expectancy and lifestyle changes, the global burden of DR is expected to grow rapidly. Although hyperglycaemia, diabetes duration, and hypertension are considered as major risk factors for DR, they merely account for a small amount of the variation in the DR risk [5]. Several novel pathogeneses have raised that the abnormal homeostasis of trace elements including iron may associate with the progression of DR [6]. However, little is known about the association between dietary iron intake and DR.

Iron plays a critical role in dynamic redox balance, disruption of which can lead to oxidative stress and damage to target organs including the retina [7, 8]. Nevertheless, previous studies showed contradictory results regarding whether iron could protect the retina from harmful stimuli. On one hand, iron overload can exacerbate the development of retinopathy in mice, implying that iron depletion strategies may ameliorate diabetic microvascular complications [9, 10]. On the other hand, epidemiologic and clinical studies showed that iron deficiency anaemia (IDA) was related to increased risk of DR [11, 12], indicating the efficacy of anaemia treatment in decreasing retinopathy risk in type 2 diabetes (T2D). Such findings were supported by a recent study [13], in which serum iron was negatively correlated with the occurrence of DR in diabetic adults. Notably, all previous studies concentrated on the relationship of body iron status rather than dietary intake of iron and DR. The current assessment of iron status is derived from a battery of hemaetological indicators, including serum ferritin, transferrin saturation, and mean corpuscular volume, etc. However, measuring these hemaetological indicators is invasive and might be infeasible as for cost and complexity of procedure in certain settings. Thus, there is a need for a simpler and more accessible method to evaluate iron status. A complementary and easy-to-use option is to consider iron intake from diet and dietary supplements. When associations between DR risks and iron intake through nutritional assessment could be accurately estimated, it is promising to establish recommendations for appropriate dietary iron intake as to prevent DR in T2D patients. Nutritional interventions with optimal combinations of iron and other nutrients that support conventional therapies could be developed to reduce disease risk and severity in T2D patients.

Herein, we investigated the association between the level of dietary iron intake, measured by a standardised questionnaire, and the risk of DR and VTDR among patients with T2D in a series of national representative samples. We further assessed whether such associations varied in specific subgroups of population defined by major stratification factors.

Materials and methods

Study design and population

This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting statement (Supplemental Material 1). We conducted this cross-sectional study using the data from the National Health and Nutrition Examination Survey (NHANES) 2005 to 2008. NHANES is a population-based, multipurpose survey to evaluate the health and nutritional status of the US population. The National Centre for Health Statistics (NCHS) ethics review board reviewed and approved the study protocol. The study was conducted adhered to the principles of the Declaration of Helsinki, and written informed consent was obtained from each participant. Detailed data and methodology files are accessible online [14]. All potentially identifiable information has been removed to ensure the confidentiality of participants and their households. Briefly, every 2-year cycle of survey comprises of questionnaires administered at home, and a standardised physical examination in a mobile examination centre (MEC) including physical measurements and collection of biospecimens for laboratory tests. For 2005–2006 and 2007–2008 cycles, retinal photography and dietary interview were conducted to subjects older than 40 years, making this analysis available. Herein, T2D was defined according to participants’ meeting one or more of the following criteria: (1) self-reported physician diagnosis of diabetes; (2) using oral glucose-lowering medicines or insulin; and (3) fasting plasma glucose level of at least 126 mg/dL, or haemoglobin A1c (HbA1c) level of at least 6.5% [15]. In this study, only those with complete data of dietary iron intake, diabetes and retinopathy were eligible.

Patient and public involvement

Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.

Assessment of dietary iron intake

All NHANES examinees were eligible for two 24-hour dietary recall interviews. The first was performed in-person in the MECs, and the second by telephone 3 to 10 days later. Intakes of energy, nutrients, and other components from foods and beverages were estimated from the two detailed dietary interview. Detailed questionnaire and methodology of dietary intake calculation is publicly accessible online [14]. In this study, dietary intake of iron was calculated by averaging data of two 24-hour recalls if available, otherwise the single reliable dietary recall data was used. The continuous iron intake data were further dichotomised as adequate and inadequate categories according to the Recommended Dietary Allowance (RDA) issued by the Food and Nutrition Board of the National Academies [16].

Ascertainment of diabetic retinopathy

During 2005–2008 NHANES survey, the presence of major retinal diseases including DR was tested for participants aged over 40 years by the Retinal Imaging. Two 45-degree digital retinal images for each eye were captured utilising the Canon Non-Mydriatic Retinal Camera CR6-45NM (Canon, Tokyo, Japan), one focused on the optic nerve and the other on the macula. The digital images were transferred to the University of Wisconsin Ocular Epidemiologic Reading Centre, Madison for grading retinopathies according to the standardised protocol. At least two raters graded a same set of images. If the first two graders disagreed, the third graded the image. If two of three disagreed, an adjudicator would make a final decision.

DR severity was broadly classified into 4 levels: no DR, mild non-proliferative diabetic retinopathy (NPDR), moderate/severe NPDR, and proliferative diabetic retinopathy (PDR) according to the Early Treatment Diabetic Retinopathy Study (ETDRS) classification standards [17]. In this study, we aggregated these 4 levels as no DR versus any DR (incorporating mild NPDR, moderate/severe NPDR, and PDR). We also defined VTDR as the presence of severe NPDR, PDR, or clinically significant macular oedema (CSME). Herein CSME was defined when (1) the oedema involved the fovea or within 500 microns of the fovea, and/or (2) a 1 + disc area of oedema present with at least a portion of it involving the macula. Outcomes of our study were defined according to the worse one of two eyes.

Assessment of covariates

Sociodemographic variables, including age, sex, race/ethnicity, education attainment, and poverty income ratio (PIR) was collected through a questionnaire. Race/ethnicity was self-reported according to NCHS categories (Mexican American, non-Hispanic Black, non-Hispanic White, other Hispanic, or other). We combined other Hispanic and other race/ethnicity groups as the broader “other group”. Education was grouped into three categories: (1) less than high school, (2) high school or equivalent, and (3) greater than high school. PIR was classified into 3 categories: less than 1.30, 1.30 to 3.49, and 3.5 or higher. Body mass index (BMI) was calculated as weight in kilograms divided by height in metres squared, and was categorised as three groups: normal or underweight (less than 25.0 kg/m2), overweight (25.0 to 30.0 kg/m2), and obese (greater than 30.0 kg/m2). Smoking status was classified into 3 categories: never (less than 100 cigarettes in lifetime), former (greater than 100 cigarettes in lifetime, but had given up at the time of interview), and current smoker (greater than 100 cigarettes in lifetime and currently smoking). Alcohol consumption was classified into 3 categories: never (never drank alcohol in lifetime and the past 12 months), former (had ever drunk in lifetime, but not in the past 12 months), and current drinker (drunk at least 12 alcoholic drinks during the past year or lifetime and consumed alcohol at least 1 day during the past year).

Duration of diabetes was self-reported by interviewees, and was dichotomised as < 10 and ≥ 10 years. Haemoglobin A1c (HbA1c) raw data were continuous, and were further grouped into < 7.0% versus ≥ 7.0%. Blood haemoglobin level was dichotomised as normal/high (> 12 g/dL) and low (< 12 g/dL) based on the standard of World Health Organization [18].

Information about medical comorbidities was obtained from physical examination and questionnaire. We defined hypertension if participants with (1) diastolic blood pressure ≥ 80 mmHg, or systolic blood pressure ≥ 130 mmHg according to the mean value of 3 measurements, or (2) self-reported hypertension history, or (3) taking blood pressure medications [19]. Individuals whose total cholesterol ≥ 240 mg/dL (6.2mmol/L), or taking lipid-lowering medications were considered as having hypercholesteraemia. Congestive heart failure, coronary heart disease, heart attack, angina/angina pectoris, and stroke were confirmed based on self-reported physician diagnosis. The full questionnaire can be found in the Supplemental Material 2.

Statistical analysis

In this study, we reported descriptive statistics as numbers (percentages) for categorical variables, and means (standard deviations, SDs) for continuous variables. We used χ2 and unpaired t-tests to assess differences in sociodemographic, clinical and dietary characteristics between groups. We established logistic regression models to evaluate odds ratios (ORs) of DR and VTDR based on iron intake quartiles after adjustment for demographic variables, lifestyle variables, DM related variables, and medical comorbidities.

To explore nonlinear association between DR and iron intake, we used restricted cubic spline (RCS) logistic regression analyses after adjusting the abovementioned confounders. We set four knots (5th, 35th, 65th, and 95th percentiles of iron intake) according to Harrell’s recommendation that four knots can offer an adequate fit of the model and well balance flexibility and imprecision caused by overfitting [20]. The R package plotRCS (version 0.1.4) was used to visualise splines [21]. Logistic regression and RCS analysis were stratified by sex, race/ethnicity, weight status, HbA1c level (< 7.0% or ≥ 7.0%), duration of diabetes (< 10 years or ≥ 10 years), and blood haemoglobin level (normal/high or low). We performed sensitivity analyses to evaluate the robustness of major results by applying the following strategies: (1) excluding participants with outlying iron intake data (≥ 3SD from the mean value), (2) reconstructing logistic regression and RCS models through ignoring missing data rather than multiple imputation (i.e., complete case analysis). All data analyses were performed by R statistical package (R Core Team, Vienna, Austria, version 4.2.3). Variables with missing values were imputed through multiple imputation approach using mice package (version 3.15.0) [22]. P value less than 0.05 was considered to be statistically significant.

Results

Participant characteristics

Of all participants from the NHANES 2005–2006 (n = 10348) and 2007–2008 (n = 10149), 18,923 participants younger than 40 years (n = 13416) or without diabetes (n = 5507) were excluded. Participants with ungradable images (n = 374) or without dietary iron intake data (n = 28) were excluded as well, leaving 1172 participants ultimately included (Fig. 1). Nine covariates of the dataset had missing values over 1% of observations (Supplemental Figure S1), and PIR had the highest proportion (8%). To take into account the effect of missing data, we handled missing data with multiple imputation technique for the subsequent analyses.

Fig. 1
figure 1

Flow diagram of participant enrolment of this study

Full size image

The general characteristics of study population by DR status were presented in Table 1. There were 358 participants (30.5%) with any DR, among whom 188 were males (52.5%). Compared with disease-free individuals, DR patients had comparable average age, sex, education, marriage status, PIR, smoking status, and alcohol consumption, but were more likely to be non-Hispanic black, Mexican American, and have higher BMI. DR population showed poorer general health condition, longer diabetes duration, higher levels of HbA1c, lower levels of blood haemoglobin, as well as higher prevalence of systemic comorbidities including congestive heart failure, heart attack and stroke (all P < 0.05, Table 1). The mean dietary iron intake of subjects with DR was 13.2 ± 6.24 mg/d, which was significantly lower than that of subjects without DR (14.3 ± 6.76 mg/d, P = 0.008). However, the proportion of participants meeting the sufficient daily iron intake was comparable between two groups (23.5% vs. 19.8%, P = 0.176, Table 1). When analysed according to the severity of DR, the iron intake in the VTDR subgroup was significantly lower than that in the groups without DR and without VTDR (P = 0.007, Supplemental Figure S2).

Table 1 Characteristics of the study population according to status of diabetic retinopathy
Full size table

Association of dietary iron intake and DR and VTDR

Table 2 presents the results of multivariable logistic regression analyses. In crude models, participants in the third and fourth quartiles showed significantly lower risk of DR as compared with the first (reference) group (quartile 3: OR = 0.57; 95%CI, 0.40–0.82; quartile 4: OR = 0.68; 95%CI, 0.48–0.96). After adjustment for multiple covariates, significant associations remained for the third quartile in all logistic regression models (Table 2; for the saturated model: OR = 0.59, 95%CI, 0.39–0.90). Besides, no significant relationship was shown between the second quartile and the reference group in any models (all P > 0.05). When dietary iron intake was dichotomised as adequate versus inadequate by RDA, no significant relation was found between iron intake and DR (Table 2, all P > 0.05).

Compared with the first quartile, all other quartiles yielded significantly lower odds of VTDR in crude models (quartile 2: OR = 0.47; 95%CI, 0.25–0.89; quartile 3: OR = 0.43; 95%CI, 0.22–0.83; quartile 4: OR = 0.44; 95%CI, 0.22–0.86). With adjustment of covariates, this relationship remained significant for the third quartile in all models (Table 2; for the saturated model: OR = 0.42, 95%CI, 0.19–0.94). Nonetheless, correlation between VTDR risk and the fourth quartile of iron intake became insignificant after multiple adjustments (all P > 0.05). Inadequate intake of iron according to RDA showed no added risk of VTDR in either crude or multi-adjusted models (all P > 0.05).

Table 2 Association of dietary iron intake with diabetic retinopathy and vision-threatening diabetic retinopathy
Full size table

Stratified analyses for association of dietary iron intake and DR

Table 3 demonstrated the associations between dietary iron and DR by various stratifying factors, including sex, race/ethnicity, weight status, HbA1c level, duration of diabetes, and blood haemoglobin level. Compared with the first quartile, subjects within the third quartile yielded significantly lower odds of DR in female, non-Hispanic Black, obese, HbA1c ≥ 6.5%, and DM duration ≥ 10 years groups. Remarkably, for individuals with low blood haemoglobin, both the third and the fourth quartiles showed decreased risks of DR (quartile 3: OR = 0.17; 95%CI, 0.05–0.60; quartile 4: OR = 0.23; 95%CI, 0.06–0.90), but were insignificant in those with normal or high blood haemoglobin. No significant interaction between any stratification factors and dietary iron intake was found (Table 3, all P > 0.05).

Table 3 Stratified analyses for association of dietary iron with diabetic retinopathy*
Full size table

Nonlinear relationship between dietary iron intake and DR

The non-monotonic relationship revealed by logistic regression analysis promoted us to further investigate the nonlinear relationship of DR and dietary iron intake. RCS analyses showed nonlinear associations between DR and dietary iron intake varied by sex (Fig. 2). In males, no significant relationship was found between dietary iron and DR risk (Poverall=0.230, Pnonlinear=0.490) (Fig. 2A). In females, however, an approximately U-shaped relationship was revealed; only moderate level of dietary iron intake was associated with decreased risk of DR (Poverall=0.022, Pnonlinear=0.023) (Fig. 2B). No significant nonlinear association was found in subgroups by stratification factors other than sex (Supplemental Figure S3).

Fig. 2
figure 2

Associations between dietary iron intake and diabetic retinopathy by sex using restricted cubic spline model. A. Male; B. Female. Graphs show odds ratio (OR) for any DR according to dietary iron level adjusted for age, race, education, marital status, poverty income ratio, body mass index, smoking status, alcohol consumption, duration of diabetes, HbA1c level, hypertension, hypercholesterolaemia, history of comorbidities, including congestive heart failure, heart attack and stroke. Data were fitted by a logistic regression model, and the model was conducted with 4 knots at the 5th, 35th, 65th, 95th percentiles of iron intake (reference is the 5th percentile). Solid lines indicate ORs, and shadow shape indicate 95% CIs. OR, odds ratio; CI, confidence interval

Full size image

Sensitivity analyses

Sensitivity analyses showed the robustness of our major results. The association between DR and quartiles of iron intake did not substantially change when participants with outlying data were excluded (Supplemental Table S1 for any DR, Supplemental Table S2 for VTDR). Such associations remained significant when we removed missing data instead of adopting multiple imputation (Supplemental Table S3 for any DR, Supplemental Table S4 for VTDR). This indicated that the outliers did not significantly deflate or inflate the mean of the sample, and had minimal influence on the association derived from the mean. Similarly, when excluding the participants with outlying data (Supplemental Figure S4) or ignoring missing data (Supplemental Figure S5), the RCS analysis results did not substantially change, indicating that missing data caused little noise or bias to estimation.

Discussion

In this large-scale, nationally representative T2D cohort, we found significantly lower daily dietary iron intake in individuals with DR, particularly those with VTDR. After adjustment for major confounding factors, medium-high level of dietary iron intake (13.2–18.1 mg/d) was associated with 59% risk reduction for DR, and 42% risk reduction for VTDR. Of note, beneficial effects of adequate iron intake were more profound in several specific subpopulations, including females, non-Hispanic Blacks, individuals with longer diabetes duration, higher level of HbA1c, concurrent obesity, or anaemia. Spline regression analysis demonstrated that there was a nonlinear U-relationship between the daily iron intake amount and DR risk in females, but not in males. Sensitivity analyses confirmed the robustness of the above findings.

As one of the essential minerals, iron is vital for maintaining the normal structures and functions of a number of macromolecules in cells. Dysregulation of iron homeostasis, either excess or deficiency, might lead to a variety of chronic diseases including diabetes. In patients with pre-existing diabetes, iron deficiency anaemia (IDA) can exacerbate retinopathy through inducing long-term hypoxia in the retina [23]. Additionally, increased lipid peroxidation induced by IDA could elevate HbA1c level [24], which is a strong indicator to predict the onset and progression of DR. Luckily, elevated HbA1c in T2D patients with IDA could be ameliorated by 3-month iron supplementation therapy [25]. Iron overload, however, generates various oxygen and nitrogen species via Fenton reaction, which is one of the major causative factors for diabetes and its complications [26]. Recently, an animal study demonstrated that excessive iron can exacerbate the development of DR by increasing retinal renin expression in mice [9]. Another study also verified that iron accumulation induced the diabetic-related pericyte loss in eyes and aggravated diabetic microvascular complications [10]. Collectively, ensuring a balanced iron status in the body is critical for preventing the occurrence and progression of diabetic ocular complications. Our findings in this study were well consistent with this concept, and for the first time we revealed possible beneficial effects of medium-high dietary iron intake on preventing DR and VTDR; neither higher nor lower amount was significantly associated with the occurrence of DR.

Another interesting finding of this study is that sex can potentially modify the relationship between dietary iron intake and DR. Both multivariate logistic regression and spline analysis models demonstrated that medium-high daily iron intake was associated with a reduced risk of DR in females but not in males. The underlying mechanism is difficult to interpret but may be related to sex differences in iron homeostasis under the influences of hormonal, genetic, and dietary factors. Females have lower iron storages than males because of menstruation. On average, females lose about 20–60 mg of iron per menstrual cycle, an amount comparable to daily dietary iron intake requirements for males. Besides, differences in dietary habit by sex may also affect iron absorption: men consume more haem iron, while women consume more non-haem iron. Haem iron is more available for absorption from the diet than non-haem iron, which may substantially affect iron status and health outcomes [27].

The stratified analysis in this study found that factors other than sex may also modify the relation between dietary iron intake and DR. Participants who are non-Hispanic Black, obese, with low haemoglobin levels, with poorer glycaemic control (HbA1c ≥ 6.5%) and having a longer duration of diabetes (≥ 10 years) may benefit more from sufficient iron intake than others. A possible reason is that T2D patients with obesity, of non-Hispanic Black ethnicity, and unsatisfactory glycaemic control are more likely to have concurrent iron deficiency or IDA [28,29,30]. In response, our study also found a significant inverse association between dietary iron intake and the risk of DR in T2D patients with anaemia. Given that anaemia is an established risk factor for DR [11, 12], it is particularly important for the abovementioned at-risk populations to routinely evaluate and maintain sufficient iron intakes.

The findings of our study are relevant for clinical practice, and provide implications for future research as well. To help prevent DR, we encourage adults with T2D to consume a diet with sufficient iron. The optimal amount of daily iron intake showing protective effects on DR (13.2–18.1 mg/d) is higher than the RDA for male adults (8 mg/d) and close to that for premenopausal women (18 mg/d). As this range of daily iron intake is far below the upper limits established by the Food and Nutrition Board (FNB) for adults (< 45 mg/d), it poses very little risk of iron overload and toxicity. Higher iron intake than 18.2 mg/d may not be recommended because excessive iron intake is not associated with further reduced DR risk, but may cause other health problems like neurological disorders. Moreover, due to the nonlinear relationship between dietary iron and DR, our study also indicated that dichotomised cut-off values of RDA may not be capable of guiding iron intake for diabetes patients in terms of minimising retinopathy risk. In the future, a more sophisticated “protective dietary pattern” incorporating iron intake against diabetic retinopathy is to be developed and validated.

There are several strengths of our study. Compared with previous studies (Supplemental Table S5), this is the first study to illustrate the nonlinear relationships between dietary iron intake and DR using spline analyses, a powerful technique to delineate the nonlinear nature of many phenomena in clinical research. Moreover, the NHANES data are high-quality and of great representativeness which ensured the generalisability of our findings and conclusions. The comprehensive survey data allowed us to adjust potential confounders in statistical models as well.

Nevertheless, this study has several limitations. Firstly, due to the cross-sectional design, we only identified correlation rather than causation between dietary iron intake and DR. Further dietary trials are needed to establish causality and to test the efficacy of dietary iron interventions. Secondly, the dietary data in NHANES were acquired by two 24-hour recalls, largely depending on participants’ memory. Recall bias could not be excluded, and the actual daily nutrient intake level might slightly differ from the self-reported data. Thirdly, due to the unavailability of data on supplement intake of iron, our study only included dietary iron intake but not on metal supplements. Fourthly, although we adjusted for a comprehensive range of confounding factors, residual or unknown confounding cannot be entirely excluded. Fifthly, cultural and dietary differences across the population have the potential to impact the findings. Sixthly, although the sample population is highly representative and with large size, the prevalence of VTDR was relatively low which may have contributed to the non-statistically significant results.

In summary, our study found a nonlinear association between dietary iron intake and DR risk. Medium-high level of iron intake was associated with a reduced risk of DR and VTDR, especially for females, non-Hispanic Blacks, obese people, those with HbA1c ≥ 6.5%, and diabetes duration ≥ 10 years. High or low levels of iron intake may not be conducive to preventing the development of DR. A precise efficacy of dietary iron intervention strategy and its impact on DR and VTDR are to be determined in longitudinal studies and controlled trials.

Data availability

All original data related to this study are available in the public, open access NHANES data repository (https://wwwn.cdc.gov/nchs/nhanes/Default.aspx).

References

  1. Wong TY, Cheung CM, Larsen M, Sharma S, Simó R. Diabetic retinopathy. Nat Rev Dis Primers. 2016;2:16012.

    Article  PubMed  Google Scholar 

  2. Collaborators GBVI. Causes of blindness and vision impairment in 2020 and trends over 30 years, and prevalence of avoidable blindness in relation to VISION 2020: the right to sight: an analysis for the global burden of disease study. Lancet Glob Health. 2021;9(2):e144–60.

    Article  Google Scholar 

  3. Yau JW, Rogers SL, Kawasaki R, Lamoureux EL, Kowalski JW, Bek T, Chen SJ, Dekker JM, Fletcher A, Grauslund J, et al. Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care. 2012;35(3):556–64.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Saeedi P, Petersohn I, Salpea P, Malanda B, Karuranga S, Unwin N, Colagiuri S, Guariguata L, Motala AA, Ogurtsova K, et al. Global and regional diabetes prevalence estimates for 2019 and projections for 2030 and 2045: results from the international diabetes federation diabetes atlas, 9(th) edition. Diabetes Res Clin Pract. 2019;157:107843.

    Article  PubMed  Google Scholar 

  5. Tan GS, Ikram MK, Wong TY. Traditional and novel risk factors of diabetic retinopathy and research challenges. Curr Med Chem. 2013;20(26):3189–99.

    Article  CAS  PubMed  Google Scholar 

  6. Harrison AV, Lorenzo FR, McClain DA. Iron and the pathophysiology of diabetes. Annu Rev Physiol. 2023;85:339–62.

    Article  CAS  PubMed  Google Scholar 

  7. Kang Q, Yang C. Oxidative stress and diabetic retinopathy: molecular mechanisms, pathogenetic role and therapeutic implications. Redox Biol. 2020;37:101799.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Galaris D, Barbouti A, Pantopoulos K. Iron homeostasis and oxidative stress: an intimate relationship. Biochim Biophys Acta Mol Cell Res. 2019;1866(12):118535.

    Article  CAS  PubMed  Google Scholar 

  9. Chaudhary K, Promsote W, Ananth S, Veeranan-Karmegam R, Tawfik A, Arjunan P, Martin P, Smith SB, Thangaraju M, Kisselev O, et al. Iron overload accelerates the progression of diabetic retinopathy in association with increased retinal Renin expression. Sci Rep. 2018;8(1):3025.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Altamura S, Müdder K, Schlotterer A, Fleming T, Heidenreich E, Qiu R, Hammes HP, Nawroth P, Muckenthaler MU. Iron aggravates hepatic insulin resistance in the absence of inflammation in a novel Db/db mouse model with iron overload. Mol Metab. 2021;51:101235.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Wang J, Xin X, Luo W, Wang R, Wang X, Si S, Mo M, Shao B, Wang S, Shen Y, et al. Anemia and diabetic kidney disease had joint effect on diabetic retinopathy among patients with type 2 diabetes. Invest Ophthalmol Vis Sci. 2020;61(14):25.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Chung JO, Park SY, Chung DJ, Chung MY. Relationship between anemia, serum bilirubin concentrations, and diabetic retinopathy in individuals with type 2 diabetes. Med (Baltim). 2019;98(43):e17693.

    Article  CAS  Google Scholar 

  13. Chen YJ, Chen JT, Tai MC, Liang CM, Chen YY, Chen WL. Serum iron and risk of diabetic retinopathy. Nutrients 2020, 12(8).

  14. NHANES Survey Methods and Analytic Guidelines. [https://wwwn.cdc.gov/nchs/nhanes/analyticguidelines.aspx]

  15. Association AD. 2. Classification and diagnosis of diabetes: standards of medical care in Diabetes-2020. Diabetes Care. 2020;43(Suppl 1):S14–31.

    Article  Google Scholar 

  16. Iron - Health. Professional Fact Sheet [https://ods.od.nih.gov/factsheets/Iron-HealthProfessional/]

  17. Zhang X, Saaddine JB, Chou CF, Cotch MF, Cheng YJ, Geiss LS, Gregg EW, Albright AL, Klein BE, Klein R. Prevalence of diabetic retinopathy in the united States, 2005–2008. JAMA. 2010;304(6):649–56.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Anaemia - World Health Organization. (WHO) [https://www.who.int/health-topics/anaemia]

  19. Whelton PK, Carey RM, Aronow WS, Casey DE Jr., Collins KJ, Dennison Himmelfarb C, DePalma SM, Gidding S, Jamerson KA, Jones DW et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation 2018, 138(17):e426-e483.

  20. Harrell FE. Regression modeling strategies: with applications to linear models, logistic regression, and survival analysis. New York: Springer-; 2010.

    Google Scholar 

  21. rcsplot. Plot restricted cubic splines curves [https://cran.r-project.org/web/packages/plotRCS/index.html]

  22. van Buuren S, Groothuis-Oudshoorn K. Mice: multivariate imputation by chained equations in R. J Stat Softw. 2011;45(3):1–67.

    Article  Google Scholar 

  23. Lee MK, Han KD, Lee JH, Sohn SY, Jeong JS, Kim MK, Baek KH, Song KH, Kwon HS. High hemoglobin levels are associated with decreased risk of diabetic retinopathy in Korean type 2 diabetes. Sci Rep. 2018;8(1):5538.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Balamurugan R, Selvaraj N, Bobby Z, Sathiyapriya V. Increased glycated hemoglobin level in non-diabetic nephrotic children is associated with oxidative stress. Indian J Physiol Pharmacol. 2007;51(2):153–9.

    CAS  PubMed  Google Scholar 

  25. Tarim O, Küçükerdoğan A, Günay U, Eralp O, Ercan I. Effects of iron deficiency anemia on hemoglobin A1c in type 1 diabetes mellitus. Pediatr Int. 1999;41(4):357–62.

    Article  CAS  PubMed  Google Scholar 

  26. Halliwell B, Gutteridge JM. Oxygen toxicity, oxygen radicals, transition metals and disease. Biochem J. 1984;219(1):1–14.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Abbaspour N, Hurrell R, Kelishadi R. Review on iron and its importance for human health. J Res Med Sci. 2014;19(2):164–74.

    PubMed  PubMed Central  Google Scholar 

  28. Qiu F, Wu L, Yang G, Zhang C, Liu X, Sun X, Chen X, Wang N. The role of iron metabolism in chronic diseases related to obesity. Mol Med. 2022;28(1):130.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Le CH. The prevalence of anemia and Moderate-Severe anemia in the US population (NHANES 2003–2012). PLoS ONE. 2016;11(11):e0166635.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Guo W, Zhou Q, Jia Y, Xu J. Increased levels of glycated hemoglobin A1c and iron deficiency anemia: A review. Med Sci Monit. 2019;25:8371–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

None.

Funding

This study was supported by the Guangdong Basic and Applied Basic Research Foundation (2023A1515012192, 2023A1515030108).

Author information

Author notes

  1. Xiaoyun Chen, Yihang Fu these authors contribute to this study equally.

Authors and Affiliations

  1. State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Centre, Guangdong Provincial Key Laboratory of Ophthalmology and Vision Science, Guangdong Provincial Clinical Research Centre for Ocular Diseases, Sun Yat- Sen University, Guangzhou, China

    Xiaoyun Chen, Yihang Fu, Hongyu Si, Wenfei Li, Weimin Yang & Wei Xiao

Authors
  1. Xiaoyun Chen
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Yihang Fu
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Hongyu Si
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Wenfei Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Weimin Yang
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Wei Xiao
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

X.C. and W.X. was responsible for the concept, design, and supervision. Y.F. and H.S. collected the data from the database. W.L. and W.Y. conducted data analyses and visualisation. Y.F. and H.S. wrote the original draft. X.C. and W.X. revised the manuscript critically. All authors reviewed the manuscript.

Corresponding authors

Correspondence to Xiaoyun Chen or Wei Xiao.

Ethics declarations

Ethical approval

The NHANES protocol was approved by the National Center for Health Statistic (NCHS) Research Ethics Review Board (Protocol #2005-06 for NHANES 2005–2006; Continuation of Protocol #2005-06 for NHANES 2007–2008). The Ethics Review Board approval information is accessible at: https://www.cdc.gov/nchs/nhanes/irba98.htm (accessed on 8 August, 2024).

Competing interests

The authors declare no competing interests.

Clinical trial number

Not applicable.

Patient consent for publication

Not applicable.

Additional contributions

None reported.

Role of the funder/sponsor

The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Chen, X., Fu, Y., Si, H. et al. Dietary iron intake is nonlinearly associated with the risk of diabetic retinopathy in adults with type 2 diabetes. BMC Endocr Disord 25, 102 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12902-025-01926-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12902-025-01926-z

Keywords

BMC Endocrine Disorders

ISSN: 1472-6823

Contact us