Body Iron Store and its Association with Risk of First Episode of Spontaneous Lower Extremity Deep Vein Thrombosis/ Pulmonary Embolism: A Case-Control Study

Abstract

Background and objectives

The association between body iron stores and risk of deep vein thrombosis/ pulmonary embolism (DVT/ PE) has not been studied among Indian subjects. This study aimed to evaluate the same and also study the association between iron stores and recanalization of affected veins at week-12.

Methods

This Case-Control with follow-up study enrolled 85 consecutive adult (≥ 18 years) cases presenting with first episode of spontaneous, proximal lower extremity DVT/ PE and 170 age (± 3 years) and sex matched adult controls without DVT/ PE. Those with haemoglobin(Hb) < 9 g/dl, malignancies, serum creatinine ≥ 2 mg/dL, heart failure and concurrent infections/ inflammatory disorders were excluded. All participants underwent iron profile, serum ferritin light-chain (FtL) and hepcidin testing.

Results

Anaemia [OR = 2.3 (95% CI = 1.3–4.0), p = 0.001] and elevated RDW (RDW-CV > 15%) [OR = 2.3 (95% CI = 1.2–4.3), p = 0.012] were significantly associated with increased risk of DVT/ PE. Iron deficiency (ID, defined as serum ferritin < 30 µg/L, along with TSAT < 20%) was not associated with DVT/ PE risk [OR = 0.8 (95% CI = 0.4–1.7), p > 0.05]. Serum FtL in the highest quartile (> 75th centile) was associated with higher risk of DVT/ PE (OR = 5, 95% CI = 2.6–9.6) and levels < 25th centile with protection against DVT/ PE (OR = 0.1, 95% CI = 0.01–0.32), compared to levels between 25th and 75th centiles (referent range). Highest DVT/ PE risk was associated with FtL > 90th centile [OR≈12 (95% CI = 3.9–37.2)]. No associations were noted between serum hepcidin and DVT/ PE risk and ID and DVT recanalization at week-12.

Conclusion

Higher iron stores, rather than ID, were associated with increased risk of DVT/ PE among those with Hb ≥ 9 g/dL. Anaemia and elevated RDW were also associated with risk of DVT/ PE. ID was not associated with poorer DVT recanalization at week-12.

Introduction

Venous thromboembolism (VTE), which encompasses both deep vein thrombosis (DVT) and pulmonary embolism (PE), is estimated to develop in 1/1000 people/year and contribute to 60,000–100,000 deaths annually in the United States [1, 2]. DVT, which contributes to two-thirds of VTE cases [2], is a preventable cause of morbidity and mortality. Prior DVT, prolonged immobilisation, surgery/ trauma, pregnancy/ postpartum and malignancy are some of the well-known predisposing factors for DVT/ PE. About 30–50% of DVT patients have idiopathic or unprovoked DVT [3], wherein no identifiable predisposing factors are present. Factors such as ageing population and increasing burden of obesity and malignancy have contributed to increasing incidence of VTE in the general population [1]. New associations for VTE are being explored to unravel the underlying pathogenetic mechanisms. Identification of novel risk factors and risk predicting biomarkers may reduce the burden of VTE by improved risk prediction and prevention strategies.

Serum ferritin, though most commonly used test for assessment of body iron stores, has some limitations. Ferritin values are affected by inflammation and liver disease and it is less sensitive for diagnosing ID in individuals without anaemia [4]. Ferritin light chain (FtL), which is a component of the larger ferritin molecule, is important for iron uptake and storage and is less affected by inflammation compared to ferritin [5]. Hepcidin is important for maintenance of iron homeostasis and regulates enteral iron absorption and storage. Serum FtL and hepcidin levels, of late, are being used for assessment of body iron stores [5, 6]. Red cell distribution width (RDW), which is a measure of variability of erythrocyte volume (anisocytosis) in a given person, has been found to be associated with risk of VTE [3]. RDW, a parameter useful for classification of anaemia, is elevated in iron deficiency anaemia (IDA). IDA has been associated with new-onset VTE [7], VTE recurrences [8, 9], PE [10] and cerebral venous thrombosis [11] as well. Thus, iron deficiency (ID) may explain the association between elevated RDW and VTE risk. However, the pathogenetic mechanisms by which ID is associated with VTE risk are not clearly understood and it has been argued that other factors associated with IDA may be confounding the observed association between ID and VTE [5]. Contrary to previous findings, a recent Nordic study [5] has reported increased VTE risk associated with elevated serum hepcidin and FtL levels. The findings from the latter study suggest that higher body iron stores (as suggested by elevated serum FtL and hepcidin), rather than ID, are associated with VTE risk. Thus, there is conflicting literature on the association between body iron status and VTE risk. To the best of our knowledge, there are no studies from India evaluating the association between iron status and DVT/ PE risk. With this background, the primary objective of this Case-Control study was to study the association between body iron status and risk of first episode of spontaneous, proximal lower extremity DVT/ PE. Secondary objectives were to study the association between serum FtL, hepcidin and RDW and risk of DVT/ PE and also to assess DVT recanalization at week-12 and its association with iron status.

Materials and Methods

This Case-Control study with follow-up was carried out at a tertiary care teaching hospital located in southern India. The study was approved by the Institute Ethics Committee and written informed consent was obtained from all study participants before enrolment.

Patients presenting to medicine outpatient services or emergency services at our hospital between March 2020 to February 2022, with symptomatic lower extremity DVT and/or PE (DVT/ PE) were screened prospectively for eligibility to be included as cases in this study. Consecutive adult (≥ 18 years) patients presenting with first episode of spontaneous, proximal lower extremity DVT (involving popliteal vein or above)/ PE diagnosed ≤ 2 weeks from study enrolment were included as cases in this study. Recurrent DVT/ PE, haemoglobin concentration < 9 g/dL, known malignancy, liver disease, varicosities of lower extremity veins, serum creatinine ≥ 2 mg/dL, heart failure, concurrent infection/ inflammatory disorders, DVT/ PE following trauma or central venous catheter placement or hospitalization for other illnesses and polycythemia (haemoglobin > 16.5 g/dL in men and > 16.0 g/dL in women) were exclusion criteria for cases. Adults (age ≥ 18 years) without history of DVT/ PE or venous/ arterial thrombosis at other sites and with haemoglobin concentration ≥ 9 g/dl were chosen as controls for comparison of iron status. Controls had to be age (± 3 years) and sex matched to cases. Those with history of bleeding haemorrhoids, menorrhagia, worm infestations, concurrent infections/ inflammatory conditions, varicose veins and serum creatinine ≥ 2 mg/dL were ineligible to be enrolled as controls. The ratio of cases to controls enrolled was 1:2. To reduce selection bias, controls for DVT/ PE cases with comorbidities (such as diabetes, hypertension etc.) were chosen from medicine outpatient services and postmenopausal women controls were chosen for postmenopausal DVT/ PE cases. The rest of controls were selected from attendants of inpatients/ outpatients in our hospital and hospital staff. The diagnosis of DVT was confirmed among cases by compression ultrasonography (CUS) or colour duplex ultrasonography (CDU) and that of PE by computed tomography pulmonary angiography. The treatment and follow-up of DVT/ PE cases were as per standard recommendations[2]. DVT/ PE cases were followed up at week-12 with repeat CUS of the affected veins to assess for recanalization of thrombus.

Laboratory Evaluation

Complete blood counts were done for all participants with Sysmex XT-2000i automated haematology analyser (Sysmex corporation, Kobe, Japan). RDW was expressed as coefficient of variation from the mean (RDW-CV) and was calculated by the formula given below.

$$RDW-CV(\%)=(1StandardDeviationofredcellvolumex100)/(Meancorpuscularvolume)$$

Serum iron and unsaturated iron binding capacity (UIBC) were measured by spectrophotometry and serum ferritin by chemiluminescent assay, using Beckman Coulter™ analyser. Total iron binding capacity (TIBC) was calculated by adding serum iron and UIBC. Transferrin saturation (TSAT) was calculated as ratio of serum iron to TIBC. Serum FtL and hepcidin levels were measured by commercially available ELISA kits (Abbkine Scientific Co., Ltd).

Definitions Used in the Study

Anaemia was defined as haemoglobin concentration of < 13 g/dL in men and < 12 g/dL in women, as per the World Health Organization criteria[12]. Normal range of RDW-CV in our laboratory is 11–15% and a value > 15% was considered elevated. For the purpose of this study, to have a more specific diagnosis, ID for both men and women was defined as serum ferritin < 30 µg/L, along with TSAT < 20%.

Statistical Analysis

Assuming difference in prevalence of ID or high iron stores between cases and controls to be 20% and type I error of 5% and power of 80%, minimum number of cases estimated to be enrolled was 67 [using OpenEpi software (version-3.1)]. Chi-square/ Fisher’s exact tests were used for comparison of categorical variables and continuous variables were compared with student t test/ Mann Whitney U test. The correlation between serum ferritin, FtL and hepcidin levels among study participants was determined using bivariate correlations (Spearman’s rank correlation). Statistical analysis was done with Statistical Packages for Social Sciences software version 19.0.

Results:

A total of 85 DVT/ PE cases (54 men, 31 women) and 170 (108 men, 62 women) controls were enrolled (Fig. 1). Majority (59, ≈70%) of cases belonged to the age-group of 30–59 years. Among 85 cases, 75 presented with DVT and 10 with symptomatic PE. Among 10 cases presenting with symptomatic PE, six showed evidence of lower extremity DVT on CUS/CDU. Thus, among 81 cases with proximal lower extremity DVT, 48 had left lower limb involvement, 24 right lower limb and nine had bilateral involvement. None of the women with DVT/ PE had received oral contraceptives/ hormonal preparations. Clinical and laboratory characteristics of study participants are shown in Table 1. Barring six postpartum women participants, none of the others were on iron replacement at the time of enrolment.

Fig. 1.

Shows enrolment of participants and follow-up of cases

Table 1.

Shows clinical and laboratory parameters among cases and controls

Clinical/laboratory parameter	Cases (N = 85)	Controls (N = 170)	p Value
Age (years)—mean (± SD)	43.2 (± 15)	43.3 (± 15.2)	–
Family history of DVT/ PE—No. (%)	1	0	–
Smoking-No. (%)	19 (22.4)	27 (15.9)	0.205*
Heavy alcohol intakea—No. (%)	12 (14.1)	26 (15.3)	0.804*
Diabetes Mellitus—No. (%)	22 (25.9)	33 (19.4)	0.236*
Hypertension—No. (%)	12 (14.1)	17 (10)	0.329*
Postpartum stateb—No. (%)	3 (9.6)	3 (4.8)	0.397¥
Obesity (BMI ≥ 25 kg/m2)—No. (%)	29 (34.1%)	27 (15.9%)	0.001*, OR = 2.7 (95% CI: 1.4–5.0)
Hemoglobin concentration (g/dL)—median (IQR)	12.2 (10.5–14.2)	13.2 (12–14.1)	0.012$
RDW-CV (%)—median (IQR)	14.2 (13.4–15.1)	13.5 (12.7–14.4)	< 0.001$
Serum iron (µg/dL)—median (IQR)	49 (30–70)	73 (51–100)	< 0.001$
Serum ferritin (µg/L)—median (IQR)	116.4 (65.7–276.2)	74.8 (37.3–112.7)	< 0.001$
Serum ferritin light chain (ng/mL)—median (IQR)	2.98 (2.76–3.30)	2.30 (1.78–2.71)	< 0.001$
Serum hepcidin (ng/mL)—median (IQR)	22.5 (20.5–25.4)	22.9 (11–27.5)	0.114$

BMI body mass index; IQR interquartile range; RDW-CV red cell distribution width- coefficient of variation; SD standard deviation

^aAlcohol intake more than thrice/ week

^bPercentages among women

*Using Chi-square test

^¥Using Fisher’s exact test

^$With Mann–Whitney U test

Correlation Between Serum Ferritin, Ferritin Light Chain and Hepcidin Levels Among Study Participants (N = 255)

Serum ferritin and FtL levels had a weak, but statistically significant positive correlation on applying Spearman’s correlation [r_s = 0.26 (95% CI = 0.19–0.31), p < 0.001]. Serum FtL and hepcidin also had a moderate positive correlation which was statistically significant [r_s = 0.49 (95% CI = 0.44–0.55), p < 0.001]. However, no significant correlation was found between serum ferritin and hepcidin levels [r_s = 0.09 (95% CI = 0.03–0.16), p > 0.05].

Anaemia, RDW and Iron Deficiency Among Study Participants

Anaemia was more prevalent among DVT/ PE cases than controls [51.8% vs 31.2%, Table 2]. Thus, anaemia was significantly associated with risk of DVT/ PE [odds ratio (OR) = 2.3 (95% CI = 1.3–4.0), p = 0.001]. Cases had significantly higher RDW-CV values compared to the control group (Table 1). Elevated RDW-CV of > 15% was noted in 23 (27.1%) of cases and 24 (14.1%) of controls. Elevated RDW was significantly associated with risk of DVT/ PE [OR = 2.3 (95% CI = 1.2–4.3), p = 0.012]. Prevalence of ID among cases and controls was 14% and 16.5% respectively (Table 2). Thus, ID was not associated with risk of DVT/ PE [OR = 0.8 (95% CI = 0.4–1.7), p = 0.63]. About a fourth (12/44, 27.3%) of anaemic cases and half (28/53, ≈53%) of anaemic controls fulfilled the criteria for ID.

Table 2.

Shows prevalence of anaemia and iron deficiency among cases and controls

Sex	Cases (N = 85)		Controls (N = 170)		Cases (N = 85)		Controls (N = 170)
	Anaemiaa	No anaemia	Anaemiab	No anaemia	ID	No ID	ID	No ID
Men	21	33	27	81	5	49	9	99
Women	23	8	26	36	7	24	19	43
Total- No. (%)	44 (51.8)	41 (48.2)	53 (31.2)	117 (68.8)	12 (14.1)	73 (85.9)	28 (16.5)	142 (83.5)

ID iron deficiency

^a31 (70.5%) of cases with anaemia had normocytic normochromic, 12 (27.3%) microcytic hypochromic and 1 had macrocytic anaemia

^b33 (62.2%) of controls with anaemia had normocytic normochromic, 18 (33.9%) microcytic hypochromic and 2 (3.8%) had macrocytic anaemia

Serum Ferritin Light Chain (FtL) and Hepcidin Levels and Risk of DVT/ PE

Table 3 shows the serum FtL and hepcidin among 255 study participants. Those with serum FtL in the highest quartile (> 75^th centile) had higher risk of DVT/ PE (OR = 5, 95% CI = 2.6–9.6) and the ones with FtL in the lowest quartile (< 25^th centile) were protected against DVT/ PE (OR = 0.1, 95% CI = 0.01–0.3), compared to the referent group having FtL between 25th and 75th centiles. Further, serum FtL chain > 90^th centile was associated with the highest risk of DVT/ PE [OR ≈12 (95% CI: 3.9–37.2), Table 4]. However, there was no association between serum hepcidin levels and DVT/ PE risk.

Table 3.

Shows serum ferritin light chain and hepcidin levels among study participants (N = 255)

Percentiles	Range (ng/mL)	Cases (N = 85)	Controls (N = 170)	Odds ratio (95% confidence intervals)
Ser. FtL
< 25th	0.57–1.96	2	61	0.07 (0.01–0.32)
25th-75th	1.98–2.97	39	89	Referent
> 75th	2.97–4.16	44	20	5.02 (2.6—9.6)
Ser. hepcidin
< 25th	0.55–11.98	2	61	0.02 (0.01 – 0.11)
25th-75th	11.99–26.24	71	57	Referent
> 75th	26.31- 73.62	12	52	0.18 (0.09 – 0.38)

Table 4.

Shows serum ferritin light chain levels among study participants (N = 255)

Percentiles	Range (ng/mL)	Cases (N = 85)	Controls (N = 170)	Odds ratio (95% confidence intervals)
Ser. FtL
< 10th	0.57–1.64	1	24	0.09 (0.01–0.73)
< 25th	0.57–1.96	2	61	0.07 (0.01–0.32)
25th-75th	1.98–2.97	39	89	Referent
> 75th	2.98–4.16	44	20	5.02 (2.6—9.6)
> 90th	3.31–4.16	21	4	11.98 (3.86–37.22)

Recanalization of Affected Lower Extremity Deep Veins at Week-12 Among DVT/ PE Cases

Among the 68 DVT/ PE cases who could be followed up and reassessed at week-12 for recanalization of affected lower extremity veins (Fig. 1), 21 (30.9%) had completely recanalized veins, 19 (27.9%) partially recanalized veins and the remaining 28 (41.2%) still had completely occluded veins. Though ID was more common among cases who achieved partial/ complete recanalization of affected veins compared to those who still had occluded veins [8/40 (20%) vs. 1/28 (3.5%), OR = 6.8 (95% CI = 0.8–57.4)], the difference did not achieve statistical significance (p = 0.07, Fisher’s exact test). Thus, ID was not associated with DVT recanalization at week-12 in this study.

Discussion

In this study, iron status among participants was evaluated using multiple parameters such as serum ferritin, TSAT, FtL and hepcidin levels. On expected lines, significant positive correlations were noted between serum ferritin and FtL and also serum FtL and hepcidin levels. A previous study [5] has also reported significant positive correlation between serum FtL and hepcidin. However, unlike another Indian study [13], no correlation was found in this study between serum ferritin and hepcidin levels. The former study involved anaemic children (half had IDA and other half anaemia of chronic disease) and only 38% of participants had anaemia in the present study. Differences in the age-group and anaemia prevalence among enrolled participants may explain the observed differences in the correlation between serum ferritin and hepcidin in the two studies.

All participants enrolled in this study had haemoglobin of ≥ 9 g/dL. Despite this selection bias, anaemia was significantly associated with risk of DVT/ PE (OR = 2.3). Another study involving acutely ill, hospitalised medical patients has reported that anaemia was independently associated with risk of symptomatic VTE, despite thromboprophylaxis [14]. Further, association has been reported between anaemia and cerebral venous thrombosis as well [11, 15].

The ID criteria used in this study reflect the emphasis on a more specific diagnosis of ID and were similar to those used in previous studies [8, 9]. Contrary to previous reports of association of ID with first or recurrent episodes of VTE [7–9], no association could be found between ID and DVT/ PE risk in this study. Further, higher body iron stores (i.e., iron stores in the upper part of normal range, as suggested by higher serum FtL levels), were associated with increased risk of DVT/ PE. A Nordic study [5] has also reported association between higher serum hepcidin levels (suggesting higher iron stores) and increased risk of VTE. Higher body iron stores may lead to increased oxidative and nitrosative stress [16], endothelial dysfunction (by reduced bioavailability of nitric oxide) and also to platelet activation [5]. These mechanisms may explain increased risk of VTE associated with higher body iron stores. It is interesting to note that though anaemia burden was significantly higher among cases compared to controls, ID prevalence was not significantly different between cases and controls in this study. Nearly three-fourths of DVT/ PE cases with anaemia did not fulfil criteria for ID and had normocytic normochromic blood picture in this study (Table 2). This emphasizes that anaemia per se, and not ID, was associated with risk of DVT/ PE. Thus, as for as clinical practice is considered, the implication from this study is that correction of ID per se (with or without anaemia) may not be of help in reducing the burden of DVT/ PE.

RDW values were significantly higher among cases compared to controls and elevated RDW (RDW-CV > 15%) was associated with increased risk of DVT/ PE (OR = 2.2) in this study. Previous studies have also reported association between elevated RDW and increased VTE risk[3, 17] and also worser outcomes after VTE[18, 19]. Though RDW-CV was significantly higher among cases compared to controls, there was no significant difference in the prevalence of ID between them in this study. Thus, as has been reported previously[5], elevated RDW among DVT/ PE cases in this study could not be explained by ID. Apart from ID, other factors impairing erythropoiesis or affecting erythrocyte survival such as vitamin B12/ folic acid deficiencies, oxidative stress, shortened telomere length and alteration of erythropoietin function can also account for elevated RDW[20]. These factors might account for elevated RDW and also partly its association with risk of VTE in the present study.

As it has been reported that residual venous obstruction after treatment is associated with increased risk of recurrent VTE [21], we sought to evaluate recanalization of affected deep veins at week-12. The recanalization at week-12 in present study (≈ 59% had complete/ partial recanalization) was comparable to 40–53% recanalization reported at three months with warfarin treatment in previous studies [22, 23]. A previous study has reported that ID is associated with risk of VTE recurrences [8]. However, no association could be found between ID and recanalization at week-12 in the present study. There are no previous studies evaluating the association between iron status and DVT recanalization as such.

Limitations

This was a hospital-based Case-Control study. Hence, selection and confounding bias can’t be ruled out from the present study. Though infections and inflammatory conditions were excluded before study enrolment, because of financial constraints, markers of inflammation such as serum C reactive protein could not be studied. Confounding factors such as hyperhomocysteinemia, which is associated with risk of DVT/ PE [24, 25] and vitamin B12 and folic acid deficiencies (which give rise to hyperhomocysteinemia) could not be evaluated among participants because of financial constraints. Data on effectiveness of oral anticoagulation with warfarin and the time spent with international normalized ratio (INR) maintained in the therapeutic range (between 2.0 and 3.0) was not available for DVT/ PE patients. Adequacy of anticoagulation is also a determinant of recanalization.

Conclusion

Among those with haemoglobin concentration ≥ 9 g/dL, there was no association between ID and risk of DVT/ PE. Rather, there was association between higher serum FtL levels and DVT/ PE risk, suggesting that higher body iron stores (i.e., stores in the upper part of normal range) are associated with DVT/ PE risk. Further, there was no association between ID and recanalization of affected lower extremity deep veins at week-12. Anaemia and elevated RDW-CV (> 15%) were significantly associated with increased risk of DVT/ PE.

Abbreviations

CI

Confidence intervals

CUS

Compression ultrasonography

CDU

Colour duplex ultrasound

DVT

Deep vein thrombosis

ELISA

Enzyme linked immunosorbent assay

FtL

Ferritin light chain levels

OR

Odds ratio

ID

Iron deficiency

IDA

Iron deficiency anaemia

PE

Pulmonary embolism

RDW-CV

Red cell distribution width-coefficient of variation

TIBC

Total iron binding capacity

UIBC

Unsaturated iron binding capacity

TSAT

Serum transferrin saturation

VTE

Venous thromboembolism

Author contributions

RD Study design; methodology; investigations, data collection; formal analysis; statistical analysis; writing—original draft; reviewing and editing and preparing final version of the draft. VKV: concept; study design; methodology; formal analysis; statistical analysis; funding acquisition; project administration; writing—original draft; reviewing and editing and preparing final version of the draft. PSA: methodology; investigation (performing biochemical tests); reviewing and editing and preparing final version of the draft. All authors have reviewed and approve the final version of the manuscript submitted to the journal.

Funding

This work was supported by Intramural Research Grant from Jawaharlal Institute of Postgraduate Medical Education and Research (JIPMER), Puducherry, India. Intramural grant Reference No.: JIP/Res/Intramural/phs 2/2020–21/Sl. No. 2, dated 23rd Dec. 2020.

Declarations

Conflict of interest

The authors have no relevant financial or non-financial interests to disclose.

Ethical approval

All procedures performed in this study involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The study was approved by the Institutional Ethics Committee (Human Research) of Jawaharlal Institute of Postgraduate Medical Education and Research (JIPMER), Pondicherry. [Ethics Committee Approval Ref. No.: JIP/IEC/2019/549, approved on 20th February 2020].

Informed consent

Informed written consent has been obtained from all participants of this study and we declare that anonymity of participant data has been maintained, ensuring that patient confidentiality has been ensured.