**4. Iron deficiency anemia and cardiovascular risk**

Oxidative stress results from disequilibrium between oxidants and antioxidants. While iron excess may be involved in the generation of ROS, as commented above, anemia due to iron deficiency anemia (IDA) may affect the functioning of many enzymatic systems (cytochromes, catalases, hydroxylases, etc.) related to immunity, antioxidant status, and DNA integrity, among others [56].

In this regard, Aslan et al. [57] compared total plasma antioxidant capacity and lymphocyte DNA damage between two groups of IDA and control adults and concluded that both oxidative stress and DNA damage increased in IDA. In another study, four groups were compared: patients recently diagnosed with IDA who were not receiving any treatment at the beginning of the study; patients with IDA at the sixth week of an iron-replacement program (considered the time of hemoglobin normalization); patients with IDA at the end of the iron-replacement treatment (time of saturation of body iron stores); and age- and sex-matched healthy controls. Results show that untreated IDA patients present high lipid peroxidation, assessed by plasma malondialdehyde, and low activities of the antioxidant enzymes glutathione peroxidase, superoxide dismutase, and catalase and that the values did not differ between the sixth week and the end of the treatment, suggesting that recovery from IDA reduces oxidative stress [58]. Unfortunately, in these studies, the changes within a patient were not analyzed. In animal models of IDA, where all experimental conditions are controlled, high oxidative stress and DNA damage were not demonstrated [59, 60].

Another aspect that has been studied is the possible changes in lipid levels in IDA. Old animal studies reported dyslipidemia with altered triglycerides and total cholesterol in serum. However, in most animal experiments, there were important confounders. For instance, iron deficiency induces low appetite and a reduction in food intake that often was not adequately controlled. In this regard, there are inconsistent results, but clearly the direction of change was toward reduction in circulating lipids and profound modifications in lipoprotein metabolism [61].

in HH and most of the C282Y +/+ have no iron overload phenotype or are asymptomatic until adulthood. Cash et al. [30] compared vascular function, biochemical endothelial markers, and antioxidant status between HH patients (C282Y homozygous and high serum-ferritin levels) and controls. Noninvasive pulse wave analysis and pulse wave velocity were applied to carotid and radial arteries to estimate endothelial dysfunction. They reported that male HH patients had higher pulse wave velocity; however, this effect disappeared after adjusting for hypertension. In both sexes, HH was associated with diminished antioxidant levels but neither increase in lipid peroxidation nor alteration in the systemic inflammation marker (i.e., C-reactive protein) could be demonstrated [30]. Thus, controversy remains on the idea

Iron overload is usually a complication of thalassemia, particularly in patients who require red blood cell transfusions. Among the three types of thalassemia, thalassemia intermedia is characterized by ineffective erythropoiesis, anemia, medullary expansion, and extramedullary hematopoiesis. In contrast to HH, thalassemic patients show a proatherogenic biochemi-

There are other iron overload pathological situations, where bone medulla is inefficient and iron overload results from repeated transfusions. In such syndromes, the amount of body iron can reach very high values and myocardial damage is the most frequent collateral effect of the treatment. In these cases, iron chelation therapy may result in higher quality of life and

Oxidative stress results from disequilibrium between oxidants and antioxidants. While iron excess may be involved in the generation of ROS, as commented above, anemia due to iron deficiency anemia (IDA) may affect the functioning of many enzymatic systems (cytochromes, catalases, hydroxylases, etc.) related to immunity, antioxidant status, and DNA integrity,

In this regard, Aslan et al. [57] compared total plasma antioxidant capacity and lymphocyte DNA damage between two groups of IDA and control adults and concluded that both oxidative stress and DNA damage increased in IDA. In another study, four groups were compared: patients recently diagnosed with IDA who were not receiving any treatment at the beginning of the study; patients with IDA at the sixth week of an iron-replacement program (considered the time of hemoglobin normalization); patients with IDA at the end of the iron-replacement treatment (time of saturation of body iron stores); and age- and sex-matched healthy controls. Results show that untreated IDA patients present high lipid peroxidation, assessed by plasma malondialdehyde, and low activities of the antioxidant enzymes glutathione peroxidase, superoxide dismutase, and catalase and that the values did not differ between the sixth week and the end of the treatment, suggesting that recovery from IDA reduces oxidative stress [58]. Unfortunately, in these studies, the changes within a patient were not analyzed. In animal models of IDA, where all experimental conditions are controlled, high oxidative stress and DNA damage were not demonstrated [59, 60].

that cardiovascular risk is high in hemochromatosis patients [54].

**4. Iron deficiency anemia and cardiovascular risk**

reduction of cardiac events [55].

104 Recent Trends in Cardiovascular Risks

among others [56].

cal phenotype which may contribute to enhance cardiovascular risks [31].

In humans, our research group found low values of total cholesterol, HDL-cholesterol, glucose, and uric acid in IDA women at fertile age, which significantly increased during anemia recovery [36]. These results coincide with that of others who also reported low serum triglycerides [62] and LDL-cholesterol [63]. It is noteworthy that despite significant increases after treatment, the observed lipid values were very low in the severe anemic patients from these studies and still did not reach levels of non-anemic controls after recovery (reported mean values were approximately 150–170 mg/dL for total cholesterol and 60–70 mg/dL for triglycerides).

The above results can be explained by inhibition of lipid biosynthesis due to iron deficiency. Kamei et al. [64] performed a transcriptome analysis to determine the effects of iron deficiency on hepatic gene expression. Rats on an iron-deficient diet were compared with rats pair-fed a control diet with a normal iron level. In agreement with human studies, these authors observed that iron deficiency decreases cholesterol and triglycerides in serum and liver. In addition, they found that serum glucose and insulin increased. Expressions of genes encoding gluconeogenic enzymes were upregulated, lactate was increased, and the urea cycle was activated. These results are explained by the insufficiency of iron for its enzymatic functions and the situation of hypoxia due to anemia. Nevertheless, the results of high glucose and insulin do not agree with human observations.

Iron deficiency may affect cardiovascular health by indirect mechanisms. In this regard, iron participates in the hydroxylation of vitamin D to the active metabolites, 25 hydroxyvitamin D and 1,25-hydroxyvitamin D, and vitamin D acts as an antioxidant and may have protective cardiovascular effects, decreasing LDL-cholesterol, and blood pressure [37]. Moreover, iron supplementation alone increases vitamin B12 and folic acid levels [36]. This is attributed to a general increase in intestinal mucosa that favors nutrient absorption.

## **5. Dysmetabolic iron, type 2 diabetes, and cardiometabolic alterations**

### **5.1. Iron excess and type 2 diabetes mellitus (T2DM)**

T2DM is the most common and an ever-increasing form of diabetes [65]. It is characterized by disorders in insulin secretion or action either of which maybe the predominant feature. The association between iron overload and T2DM came from the observation that the frequency of diabetes is increased in classic hereditary hemochromatosis [65]. A link between red meat consumption, one of the highest iron bioavailability source, and T2DM has been reported [66]. Moreover, some reports show a relationship between high ferritin and the risk of gestational diabetes [67].

The positive association between iron excess and T2DM is feasible although the underlying mechanisms still remain to be fully determined. First, iron is a powerful pro-oxidant and catalyst molecule, which promotes the formation of hydroxyl radicals and could attack pancreatic β cells by increasing oxidative stress thus resulting in impaired insulin synthesis and secretion [68]. Second, iron excess can diminish insulin utilization in muscle tissue leading to a shift from glucose to free acid oxidation, which may result in enhanced insulin resistance [69]. Third, increasing free fatty acid, main substrate for hepatic gluconeogenesis, would provoke higher glucose production [69]. Thus, the possible mechanisms are insulin deficiency, insulin resistance, and hepatic dysfunction [70].

Several studies report an association between the heme iron intake and risk of T2DM. The prospective cohort within the Nurses' Health Study found that higher intake of heme iron was associated with higher intake of fat (total and saturated), red meat, and protein and with lower intake of carbohydrates. However, the association was not entirely explained by the red meat intake. Total dietary iron, non-heme iron, or supplemental iron were not related to diabetes risk [71].

Other studies reveal that vegetarians have higher insulin sensitivity than omnivores, and this was mainly attributed to their lower body iron [72]. In this regard, blood donation, by reducing iron stores, may increase insulin sensitivity [72, 73]. However, there is controversy in this issue [74–76].

We have studied some cardiovascular risk markers in a population of 595 T2DM from the DIabetes and CArdiovascular RIsk VAllecas (DICARIVA) study according to ferritin levels (**Table 1**).

Diabetic dyslipidemia is a cluster of altered plasma lipids and lipoproteins [77] though LDL-cholesterol levels are normal or reduced. It is characterized by high triglycerides and low HDL-cholesterol levels and by increased number of small and dense LDL particles [77]. Altogether, these features are known as the lipidic triad. In addition, other alterations are often observed:



**Table 1.** Male and female ferritin levels (ng/mL) in type 2 diabetes population belonging to the DIabetes and CArdiovascular RIsk VAllecas (DICARIVA) study P25 and P75, 25th and 75th percentiles.

• Higher clearance of apo A1 with decrease in the high-size HDLs and decrease of the cholesterol reverse transport.

The positive association between iron excess and T2DM is feasible although the underlying mechanisms still remain to be fully determined. First, iron is a powerful pro-oxidant and catalyst molecule, which promotes the formation of hydroxyl radicals and could attack pancreatic β cells by increasing oxidative stress thus resulting in impaired insulin synthesis and secretion [68]. Second, iron excess can diminish insulin utilization in muscle tissue leading to a shift from glucose to free acid oxidation, which may result in enhanced insulin resistance [69]. Third, increasing free fatty acid, main substrate for hepatic gluconeogenesis, would provoke higher glucose production [69]. Thus, the possible mechanisms are insulin deficiency, insulin

Several studies report an association between the heme iron intake and risk of T2DM. The prospective cohort within the Nurses' Health Study found that higher intake of heme iron was associated with higher intake of fat (total and saturated), red meat, and protein and with lower intake of carbohydrates. However, the association was not entirely explained by the red meat intake. Total dietary iron, non-heme iron, or supplemental iron were not related to

Other studies reveal that vegetarians have higher insulin sensitivity than omnivores, and this was mainly attributed to their lower body iron [72]. In this regard, blood donation, by reducing iron stores, may increase insulin sensitivity [72, 73]. However, there is controversy in this

We have studied some cardiovascular risk markers in a population of 595 T2DM from the DIabetes and CArdiovascular RIsk VAllecas (DICARIVA) study according to ferritin levels

Diabetic dyslipidemia is a cluster of altered plasma lipids and lipoproteins [77] though LDL-cholesterol levels are normal or reduced. It is characterized by high triglycerides and low HDL-cholesterol levels and by increased number of small and dense LDL particles [77]. Altogether, these features are known as the lipidic triad. In addition, other alterations are

• Increased concentration of very low density lipoproteins (VLDL) due to an increased pro-

• Increased production of apo B-LDL as well as an increment in glycosylation and oxidation

**P25 Median P75**

**deviation**

**Table 1.** Male and female ferritin levels (ng/mL) in type 2 diabetes population belonging to the DIabetes and CArdiovascular

Males 265 150.5 149.3 50.5 107 200.5 Females 330 67.6 83 21.8 41.5 78

The distribution of ferritin in male and femaleT2DM was significantly different (p < 0.0001).

duction or a lower clearance of triglycerides and apolipoprotein (apo) B.

**n Mean Standard** 

RIsk VAllecas (DICARIVA) study P25 and P75, 25th and 75th percentiles.

resistance, and hepatic dysfunction [70].

diabetes risk [71].

106 Recent Trends in Cardiovascular Risks

issue [74–76].

(**Table 1**).

often observed:

of LDL particles.

• Lower clearance of chylomicrons and remnant particles (i.e., intermedium density lipoproteins or IDL).

The diabetic dyslipidemia is associated with insulin resistance, visceral obesity, and liver fat content. Furthermore, insulin resistance is related to an excessive flux of substrates (free fatty acids and glucose) to participate in the formation of VLDL in the liver and with a positive control of mechanisms that produces an excess of large VLDL. These lipoprotein metabolism anomalies are not disconnected facts but are closely related to each other [78]. It is known that lipid metabolism in T2DM is modulated by several factors, such as the degree of glucose control and insulin resistance. The hypertriglyceridemia is very prevalent in T2DM and is also frequent in prediabetes, preceding the presence of chronic hyperglycemia [79]. When there is insulin resistance, mesenteric or "central" adipocytes are full and are unable to retain more fatty acids, consequently fatty acids reach the liver in very high quantities.

Since LDL-size assessment requires special methodology, other approaches have been proposed. The triglycerides/HDL-cholesterol molar ratio has been widely used as a surrogate marker of LDL-size in clinical practice [80]. Earlier a value < 1.33 for this ratio was considered adequate and indicative of large LDL particles. In contrast, individuals with high triglyceride/ LDL-cholesterol molar ratio present a high amount of small, dense, oxidizable, and, thus, highly atherogenic LDL particles [81].

It has been confirmed that triglyceridemia is the determinant of LDL size [82]. In fact, it has been proposed that highly enriched in triglyceride VLDL subtype (VLDL1) are the predecessors of dense and small LDL particles [83].

According to the data in **Table 2**, T2DM women presented higher triglyceridemia and higher HDL-cholesterol levels but lower triglyceride/HDL-cholesterol levels than men.

This study also shows that triglyceride levels increase in parallel to the level of ferritin in men and women. Triglycerides were 36 and 23% higher in men and women, respectively, belonging to the 4th ferritin quartile *versus* their 1st counterparts. LDL particles appear 38% smaller in men and 24% smaller in women at the highest quartile *versus* the lowest, according to the triglyceride/HDL-cholesterol molar ratio. Taking into account these data and the significant correlation between ferritin and this molar ratio (p < 0.001), it can be speculated that body iron contributes to this theoretically higher oxidability and atherogenicity of the LDL.

When the T2DM sample belonging to the 1st or 4th quartile for ferritin was stratified according to the presence of normo or hypertriglyceridemia, low or high levels of HDL-cholesterol, and small or large LDL particles, it was observed that the prevalence of altered triglycerides was higher (odd ratio 1.78; p=0.011) in T2DM patients belonging to the highest quartile for ferritin. Similarly, the odd ratios for high levels of HDL-cholesterol or the presence of small LDLs was 0.54 (p = 0.010) and 1.93 (p = 0.004), respectively in the T2DM patients of the 4th quartile *versus* the 1st quartile of ferritin.

To insist even more in this idea, the prevalence of T2DM presenting the lipid triad was compared to that of patients who did not present any of the three components of the triad. The


concurrence of high ferritin and all the three components of the triad was higher than the concurrence of low ferritin and the three components of the triad. On the contrary, the absence of any of the three components of the triad was less prevalent in T2DM patients with high ferritin values than in their low ferritin counterparts. The odd ratio for the lipid triad/ferritin association was 2.23 (p = 0.010), suggesting the hypothesis that altered CVD risk factors is more prevalent in T2DM patients presenting high iron body stores.

### **6. Conclusions, remarks, and future research**

**95% CI**

**Ferritin** 

**N**

**Mean**

**SD**

**Lower limit**

**Upper limit**

**ANOVA** *p*

**P25** *vs* **P75**

**quartiles**

Men

Triglycerides

mg/dL

P25-<P75

≥P75

HDL-cholesterol

mg/dL

P25-<P75

≥P75

> TG/HDLc\*

mol/mol

P25-<P75

≥P75

> Women

Triglycerides

mg/dL

P25-<P75

≥P75

HDL-cholesterol

mg/dL

P25-<P75

≥P75

> TG/HDLc\*

mol/mol

P25-<P75

≥P75 P, percentile; ANOVA *p*, P value for <P25, P25-<75 and ≥P75.

\*TG/HDLc, Triglycerides/HDL-cholesterol molar ratio.

**Table 2.** (DICARIVA) study stratified according to ferritin quartiles.

84

1.65

1.40 Triglyceride, HDL-cholesterol, and the Triglyceride/HDL-cholesterol molar ratio in men and women from the DIabetes and CArdiovascular RIsk VAllecas

1.35

1.96

163

1.21

1.27

1.01

1.41

<P25

82

1.33

1.20

1.07

1.59

0.038

0.11

84

53.2

13.3

50.3

56.1

163

56.5

14.1

54.3

58.7

<P25

82

59.2

14.9

55.9

62.4

0.025

0.007

84

187.9

213.1

141.7

234.2

163

138.1

101.6

122.4

153.8

<P25

82

152.8

95.4

131.9

173.8

0.027

0.17

66

1.78

1.47

1.42

2.14

133

1.63

1.66

1.35

1.92

<P25

66

1.29

0.82

1.08

1.49

0.013

0.020

66

48.2

12.5

45.4

51.5

133

48.6

12.7

46.4

50.7

<P25

66

48.4

12.5

45.3

51.5

0.99

0.96

66

174.8

118.5

145.7

203.9

133

157.8

114.3

138.2

177.4

<P25

66

128.6

63.1

113.0

144.1

0.038

0.006

108 Recent Trends in Cardiovascular Risks

From all the above, there clearly exists a connection between iron regulation, lipoprotein metabolism, and insulin resistance.

Experimental evidence in animals and humans indicates that dietary fat may be important in iron metabolism. Despite increasing evidence that dietary fat can influence iron absorption and retention, there is a paucity of information about the mechanism implicated [84, 85]. These may be related directly to changes occurring within the intestinal lumen in the enterocytes at luminal or apical membranes. Many aspects of iron absorption and its regulation are still unknown, such as the mechanisms of ferric iron transport, the role of mucines, and so on [8, 9, 11].

Droke et al. [86] demonstrate that palmitate increased iron transport to a greater extent than stearate, and this is followed by far by oleate, which could be due to fatty acid metabolism within the cells and the elongation of palmitic to stearic acid. However, the results suggest that fatty acids affected iron uptake to a greater extent than iron transport. One of the most striking effects of dietary fat on mineral metabolism is the finding of the enhancement of iron uptake and utilization by saturated fat. The effects are prominent when dietary iron is limiting and thus indicate a novel role in promoting an adequate iron status in human [86].

A genome-wide association study (GWA study or GWAS) and epigenome-wide association study (EWAS) together with metabolomic studies would help much to understand the mechanisms involved in the conjoint iron-lipid metabolism that, in turn, affects CVD risks. This information will be useful in the dietary personalization to optimize human health and function.

Meanwhile, as saturated fatty acids (mainly palmitic) increase the iron store [86], total cholesterol, and LDL cholesterol, and induce negative effects on insulin resistance compared to unsaturated fat [87], the authors of the present review claim insisting in the need that T2DM patients show a high adherence to present dietary recommendations for diabetes (American Diabetes Association [88]), which textually include that fat quality (eating monounsaturated and polyunsaturated fats and avoiding *trans* fats and saturated fats) appears to be more important than quantity.

In conclusion, iron is a key metal involved in cardiovascular health. Mild iron deficiency may reduce cardiovascular risk; contrarily, severe anemia induces alterations in the antioxidant iron-dependent enzymes and can be a threat. Iron overload appears to be more important than deficiency in triggering insulin resistance. In this regard, dysmetabolic iron overload syndrome has been related to liver fat accumulation and visceral adiposity [89]. Whether hepcidin resistance is linked to insulin resistance should be a matter of further research.

Our data in an ample sample of adults diagnosed with T2DM suggest that body iron stores, evaluated as serum ferritin, are clearly related with some key markers of the so-called lipidic triad of the T2DM (high triglyceride and low HDL-cholesterol levels together with the presence of small and dense LDL particles) which also is in the frame of the dysmetabolic iron overload syndrome.

### **Acknowledgements**

This study was partially supported by the Spanish project AGL2014-53207-C2-2-R. We also acknowledge type 2 diabetes mellitus patients from de DICARIVA study for their voluntary participation and the Infanta Leonor Hospital of Madrid (Spain).

### **Abbreviations**

