Introduction

Iron is the most commonly ingested nutrient in the human diet [1]. Iron deficiency affects many cellular processes, which include oxygen transport, myelination, storage, electron transport, oxidative phosphorylation, neurotransmitter metabolism, immune functions, and DNA synthesis [2, 3].

Previous studies showed that anxiety and/or depression, social and attention deficit problems were more frequent in children with iron-deficiency anemia (IDA) [4]. Similarly, an association between serum ferritin (SF) levels and depression in the absence of IDA has been reported in adults [5].

In spite of the fact that headache is very prevalent in the general population, there is no too much data about the frequency of headache and associated factors in literature. It was stated that headache was more common in hemachromatosis in which there is iron overload [6]. The most frequent type of chronic episodic headache in general practice is migraine headache [7]. It is known that the presence of headache and migraine are associated with anxiety and depression. Although both migraine and IDA are relatively more frequent in young women, there has been no study about an association between the two.

In our previous study [8], we found an increased prevalence of chronic widespread pain and fibromyalgia in our IDA patients, and we observed that these patients had very prominent sleep disturbances.

Until now, the frequency of headache and migraine in IDA and the influence of the presence of IDA on migraine have not been evaluated. In our study, we tried to determined factors associated with migraine and anxiety/depression in IDA.

Materials and methods

A total of 127 patients with IDA who came to our Internal Medicine and Hematology outpatient clinics were included into the study. As already known, IDA is most prevalent in adult-age women and content analysis scores from female subjects might allow a more correct prediction of psychological state than scores from male subjects. Therefore, only female subjects were included into our study [9]. Patients with an already diagnosed cause of anemia (including thalassemia minor) and patients with hematological malignancies, and those who have had psychiatric treatment within the last 6 months were not included into the IDA group. Patients with a prediagnosed neurological disease and solid-organ malignancy were also excluded. All the patients were residents of Edirne, a city in Northwestern Turkey, and all were of Turkish descent. The aim of the study was described in detail to all patients and written informed consent for participation was obtained.

Whole blood count and standard parameters of iron status were determined in a fasting venous blood sample obtained from IDA patients. The CellDyn 3700 analyses was used for determining whole blood counts, and hemoglobin (Hb), mean corpuscular volume (MCV), red cell distribution width, and hematocrit and platelet values were recorded down. The ARCHITECT c8000 system (Abbott Laboratories, Diagnostics Division, Illinois, USA) was used to determine serum iron (SI) (normal range: 25–156 μg/dl) and unsaturated iron-binding capacity (UIBC) (normal range: 110–370 μg/dl) with spectrophotometric methods. Total iron-binding capacity (TIBC) was calculated by summing up SI and UIBC. The equation [(SI/TIBC) × 100 %] was used to calculate transferin saturation (TS). The IMMULITE 1000 analyzer (Diagnostic Products Corporation, DPC, Los Angeles, USA) was used to determine SF. For female individuals, the normal ranges for ferritin concentration were 6–159 ng/ml.

The definition of IDA was a Hb concentration of less than 12 g/dl in women (normal female adult 12.0–15.5 g/dl) together with a SF less than 12 ng/ml, low SI, raised TIBC, and TS < 16 %. In case the MCV was < 76 fl, High performance liquid chromatography (HPLC) was carried out and subjects with a HbA2 > 3.5 % were diagnosed as Thalassemia minor (TM).

The demographic and clinical characteristics of the patients, including age, marital status, educational level, smoking status, concomitant diseases, and employment status, were interviewed. Initially, the subjects were asked the following 2 questions: “Have you ever suffered from headache in your life?” and “Have you ever had headache during your life”. Later, subjects who answered “yes” to the second question were questioned about the frequency, intensity, location, aura symptoms, other migraine or cluster headache features, triggering factors, and associated symptoms.

International Headache Society criteria [10] were used to diagnose migraine and evaluated data were interpreted by a neurologist (YÇ) and a psychiatrist (MŞT).

The Hospital Anxiety and Depression Scale (HADS) was administered by specially trained doctors (MŞU, HK, RA) [11]. The HADS aims to rate anxiety and depression in patients with physical illness, and it has two subscales: one assessing anxiety and the other, depression. The reliability and validity of the Turkish version of HADS have been established in Turkish patients: the optimum cut-off point in the anxiety subscale score was reported to be 10 and the depression subscale score was 7 [9]. In addition, all subjects were questioned for the presence of headache or migraine-related quality of life (QoL) disturbance by using a Visual analog scale (VAS) between 0 and 4. Four pointed out to the highest degree of QoL disturbance and 0 meant no QoL disturbance.

The categoric variables were compared by the chi-square test. The quantitative variables were compared by the unpaired t-test. Pearson’s test was used to make correlation analysis.

Results

The questioning of IDA patients revealed that 101 of them (79.5 %) have defined headache at any time of their lives. When patients with headache and migraine were sorted into groups based on age, it was seen that they were most common in the younger than 30 years of age group (Table 1).

Table 1 The frequency of headache and migraine in patients with iron-deficiency anemia according to age groups
Full size table

In all, 24 (52.2 %) of IDA patients with migraine had aura symptoms and in 40 (87 %), headache became worse with movement. The factors which triggered migraine attacks were emotional stress in 20 cases (44.4 %), sleep disorder in 14 (31.1 %), and the menstrual cycle in 4 (8.9 %). The symptoms most frequently accompanying migraine attacks were nausea or vomiting in 23 cases (51.1 %), photophobia in 7 (15.6 %), phonophobia in 6 (13.3 %), and odor sensitivity in 5 cases (11.1 %).

The clinical features, hemogram and iron parameters, of IDA patients with and without migraine is showed in Table 2. In the IDA with migraine group, smoking was significantly more frequent (p < 0.001) and TIBC (p = 0.008) was significantly higher than in the IDA without migraine group. On the other hand, Hb level (p = 0.013) and MCV (p = 0.02) were significantly lower in the IDA group with migraine. SI and ferritin levels were similar in both groups (all p values > 0.05).

Table 2 The comparison of clinical features and iron parameters in iron-deficiency anemia (IDA) patients with and without migraine
Full size table

The comparison of anxiety and QoL scores of IDA patients with and without migraine is showed in Table 3. In the IDA group with migraine, the frequency of anxiety (p = 0.024), the mean anxiety score (p = 0.046), and headache-related QoL disturbance score (p = 0.021) were significantly higher than in the IDA group without migraine.

Table 3 The comparison of anxiety and depression scores in iron-deficiency anemia (IDA) patients with and without migraine
Full size table

When all IDA patients were considered as a whole, it was observed that HADS-A score correlated with HADS-D score (r = 0.66, p < 0.001). HADS-D score was also seen to correlate with headache-related QoL disturbance score (r = 0.37, p < 0.001). These parameters did not have any significant correlations with anemia or iron parameters.

The HADS-A scores of IDA patients with migraine correlated with HADS-D scores (r = 0.69, p < 0.001), and also with Hb (r = 0.48, p = 0.041), MCV (r = 0.51, p = 0.032), and headache-related QoL disturbance score (r = 0.67, p < 0.001). HADS-D score, on the other hand, correlated with age (r = 0.35, p = 0.022), Hgb (r = 0.5, p = 0.029), and headache-related QoL disturbance score (r = 0.51, p < 0.001). In this latter group, migraine-related QoL disturbance also correlated with MCV (r = 0.52, p = 0.027). When migraine patients with aura were compared with others, it was observed that they had lower Hb values (10.1 ± 0.9 vs. 11.3 ± 1.1, p = 0.02), higher HADS-D scores (10.1 ± 5.3 vs. 6.4 ± 3.9, p = 0.005), and higher migraine-related QoL disturbance scores. Smoker migraine patients had lower Hgb values than others (10.5 ± 1.2 vs. 11.6 ± 0.9, p = 0.03).

Discussion

In our study, the lifetime frequencies for headache and migraine were found to be quite common in IDA patients (79.5 and 36.2 %). We did not had a control group. Nevertheless, a previous epidemiologic study conducted in our city in the same age group reported the prevalence of migraine as 19.9 % [7]. We have included only female subjects into our study. In the aforementioned study, the prevalence of migraine in female subjects (29.3 %) was lower than in our IDA patients. One of the two studies from our country reported the prevalence of lifetime headache to be 65 % and that of migraine to be 16.4 % [7]; while in the second study, the prevalence of lifetime migraine was 13 % [12]. The frequency of migraine in our IDA patients was higher than the prevalences reported from Western countries [13, 14].

It is known that geographical, sociocultural, and genetic factors have influence in the frequency of migraine. It was reported that the frequency of migraine was highest among young female subjects. The highest frequency of migraine in our study was (48.3 %) detected in subjects younger than 30 years of age in the IDA group. In general, the frequency of migraine decreases with age; however, in our study, migraine was most frequent in subjects older than 50 years of age (41.7 %).

One study reported that there was an association between Hb level and cognitive function in IDA [15]. Another study showed that iron supplementation significantly improved verbal learning and memory in nonanemic iron-deficient adolescent girls [16]. In spite of the fact that there is no strong evidence to support an association between fatigue, inability to concentrate, other nonspecific symptoms, and iron deficiency, people often attribute their symptoms to IDA [17]. Therefore, nonspecific symptoms in IDA patients are generally explained with the presence of anemia. When we interpret the results of this study, we might conclude that the higher frequency of migraine in IDA might explain a great proportion of nonspecific symptoms in that disease.

In one study, it was reported that migraine was more frequent in subjects with restless leg syndrome [18]. Nevertheless, it was reported that the pathophysiologic basis of increased frequency of migraine was iron deficiency, which was also associated with restless leg syndrome [19]. In another study, it was stated that IDA was more common in patients with menstrual migraine, and that migraine attacks could be triggered by IDA [20].

In our IDA group, smoking, hemoglobin, and low MCV were found to be associated with migraine on univariate analysis. This was an interesting result of our study. As we have mentioned earlier, association between IDA and depression, anxiety, and attention disorders were reported; nevertheless, an increased frequency of migraine in IDA, especially lower Hb levels in IDA patients with migraine, has been reported for the first time. There are data supporting that iron deficiency affects neuronal functions; however, it is currently not possible to make an explanation about that.

In addition, we observed associations of the presence of anxiety and its score with the presence of migraine in IDA patients. IDA patients with migraine were seen to have significantly higher QoL disturbance. As a result, we might suggest that the presence of frequent migraine in IDA might be partially explained with the presence of anxiety. The coexistence of anxiety and migraine in IDA might lead to worsening of the QoL disturbance score in this subgroup of patients. One possible comment about our result might be as follows: In our study, we included patients admitted to the Hematology and Internal Medicine Outpatient clinics. Our patient population was relatively young women seeking medical attention at our hospital; and probably, it was a group with high anxiety scores. If we had had the chance to evaluate IDA patients not coming to the outpatient clinics, probably, we would have found lower anxiety and depression scores.

In our study, the symptoms severities of anxiety and depression were found to be significantly correlated with each other. However, we did not detect significant correlations between hemoglobin, iron parameters, and the psychiatric tests. Interestingly, migraine patients with aura had lower Hb values. The relationship between the presence of aura and the depth of anemia should be analyzed in detail. The limitations our study could be summarized as follows. First, we did not have a positive control group.

Second, as the study had a cross-sectional design, we did not follow-up the patients to see the disappearance of migraine or its symptoms after iron replacement therapy.

The nonspecific symptoms in IDA patients are generally attributed to anemia. Considering the results of our study, we might suggest that headache and migraine contribute to the symptoms in IDA to a great extent. As a conclusion, we observed a significantly increased frequency of migraine and headache and associated symptoms of anxiety in IDA patients. The presence of nonspecific complaints in IDA patients might not always be explained with the presence of only anemia. If the patient has unexplained symptoms, physicians should search for the presence of anxiety and depression. In addition, migraine should be in the differential diagnosis of headache in IDA patients coming to the outpatient clinics, and they should be questioned for its presence.

Conflict of interest

The authors declare that there are no actual or potential conflicts of interest in relation to this article.