Abstract
Bathing headache is rarely described in literature. We report four middle-aged Taiwanese women who developed severe throbbing headache with maximum intensity of onset during bathing. Diffuse cerebral vasospasm was demonstrated in one of them. All their headaches resolved spontaneously (n = 1) or after nimodipine treatment (n = 3). Except for one patient with vasospasm in whom reversible posterior leukoencephalopathy and an asymptomatic cerebellar infarction developed, the others recovered without any complications. The clinical profile of bathing headache points to idiopathic thunderclap headache. It may not be as benign as previously reported. Nimodipine might be effective in treatment of this special headache syndrome.
Introduction
Negoro et al. (1) first reported ‘benign hot bath–related headache’ in three middle-aged Japanese women in whom soaking in a hot bath or simply pouring hot water over themselves triggered a split-second onset of severe headache. All these patients had normal neurological examinations, biochemical surveys as well as cerebral vasculature demonstrated by neuroimaging studies. Despite the explosive presentation, the headache was considered ‘benign’ because it subsided spontaneously after a period of 2 weeks to 3 months (1). The mechanism of this headache syndrome is still unclear. Excessive stimulation of thermosensitive receptors in the face and scalp has been proposed (1).
In the past year, we treated four female patients with bath-related headache. Although most of them had a favourable outcome, one developed diffuse cerebral vasospasm, reversible posterior leukoencephalopathy and, subsequently, a cerebellar infarction, as demonstrated by serial neuroimaging studies.
Case reports
Patient 1
A 51-year-old-woman, who denied a prior headache history, was admitted to our hospital because of paroxysmal severe headaches, which lasted for 8 days. Her past history included papillary carcinoma of the thyroid, and she underwent total thyroidectomy one year previously and 131I ablation therapy 3 months previously. Just before the onset of the headache, she had an argument with her son. A pulsatile headache then suddenly occurred while she was pouring hot water over herself during a bath (day 1). The headache was intense and located over the vertex. She had nausea but no vomiting, photophobia or phonophobia. The severe headache subsided 6 h later, but a milder headache lingered since then. In the following three nights, an explosive headache occurred each time she took a shower. She had visited our emergency room twice (days 6 and 7). Ketorolac or meperidine injections relieved her pain only temporarily and oral nonsteroidal anti-inflammatory drugs were not effective at all.
Results of neurological examinations on admission (day 8) were normal; neither neck stiffness nor papilloedema was found. She was normotensive throughout the hospital course except for a blood pressure of 172/84 mmHg recorded once in the emergency room. Findings from laboratory surveys, including 24-h urine vanillylmandelic acid (VMA), antinuclear antibody (ANA), anticardiolipin antibody, lupus anticoagulant, and thyroid function tests, were either negative or within normal limits. The cerebrospinal fluid (CSF) study showed slightly increased lymphocytes but no evidence of subarachnoid haemorrhage (no xanthochromia, opening pressure, 110 mm H2O; white blood cell count = 10 × 106/l (all lymphocytes), red blood cell count = 30 × 106/l, protein level = 39 mg/dl, glucose level = 63 mg/dl, immunoglobulin G index = 0.64).
Brain magnetic resonance angiography (MRA) and magnetic resonance imaging (MRI) performed on day 6 revealed multiple short-segmental narrowing of the major cerebral arteries (Fig. 1a), and T2-weighted images showed multiple high-signal-intensity spots in the subcortical regions of temporo-occipital and parasagittal parietal lobes (Fig. 2a). Cerebral angiography performed on day 9 further demonstrated multiple short-segment narrowing involving the distal braches of major cerebral arteries. These findings suggested diffuse vasospasm with posterior leukoencephalopathy.
(a) Three-dimensional (3D) time-of-flight (TOF) MRA obtained on day 6 after the onset of symptoms shows multiple short-segmental narrowing in the bilateral anterior cerebral arteries, the left middle cerebral artery, and the bilateral posterior cerebral arteries (arrows). (b) 3D TOF MRA obtained on days 21 after the onset of symptoms show resolution of vasospasm.
(a) Fluid-attenuated inversion recovery (FLAIR) MRI (repetition time/inversion time/echo time/number of excitations [TR/TI/TE/NEX], 9000/2500/105/1) obtained on day 6 after onset of symptoms shows multifocal high-signal-intensity areas in the right posterior temporal and bilateral occipital regions (arrows). (b) These areas disappeared on the FLAIR MRI(8000/2000/160/1) obtained on day 14 after onset of symptoms. (c) FLAIR MRI(8000/2000/160/1) on day 14 after onset of symptoms shows an acute infarction in the lateral aspect of right cerebellar hemisphere.
Nimodipine 30–60 mg every 4 h was given orally beginning on day 6, but severe headaches recurred as well on days 7 and 8. She then stopped bathing for the next two days. On day 11, the patient resumed washing her inguinal area with warm water; this again triggered a severe headache. Therefore, a continuous intravenous infusion of nimodipine (0.5–1 mg/ h) was given instead, and her headache resolved completely within hours. She tried washing her inguinal region with cold water on day 12 and resumed regular bathing starting day 13; no further headaches were provoked. Brain MRI performed on day 14 showed remission of the posterior white-matter lesions (Fig. 2b), except for a small new ischaemic infarct over right cerebellar hemisphere (Fig. 2c). Clinically, the patient had no cerebellar symptoms or signs. Cerebral angiography performed on day 20 showed marked resolution of the vasospasm, except for narrowing over the M1-2 segment of the right middle cerebral artery (MCA). Serial transcranial colour Doppler flow imaging (TCDFI) revealed an accelerated mean MCA flow velocity during the acute period and a normalized velocity after the disease subsided (Fig. 3).
The summary of the clinical course, headache profile, and neuroimaging findings of patient 1 (MCA: middle cerebral artery).
Nimodipine therapy was tapered off gradually in 20 days, without a recurrence of headache. Follow-up TCDFI performed on day 34 and MRI/MRA performed on day 41 revealed normal results except for the cerebellar infarct. The clinical course and imaging findings of patient 1 are summarized in Fig. 3.
Patient 2
A 53-year-old healthy woman without a headache history experienced four explosive headaches in 2 weeks. All these attacks occurred in the evening at her bathing time. When she poured hot water over her head, severe throbbing pain developed instantaneously in the frontal-vertex region. She had tried to wash her lower body with warm water, but headache occurred as well. Each headache lasted for 7– 8 h and was not relieved with ibuprofen. Associated symptoms included blurred vision, but not nausea, vomiting, photophobia or phonophobia. No precipitating factors other than bathing were found. Her blood pressure was 141/81 mmHg. Physical and neurological examinations were unrevealing. Results of biochemical surveys, including tests for ANA, the erythrocyte sedimentation rate and urine VMA were normal. Brain MRI/MRA performed on day 9 after headache onset showed normal results. Oral nimodipine 60 mg every 4 h was given on day 14, and no further attacks were noted, even with a hot-bath challenge on day 17. Nimodipine treatment was tapered off gradually in the following 4 weeks, and no headache recurred.
Patient 3
A 32-year-old woman was admitted because of daily headaches over five days. The first headache (day 1) started after a vigorous cough. In the following 4 days, three headaches occurred during bathing as she poured hot water over her body or head, and one occurred while she brushed her teeth with cold water. All of these headaches were pulsatile, they were located in neck-to-vertex region, and they lasted for 2 h. The patient had no nausea, vomiting, photophobia, or phonophobia. The headache was so intense that she visited the emergency room for each headache, and the pain was only partially relieved with an intramuscular injection of ketorolac. The patient had no significant past history except that she had delivered a baby 3 months previously. Her blood pressure was 120/67 mmHg. Findings from neurological examinations and routine surveys, including urine VMA test, were normal. Brain MRI/MRA on day 7 showed normal results. One severe headache occurred during a nightmare, which occurred after her hospitalization. Nimodipine 30 mg every 4 h was prescribed on day 7. No more headaches were reported after that, and the medication was tapered in 12 days. The patient's headaches did not recur. Results from follow-up brain MRI/MRA performed 1 month later were normal.
Patient 4
This 53-year-old woman experienced three episodes of severe left hemicrania within 2 weeks. The first headache occurred while she was taking a shower. An explosive throbbing headache occurred as the patient poured hot water over her body. The headache persisted for 5–6 h and eventually resolved after she slept. Another headache happened during bathing 2 days later. To avoid further headaches, she stopped bathing for 10 days, but headache recurred when she resumed bathing. This time, the headache lasted for 30 h. A systolic blood pressure of as high as 198 mmHg was recorded once; her baseline blood pressure was 150–160/90–100 mmHg during this period. Findings from neurological examination and surveys, including serum catecholamine, thyroid function, and urine VMA tests, were normal. Brain MRI/MRA on day 10 after the headache onset showed normal results. The patient refused any medications, and her headache remitted spontaneously.
Discussion
We report four patients with bathing headache. Their headache profiles are similar to that reported by Negoro et al. (1). The patients affected were mostly middle-aged women in their 5th or 6th decade. The headaches were pulsatile and intense, with maximal intensity at their onset. Bathing repeatedly provoked the headaches in the acute stage. However, the headaches eventually subsided after a cluster of attacks occurred, even without treatment. Multifocal segmental vasoconstriction, i.e. arterial vasospasm, occurred in one of our four patients and none of the three patients reported by Negoro et al. (1). It might indicate a poor outcome.
It is noteworthy that the features of bathing headache fulfil the criteria of idiopathic thunderclap headache (2, 3), which include the following (1): a very severe pain intensity (2), a hyperacute onset of pain (<30 s) (3), a headache duration of 1 h to 10 days (4), headache recurrence over a 1-month period without regular occurrence over the subsequent months, and (5) normal cerebral vasculature or reversible segmental vasoconstriction. These striking similarities suggest that bathing headache is a variant of idiopathic thunderclap headache, with bathing as a specific precipitating factor.
As demonstrated by serial MRA, cerebral angiography, and TCDFI (4, 5), the dynamic changes in the major cerebral arteries and their braches in patient 1 could be explained only by vasospasm and not vasculitis or atherosclerotic stenosis. Reversible posterior leukoencephalopathy (6) and a silent cerebellar infarction eventually developed. Therefore, like idiopathic thunderclap headache (2, 3, 7–11), some patients with bathing headache may also develop multifocal vasospasm or ischemic complications, although most have a benign course. Notably, in patient 1, the relationship between bathing headache and vasospasm was not completely parallel (Fig. 3). The bathing headache apparently resolved before the complete resolution of vasospasm. Bathing headache, thunderclap headache, and reversible posterior leukoencephalopathy (6) may present with acute headache and reversible vasospasm; but marked hypertension and immunosuppressive therapy are hardly encountered in the former two syndromes. The overlapping between these three syndromes suggests at least part of the pathophysiological mechanism is similar.
Thunderclap headache simply labels a syndrome, sometimes associated with vasospasm or a structural lesion but more often of unknown mechanism. The underlying mechanism of bathing as a trigger for thunderclap headache remains unknown (1). Either soaking in water, pouring water over the body, or washing different parts of the body can provoke the headache. The water temperature is probably not important, because cold, warm, or hot water could trigger bathing headache. To avoid over-emphasizing of thermal stimulation, we prefer the term ‘bathing headache’ to ‘benign hot bath–related headache’. Immediate response to bathing indicates skin stimulation is more important than temperature change. Although bathing remains the most consistent headache trigger, other stimuli can provoke this headache syndrome as well. For example, cold wind, coughing, tooth brushing, and nightmares were among the headache triggers in the previous (1) and present reports.
No effective medical treatment had been proposed for bathing headache. On the basis of the previous (1) and present reports, some parenteral medications might relieve the headache temporarily. Negoro et al. have suggested that avoiding bathing might be the best management for this headache. We empirically used nimodipine to treat bathing headache, even though not all our patients had vasospasm. The oral form of nimodipine was sufficient for patients 2 and 3, whereas intravenous nimodipine was needed to completely abort the bathing headache in patient 1. The effect of nimodipine in treating cerebral vasospasm that occurs after aneurysmal subarachnoid haemorrhage (12, 13) is well documented. However, the underlying mechanism of nimodipine in treatment of bathing headache is still unclear since not all the responsive patients showed vasospasm. Although experience is still limited, nimodipine may be a potential therapeutic choice for bathing headache.