The relationship between headache and hypertension has been debated for many years, and is a subject of great concern to our patients. Today, in many areas of medical science, the focus is almost exclusively on the most recent publications, and any paper more than 20 years old is often ignored. However, in the case of hypertension and headache, the older papers still have much to contribute because the prevalence of severe, untreated hypertension was much higher then, and the awareness of hypertension much lower. The central problem in this debate, of course, is that both headache and hypertension are very common, and hence many hypertensives complain of headache. In addition, people who experience headache are more likely to seek medical care, and thus to have their blood pressure checked.
To answer the question of whether headache is more common when the blood pressure is higher, we must look at population surveys. A good example is the U.S. Health Examination Survey of Adults (subsequently known as NHANES), which was first conducted in 1960-1962, when effective antihypertensive treatment was just beginning to make its mark. The main finding was that headaches that occurred "every few days" or bothered the subjects "quite a bit" were reported by 22.8% of people with a systolic pressure <140 mm Hg, 23.2% with systolic pressure of 140-159 mm Hg, and 24.4% with systolic pressure >160 mm Hg, a trend that did not approach statistical significance. The subjects also underwent fundoscopic examination, and when they were classified according to whether or not they had hypertensive retinopathy (grade II or higher), there was a difference, but it was surprisingly small. In people with a diastolic pressure >90 mm Hg, headache was reported in 20.6% of those with a normal fundal exam, and 22.4% of those with retinopathy.
Another early study is equally revealing. It was published in 1953, and represents a type of study that is rarely seen these days -- a series of personal observations by a single physician of patients in his clinic. Stewart reported on 200 consecutive patients, all of whom had a diastolic pressure of >=120 mm Hg, and whom he asked to describe their symptoms; 44% admitted to headaches, but the blood pressure level did not distinguish those with and without headaches. As in the NHANES study, the presence of retinopathy was associated with a slightly higher prevalence of headache. In contrast to the NHANES results, the factor that reliably distinguished the presence of headache was whether or not the patients were aware of their hypertension. Of the 96 patients who knew their diagnosis, 74% complained of headaches, whereas in the 104 who were unaware of their hypertension, only 16% reported having headaches, despite an average diastolic pressure of 135 mm Hg. This trend was seen even in patients with malignant hypertension (papilledema on the fundal exam), of whom there were 18. Headache was reported by 10 of the 12 patients who had been diagnosed, but by only one of the six who were unaware of their diagnosis. Stewart concluded that, while a true "hypertensive headache" may exist in patients with hypertensive retinopathy and malignant hypertension, it is quite rare, and that in the vast majority of patients it may be attributed to anxiety.
In his classic 1968 textbook High Blood Pressure, my father supported the view that the majority of headaches in hypertensives were not simply the consequence of the height of the blood pressure, and that psychoneurosis played a major role. He wrote: "The high incidence of psychoneurotic symptoms in patients with hypertension may have three explanations: it may be because psychical disturbances play a part in pathogenesis; it may be that the discovery of high blood pressure induces a psychoneurosis; and it may be that patients who are discovered to have hypertension present with psychoneurotic symptoms, simply because these are much the most common symptoms which bring patients to doctors. While there may be some truth in the first, the second and third considerations are certainly true both in Great Britain and the U.S. today."
Some more recent studies paint a different picture, at least at first glance. A survey of 2673 patients enrolled in seven double blind, placebo-controlled studies of the effectiveness of irbesartan, an angiotensin receptor blocker, found that in the placebo group, there was a weak correlation between the prevalence of headache and the diastolic pressure, but no correlation with the systolic pressure. Also, active drug treatment was associated with a significantly lower incidence of headache (17% of treated patients vs. 22% of those receiving placebo). The authors concluded that headache is a feature of mild hypertension and that the aggressive reduction of blood pressure can reduce this symptom.
Many patients would no doubt agree that they are more likely to have headaches when their blood pressure is up, but other recent studies cast doubt on this. One examined the association between headache and blood pressure during ambulatory blood pressure monitoring. In a population of 150 patients referred for evaluation of hypertension (many of whom were taking antihypertensive medications), episodes of headache were reported in the diaries of 30% of the patients. The study showed no relationship between the occurrence of headache and what was happening to the blood pressure, and most of the patients were free of headache when the blood pressure was at its highest. In a second study, serial observations were made of 54 patients who were participating in a double blind, placebo-controlled study involving propranolol and clonidine. At each clinic visit, patients were told the blood pressure at the previous visit and then asked to estimate the value at the present visit. They were also asked about symptoms (including headache) and general well-being. There was a significant association between the patients' estimates of their blood pressure and the actual blood pressure recorded during the clinic visit, although the correlation coefficient was quite weak (0.17). However, this estimation had nothing to do with symptoms, and negative mood correlated only with predicted pressure, not actual pressure. Patients also rated the active treatments more effective than placebo, and there was a relationship between perceptions of greater well-being, fewer symptoms, and higher ratings of medication effectiveness.
The ability of patients to predict their own blood pressure was examined more directly in a study by Cantillon et al, conducted in a general practice in London. Exactly one half of 102 hypertensives claimed that they could perceive when their blood pressure was high. When this was put to the test, only a handful could do better than chance. What did distinguish the "predictors" from the "nonpredictors" was the frequency of symptoms in general, which was much higher in the predictors. Headache, the commonest symptom, was reported by 37% of predictors and 15% of nonpredictors, and the predictors were also more anxious. When these studies are taken together, it is fair to say that there is little support for the idea that in patients with mild or moderate hypertension short term changes in blood pressure show any correlation with changes in such symptoms as headache.
On the other hand, it has been suggested for many years that there is also a specific type of headache that, in the words of my father, "is almost pathognomonic of high blood pressure, which tends to be associated with gross hypertension, and which has certain features suggesting an organic origin." He described this type of headache as occurring early in the morning on waking, and observed that it may also develop following daytime sleep, but he was at a loss to explain its mechanism. Moser et al, in 1962, described 54 patients who were referred to a headache clinic, and in whom hypertension was discovered incidentally. Thus, in most of these patients the headaches preceded the diagnosis of hypertension. There was no specific location of the headaches, but 60% were described as occurring early in the morning. In 25% of patients the headaches were substantially improved as a result of lowering the blood pressure. The importance of early morning or waking headache was demonstrated in another study comparing symptoms in untreated hypertensives, treated hypertensives, and normotensives. It was found that waking headache, blurred vision, and depression were all more common in untreated hypertensives. On re-evaluation after 10 months of treatment, there was a significant improvement in headaches, and those patients who lost their headaches altogether (10% of the total) showed a larger reduction of blood pressure than the others. The authors considered the possibility that the relief of headache might be a nonspecific effect of the patients' general improvement in well-being as a result of treatment, and rejected it on the grounds that some other symptoms, such as depression, did not improve. Although it was not appreciated at the time, morning headaches may be the result of sleep-disordered breathing, with which hypertension is associated. A recent study found a significant correlation between the severity of obstructive sleep apnea and the frequency of early morning headaches, and also revealed that successful treatment of the sleep apnea improved the morning headaches in the majority of cases, while headaches occurring at other times were unaffected. Perhaps the mystery of morning headaches in the hypertensive patient has now been solved.
Can the "old" and the "new" views of headache and hypertension be reconciled? What clinical studies tell us is that headache and hypertension are both common in the general population, and inevitably coexist without any necessary connection. Nevertheless, some, but not all, studies have shown that headache is somewhat more common in hypertensives than in normotensives. This slight excess of headaches in hypertensives may have at least two explanations. In the majority of patients the headaches are nonspecific, and may be associated with a spectrum of symptoms, including anxiety, which may be either a cause or a consequence of the hypertension. In these patients, the presence of headache is rarely, if ever, a direct consequence of the height of the blood pressure. Since the majority of patients are aware of their diagnosis, it is likely that the headaches are in many cases the consequence of labeling the patients as hypertensive, or that anxiety and tension are contributory factors both to the hypertension and to the symptoms. In these patients, treating the hypertension may relieve the headache as well as other symptoms, leading to an improved sense of well-being, but there is no reason to think that lowering the blood pressure is directly responsible for the improvement of symptoms. That leaves a small residue of patients in whom there is a direct connection between the height of the blood pressure and the occurrence of headaches, who may be of two types. One is the very rare patient with very high diastolic pressure (>=120 mm Hg), and the other is the patient with morning headaches, who may have sleep apnea as the underlying problem.
Address for correspondence/reprint requests: Thomas Pickering, MD, DPhil, Mt Sinai School of Medicine, 50 East 98th Street, New York, NY 10029.
© 2000 Le Jacq Communications, Inc.
Cite this: Headache and Hypertension -- Something Old, Something New - Medscape - Sep 01, 2000.