An association was found between headache or non-migrainous headache and later dementia. Not so for migraine.
Any headache and non-migrainous headache are more likely to be reported at baseline among patients who later are included in a dementia registry, the HUNT study found. No statistical significant association was found between migraine and dementia. In fact, migraine was less likely in patients in whom dementia developed.
To examine the association between headache and dementia, researchers in Norway used data from the Nord-TrÃ¸ndelag Health Surveys performed in 1995–1997 (HUNT2) and 2006–2008 (HUNT3). Patients aged 55 years or older who answered headache questions in HUNT2 and later participated in HUNT3 were the reference group.
Any headache was more likely to be reported in HUNT2 among those who later were included in the dementia registry compared with the reference group but less likely among confirmed non-demented patients. The relationship was stronger for non-migrainous headache; such an association was not found for migraine.
The researchers suggested there are several ways to interpret the finding that any headaches, particularly non-migrainous headaches, are more likely to be reported at baseline among patients in whom dementia develops:
• The notion that for some patients, the early or presymptomatic condition of dementia was causing headache-rather than headache later causing dementia-cannot be ruled out.
• Headache and dementia may share an underlying causal factor, eg, cardiovascular risk factors or alcohol overuse.
• Headache may be a true risk factor for vascular dementia. If it is, the underlying mechanisms should be elucidated.
Although a large number of comorbid conditions (eg, psychiatric disorders) have been identified in middle-aged and older patients with headache, relatively few studies have evaluated the relationship between headache and cognitive status and dementia, and most previous studies have focused on migraine, the researchers noted. They recommended that adding headache questions in the clinical evaluation of cognitive function be considered.