Abstract
Objectives: Morning headache has been considered as an accompanying symptom of obstructive sleep apnoea syndrome (OSAS). However the frequency, characteristics, predictors and impacts of morning headache in habitual snorers are not well defined.
Methods: We studied consecutive patients with habitual snoring in a sleep laboratory using polysomnography. All patients were interviewed by a physician regarding the presence or absence of morning headache, migraine and insomnia. Each patient completed the Short Form-36 health survey (SF-36) and the Hospital Anxiety and Depression Scale (HADS). Morning headache was defined as headache on awakening ≥1 day/week for ≥6 months.
Results: Of the 268 participants with habitual snoring, 63 (23.5%) had morning headache and 184 (69%) had OSAS. Patients with morning headache reported lower scores in all eight domains of the SF-36 than those without (difference: 10.6 to 29.7 points, all p ≤ 0.005). The independent predictors of morning headache were migraine (adjusted odds ratio (AOR) 6.3), insomnia (AOR 4.2), psychological distress (HADS ≥ 8) (AOR 3.9) and OSAS (AOR 2.6). Morning headache in 12 patients (19%) fulfilled the criteria for migraine attacks.
Conclusions: Morning headache was common in habitual snorers and associated with a pervasive impairment of health-related quality of life. Migrainous features were not uncommon. Not only OSAS, but migraine, insomnia and psychological distress were also important predictors for morning headache, even in snoring patients.
Introduction
Diagnostic criteria of sleep apnoea headache proposed by the International Classification of Headache Disorders (ICHD-2, code 10.1.3) and the frequencies of each criterion in our snoring patients with OSAS and morning headache
Habitual snoring is a common health problem (12) and also a sensitive symptom of OSAS (13), but the predictors for morning headache in patients with snoring are not known. In addition, the impact of morning headache on the health-related quality of life (HRQoL) in habitual snorers is not well defined.
This study sought to investigate the frequency of morning headache, its impact on HRQoL and relevant predictors in habitually snoring patients. All these patients underwent a study of overnight polysomnography (PSG).
Subjects and methods
We prospectively enrolled consecutive patients, males and females, ≥18 years old, who had been referred to our sleep laboratory between 1 January 2009 and 31 March 2010 with the complaint of habitual snoring, i.e. daily or near-daily snoring. Patients were excluded if they (i) had undergone an operation or continuous positive airway pressure (CPAP) treatment for snoring or OSAS; (ii) had persistent headache without remission, because this headache pattern was difficult to differentiate from morning headache; or (iii) had medication overuse headache based on the ICHD-2 criteria (14). This study had no limits regarding body mass index (BMI) levels, and infrequent painkiller usage was allowed. One physician (Chen PK) interviewed all patients to determine whether they had morning headache, migraine or insomnia. Diagnoses of migraine and other headache disorders were based on ICHD-2 (4). Diagnosis of insomnia was based on the International Classification of Sleep Disorders (ICSD-2) (15). Because there were no globally accepted diagnostic criteria for ‘morning headache’, we defined it in this study as headache on awakening ≥1 day/week for ≥6 months.
All patients filled out a structured questionnaire designed for morning headache. Information included demographics, BMI, headache frequency, duration, location, quality, intensity, and accompanying symptoms such as nausea, vomiting, photophobia and phonophobia. The study also investigated whether the patient’s morning headache fulfilled the criteria for migraine attacks based on ICHD-2 (4), i.e. duration ≥ 4 hours with two or more of the headache symptoms, including pulsatile quality, unilateral location, moderate or severe intensity and exacerbation due to physical activities; and one or more accompanying symptoms, including either ‘nausea or vomiting’ or ‘photophobia and phonophobia’. Each patient also completed the Hospital Anxiety and Depression Scale (HADS) (16), and the Short Form-36 health survey (SF-36) (17). All patients underwent an overnight PSG study after the questionnaire survey. This study protocol was approved by the Institutional Review Board at the Lin-Shin Hospital, Taichung, Taiwan. Patients signed an informed consent form before entering the study.
Study assessment
Hospital Anxiety and Depression Scale (HADS)
The HADS is a self-administered instrument designed to detect states of depression and anxiety in the setting of a hospital outpatient clinic (16,18). It does not consider the somatic symptoms of anxiety and depression, and thereby excludes the influence of the confounding factors of physical symptoms and signs. The questionnaire includes 14 questions: seven for depression and seven for anxiety. Each question is rated with a score of 0 to 3, depending on the severity of the problem, with a total score ranging from 0 to 42. Psychological distress was defined as a total score ≥8 in this study (19).
Short Form-36 health survey (SF-36)
The SF-36 is a self-administered 36-item scale widely used for the measure of generic HRQoL (17,20). It evaluates eight domains of health, including physical functioning (PF), role limitations due to physical problems (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), role limitations due to emotional problems (RE), and mental health (MH) during the preceding 4 weeks. All items were scored on a scale from 0 to 100. Higher scores indicated better health. A difference of >5 points in each domain was considered clinically significant (17,20).
Polysomnography (PSG)
The PSG was performed on Embla S4500 (Flaga, Reykjavik, Iceland) and Alice 4 (Healthdyne, Atlanta, GA, USA) sleep systems. The EEG (C3/A2-C4/A1, and O1/A2 according to the 10–20 international electrode placement system), the two-channel electro-oculogram, chin electromyogram and electrocardiogram were recorded with surface electrodes. Airflow was detected by two channels through a thermal sensor (one channel) and nasal pressure transducer (one channel), and respiratory movements with thoracic and abdominal inductive plethysmography. Oxygen saturation during sleep was measured continuously using pulse oximetry with a finger probe. Snoring was evaluated with a neck microphone. Leg movements were recorded by bilateral tibial electromyograms. Apnoea, hypopnea and periodic leg movements in sleep (PLMS) were defined based on The AASM Manual for the Scoring of Sleep and Associated Events (21). The apnoea is a period of breathing cessation. The hypopnea is defined as a ≥ 30% reduction in breathing with a ≥ 4% oxygen desaturation or arousal. The minimum duration of an event was 10 seconds. The apnoea–hypopnea index (AHI) was calculated as the number of apnoeas and hypopneas per hour of total sleep time. PLMS was scored if there were at least four consecutive leg movements of 0.5–10 seconds’ duration, and between 5 and 90 seconds apart. The minimum amplitude of a leg movement event was an 8 µV increase in EMG. The PLMS index was calculated as the number of PLMS per hour. OSAS was defined as 5 ≤ AHI < 15 if patients had at least one of the following complaints: loud snoring, daytime sleepiness, fatigue, and breathing interruptions during sleep or AHI ≥ 15 regardless of any additional complaints (15). PLMD was defined as a periodic leg movement index of >15 per hour of sleep with disturbed sleep or daytime fatigue (15).
Statistical analyses
Descriptive data were presented as mean ± standard deviation (SD) or percentages. For categorical data, a chi-square test or Fisher’s exact test was used to test the difference between groups. The Student’s t-test was used to compare the means of continuous variables. The potential predicting variables for the presence of morning headache included age, gender, smoking habits, migraine, HADS score (≥ 8 or <8), insomnia, OSAS and PLMD. The predictors for morning headache were presented as odds ratios (ORs) with 95% confidence intervals (CIs). Logistic regression was used to calculate adjusted ORs (AORs) after controlling for potential confounders. An estimated difference of each domain of the SF-36 due to morning headache was calculated after controlling for potential confounders by general linear models. All calculated p-values were two-tailed and statistical significance was defined as a p-value of <0.05.
Results
During the study period, 284 patients with habitual snoring who were referred to our sleep laboratory were recruited. After excluding two patients because of refusal, seven with persistent headache and five with medication overuse headache, 270 patients signed informed consent forms to enter the study. Two patients were excluded due to incomplete data. A total of 268 patients (178M/90F, mean age 44.4 ± 12.2, range 18–76 years) completed the study. Of them, 184 (68.7%) (135M/49F) were diagnosed with OSAS and 33 (12.3%) (21M/12F) were diagnosed with PLMD, according to both the PSG results and relevant clinical symptoms. Based on the physician interview, 116 patients (43.3%) (73M/43F) had insomnia and 58 (21.6%) (35M/23F) had migraine, including eight with and without aura and 50 without aura. Based on the HADS, 193 (72.0%) (123M/70F) had psychological distress (HADS ≥ 8 points).
Headache profile of morning headache
Clinical profiles of morning headaches (n = 63) in habitual snorers and comparisons in relation to obstructive sleep apnoea syndrome (OSAS) and migraine
Sleep apnoea headache
Habitual snorers with OSAS were more likely to have morning headache than those without (27.2% (50/184) vs. 15.5% (13/84), p < 0.001). The frequencies of the three proposed headache symptoms (A1 to A3) of ICHD-2 sleep apnoea syndrome in our morning headache patients are shown in Table 1. Forty percent of patients with morning headache could fulfil the criteria for sleep apnoea headache symptoms (A2), 32% experienced a short duration of <30 minutes (A3) and 18% had morning headache more than 15 days per month (A1). Overall, 62% (n = 31) of OSAS patients fulfilled criterion A, i.e. at least one of three proposed headache symptoms. In this study, we did not test criterion D, i.e. disappearance of morning headache after treatment of OSAS.
The impact of morning headache on the SF-36
Patients with morning headache had significantly lower scores in all eight domains of the SF-36 (Figure 1). The general linear model showed that the presence of morning headache independently accounted for more than 5 points of estimated difference in the domains of RP (−19.1 points, p = 0.006), BP (−11.5 points, p = 0.002), SF (−8.4 points, p = 0.02) and RE (−22.5 points, p = 0.02), after controlling for age, gender, BMI, smoking, insomnia, OSAS, migraine, psychological distress (HADS ≥ 8) and PLMD.
The SF-36 domain scores between snoring patients with and without morning headache. PF = physical functioning; RP = role limitations due to physical problems; BP = bodily pain; VT = vitality; GH = general health; SF = social functioning; RE = role limitations due to emotional problems; MH = mental health. *p < 0.005, **p < 0.001.
Predictors for morning headache in habitual snorers
The frequencies, odds ratios (OR) and adjusted odds ratios (AORs) of predictors for morning headache in habitual snorers
HADS = Hospital Anxiety and Depression Scale; OR = odds ratios by univariate analysis; AOR = adjusted OR by multivariate logistic regression after controlling for the other variables.
p < 0.001, *p < 0.05.
Discussion
The study found that 23.5% of habitual snorers had morning headache, with those with OSAS having a higher frequency than those without (27.2% vs. 15.5%). These findings were similar to those of prior studies, in which 7.4% to 33.6% of OSAS subjects (2,3,5–10) and 16.2% of snorers without OSAS had morning headache (9), even though different diagnostic criteria for morning headache were used. Of 63 habitual snorers with morning headache, 12 (19%) suffered from morning headaches with characteristics fulfilling the criteria of migraine attacks. Eleven of them had a history of migraine, accounting for 31.4% (11/35) of migraine patients with morning headache. Based on the ICHD-2 criteria for sleep apnoea headache, 62% of the morning headaches in our OSAS patients had at least one of three proposed headache symptoms (A1 to A3, Table 1). This is because only 40% of the patients with morning headaches had tension-type features (A2), and only one-third had a short duration (A3). We do not know if comorbidity with migraine in our snoring patients complicated the headache pattern, or the subjects fulfilling these ICHD-2 criteria were more compatible with OSAS-induced headache, because our study did not test these criteria by treating these morning headache patients with OSAS.
Our study found morning headache had a pervasive impact on all domains of the HRQoL as assessed by the SF-36. The differences in scores ranged from 10.6 to 29.7 points, much higher than the suggested scores of clinical significance (>5 points). Moreover, morning headache independently accounted for role limitation due to physical and emotional status, impairment of social functioning, and pain. This suggested that snoring patients comorbid with morning headache were more disabled in their daily activities. In fact, these involved domains and the magnitude of difference were consistent with previous studies of migraine patients (22–24).
Snoring is a sensitive symptom for OSAS (13), and was previously considered an important predictor for morning headache. In contrast to our prior belief, migraine was the most important predictor for morning headache in snoring patients, whereas OSAS was the least. The reasons for the differential impact of these factors were not clear. It is reported that more than 71% of migraine patients had experienced headache on awakening (25). Insomnia and psychological distress are well-known triggers or comorbidities of migraine (26,27), and insomnia-related migraine attacks were more predominant in the morning (28). In contrast, the frequencies of morning headache were similar among patients with different sleep disorders, not only for OSAS (5,6). In our study, more than half of habitual snorers with migraine had morning headache, and 31.4% of their morning headaches fulfilled the criteria for migraine features. The high frequency of OSAS in snorers both with and without morning headache (79.4% vs. 65.4%) downplayed the relationship between OSAS and morning headache. Because we did not have a control group of non-snorers, we do not know if there is a discrepancy in the degree of association between migraine and morning headache in non-snorers.
The pathophysiology of morning headache is still unknown. Hypoxia, hypercapnia (29) and the transient increase of intracranial pressure (30) in OSAS patients were suspected causes. Our study results showed that except for OSAS, the other PSG sleep parameters were not associated with morning headache. Therefore, morning headache in habitual snorers can be, in part, considered as morning attacks of their prior headaches. In addition, we postulated that the four predictors relative to morning headache raised the possibility of hypothalamic involvement. Migraine was related to hypothalamic activation in both neuroimaging and hormonal studies (31,32). The hypothalamus also accounts for sleep regulation and depression (33,34). Recently, abnormal hypothalamic activation was also noted in OSAS patients (35). Further studies are suggested to investigate the role of the hypothalamus in the morning headache of habitual snorers.
Our study has clinical implications. The presence of morning headache can be used as an indicator for poor quality of life and possible associations with migraine, insomnia and psychological distress in habitual snorers. Management of migraine, insomnia or psychological distress might be more helpful for morning headache than management of OSAS. The disruptive effect of CPAP on sleep should be taken into consideration in the treatment of morning headache in snorers, because the potential adverse event of sleep disturbance (36,37) may in turn worsen morning headache.
The strengths of this study are the detailed data collection of headache profiles, HRQoL and the PSG study for each participant. However, some limitations need to be addressed. First, all our patients were referred for PSG under a possible diagnosis of OSAS; therefore, one should be cautious in extrapolating to other patient populations. Second, the difference in the definitions of morning headache between our study and prior studies may produce different results. Third, morning headache was diagnosed according to self-reported headache profiles in the past 6 months, and no headache diary was collected in this study. Recall bias should be considered.
Footnotes
Funding
This study was supported in part by grants from the National Science Council of Taiwan (97-2628-B-010-007-MY3), Taipei Veterans General Hospital (V98C1-143), Ministry of Education (Aim for the Top University Plan), and Taiwan Department of Health Clinical Trial and Research Center of Excellence (DOH100-TD-B-111-004).