We investigated the relationship between photosensitivity and various metrics of sleep including sleep quality, sleep disturbance, sleep onset latency, and sleep-related impairments in patients with migraine. The results were obtained using existing data extracted from ARMR from 852 patients who met the inclusion criteria. Poor sleep quality was prevalent among our participants, with over 46% reporting lack of good sleep quality. In line with several previous inquiries [15, 38], our data exhibited insomnia as the most frequently reported sleep disorder among patients with migraine. Additionally, this study presents novel findings, demonstrating that patients with migraine with generalized photophobia tend to have poorer sleep quality compared to those without photophobia (including patients with photophilia). Both photophobia and photophilia exhibited statistically significant associations with sleep quality, sleep disturbance, and sleep onset latency. As we hypothesized, greater severity of photophobia emerged as a predictor for compromised sleep quality, increased sleep disturbance, and prolonged sleep onset latency. Conversely, a tendency for light, or photophilia, was correlated with enhanced sleep quality and lower levels of sleep disturbance and sleep onset latency. These results persisted after controlling for patient age, sex, anxiety, and depression. Our findings are aligned with a limited number of prior studies that reported photophobia as a potential risk factor for sleep disorders in migraine [22, 23] as well as with the emerging body of literature highlighting the complex effects of light exposure on sleep quality [39–41].
Prior studies have demonstrated that retinal light exposure serves as a direct stimulant for the brain’s biological clock, exerting profound effects on various facets of sleep such as sleep pattern, sleep onset latency, sleep duration, and sleep efficiency [42–44]. The rhythmic interplay of light and darkness experienced throughout the day intricately orchestrates the timing of the circadian clock. To synchronize harmoniously with the natural light-dark cycle of Earth, a consistent exposure to high levels of morning light and a contrast between daytime and nighttime light exposure are essential. In a study by Figueiro et al. [45], it was observed that exposure to high levels of lighting in the morning was associated with reduced sleep onset latency, increased phasor magnitudes (which is a measure of circadian entrainment), and increased sleep quality in adult participants. These findings were aligned with other similar inquiries on other age groups [46–48]. Whereas there is growing evidence for circadian relevance in migraine and other pain disorders [21, 49], none of the previous studies were focused on circadian effects of light on patients with migraine with photosensitivity.
A prevalent coping strategy to alleviate discomfort and pain induced by migraine-associated photosensitivity involves seeking refuge in darkness, often by spending extended periods in dimly lit or dark rooms and donning dark or tinted sunglasses. These behaviors not only reduce the retinal light exposure during the daytime in patients with migraine, but also inadvertently result in constant lighting conditions, devoid of the natural contrast between the day and night. Consequently, the biological clock is deprived of a crucial time cue to stimulate the circadian clock accordingly, potentially leading to disruptions in sleep, prolonged sleep onset, and poor sleep quality.
In those with migraine, insufficient sleep and sleep disturbance have consistently emerged as frequent headache triggers [12, 50–52]. Conversely, adequate sleep timing, duration, and quality has been reported to be protective against headache [53]. Given the therapeutic effects of high-quality sleep in managing migraine and the potential adverse effects of pharmacological treatments, implementing non-drug interventions appears to be an effective initial approach to improve sleep quality in this population. Lighting, in particular, has been demonstrated as a potent non-pharmacological intervention capable of improving sleep and circadian activity rhythms, among other benefits [39–41, 54, 55]. In a study by Burgess et al. [56], the findings revealed that bright light therapy in the morning advanced circadian phase which was associated with an increased tolerance to pain. The positive association we found between photophilia and improved sleep outcomes further supports the potential beneficial effects of light exposure on sleep outcomes amongst those with migraine. It is reasonable to assume that patients with migraine with photophilia tend to receive more retinal light exposure during the daytime which could potentially be the reason for enhanced sleep outcomes in this population. In a recent open-label study, self-reported data revealed that two hours of exposure to narrow band green light led to an improvement in photophobia and same-night sleep for 53% and 49% of all migraine attacks, on average, respectively. [18]. However, it is not clear whether the improvement in sleep was due to the effects of light exposure on circadian activity rhythms or as a result of enhanced mood, photophobia, or headache improvement.
Circadian stimulation by light is achieved through exposure to high-intensity lighting with a blue-enriched spectrum during the daytime. Nevertheless, the prevalence of photophobia in migraine poses a challenge to the implementation of circadian lighting for this population, given their severe sensitivity to high intensity light, particularly in the blue spectrum [57, 58]. The emergence of new lighting technology holds promise for enhancing well-being among those with migraine through offering customized lighting solutions with optimized intensity and spectrum tailored to individuals’ needs, while maintaining circadian rhythms by following a natural light/dark cycle. Exposure to proper circadian lighting not only has the potential to improve sleep quality in patients with migraine but also could contribute to positive effects on anxiety and depressive symptoms – both highly associated with migraine and photophobia.
The recruitment of patients from nine different sites across the United States, the large sample size, and the diverse geographical locations of enrollment enhance the potential generalizability of our findings. However, the findings of this study are limited as we only included patients from headache specialty centers who tended to have high frequency migraine and likely represent a population with severe migraine disease. To improve generalizability, future studies should include a more diverse population. Our study is also limited by its reliance on self-reported data for photosensitivity and sleep quality, collected through instruments selected by the ARMR study team. Self-reported data are inherently subjective as they rely on individuals’ subjective perceptions, introducing the possibility of recall bias. Additionally, the photosensitivity questionnaire (PAQ) utilized in ARMR, while well-validated, lacks specific assessments of headache-related light sensitivity, focusing instead on non-specific behavioral preferences. Similar limitations apply to the sleep questionnaire, which incorporates partial components from various instruments and does not include the entire set of questions from a single validated questionnaire. That makes interpretation and analysis of the sleep data challenging. Future studies should consider supplementing subjective measures with objective ones (i.e., actigraphy, wearable light trackers) to acquire precise data on sleep quality and daily light exposure behavior among patients with migraine. Furthermore, experimental trials aimed at investigating the effects of various lighting interventions on sleep quality and circadian health of those with migraine could offer valuable insights. Such research might pave the way for the development of novel lighting technologies, contributing to the improvement of well-being in this population.
Lastly, it is important to acknowledge a limitation inherent in the ARMR dataset [7], which pertains to the unbalanced racial distribution of the sample. The participants in this registry are predominantly White, a characteristic that is apparent in the sample included in our analysis (94.1% White). This feature of the dataset, undeniably, limits the generalizability of our findings. It is imperative for future research to include a more diverse sample for extrapolating the results.