Background
The onset and duration of spontaneous migraine attacks are most often difficult to predict which, in turn, makes it challenging to study the neurobiologic underpinnings of the disease in a controlled experimental setting. To address this challenge, human provocation studies can be used to identify signaling molecules (e.g. calcitonin gene-related peptide, pituitary adenylate cyclase-activating polypeptide) that, upon intravenous or oral administration, induce migraine attacks in people with migraine and mild or no headache in healthy volunteers. This approach has proven to be valid for decades and plays an integral role in mapping signaling pathways underlying migraine pathogenesis and identification of novel drug targets. However, the question arises as to whether the pathogenic mechanisms of provoked and spontaneous migraine attacks differ. In this paper, we provide an opinionated discussion on the similarities and differences between provoked and spontaneous attacks based on the current understanding of migraine pathogenesis.

Methods
The PubMed database was searched in July 2022 for original research articles on human provocation studies that included participants with migraine. The reference lists of originally identified articles were also searched and we selected those we judged relevant.

Discussion
People with migraine describe that provoked attacks resemble their spontaneous attacks and can be treated with their usual rescue medication. From a neurobiologic standpoint, provoked and spontaneous migraine attacks appear to be similar, except for the source of migraine-inducing substances (exogenous vs. endogenous source). In addition, provoked attacks can likely not be used to study the events that precede the release of migraine-inducing signaling molecules from sensory afferents and/or parasympathetic efferents during spontaneous attacks.