Prevalence of TVD
Totally 12,255 patients who visited our headache clinic during the 10-year study period were enrolled. Migraine was diagnosed in 9,966 patients, while the other 2,309 patients were excluded due to non-migraine headache. Among the migraineurs, 962 patients diagnosed with MA and 1,784 MO patients with VARS ≥ 4 were excluded from the study. After excluding those with aura or high VARS, the remaining 7,200 patients with MO constituted the final sample. The MO patients (female/male 5,620 (78.1%)/1,580 (21.9%); mean age: 40.1 ± 13.4 years) were divided into 2 subgroups based on the presence (MwTVD, n = 2,488, 34.6%) or absence (MwoTVD, n = 4,712, 65.4%) of TVD. A flow chart of the patient enrollment process is presented in Figure 1.
TVD characteristics and clinical features
Among the 2,488 MwTVD subjects, 865 (34.8%) patients reported more than one pattern of TVDs. The most common TVD was blurred or foggy vision (n = 1,766, 71.0%), followed by flickering dots/lines (n = 1,332, 53.5%) and zigzag flashes (n = 279, 11.2%). More than a quarter of the patients (n = 683, 27.5%) reported both positive TVDs (flickering dots /lines, or zigzag flashes) and negative TVDs (blurred/foggy vision). Most of the TVDs happened during the headache (74.1%), while 18.4% happened before headache and 7.5% after the headache. The onset of TVD was commonly quick (59.2% of VD occurred in ≤ 30 seconds, 22.7% between 30 seconds to 1 minute, and 15.3% between 1-5 minutes), and the duration was also short (51.6% lasted ≤ 30 seconds, 23.0% lasted for 30 seconds to 1 minute, and 15.8% lasted for 1-5 minutes). The detailed characteristics of the TVDs are shown in Table 1. The onset time, duration, and temporal relationships with headache are the three major different characteristics of TVD comparing with typical visual aura. To validate phenotype-based diagnosis of TVD, we compared these three characteristics between our MwTVD and MA groups. For visual symptoms that did not fulfill typical visual aura but had at least two of these three characteristics: A. Develops or spreads in less than 5 minutes; B. Lasts last than 5 minutes; C. Occurs during headache phase, the sensitivity and specificity to identify TVD were 96.38% and 64.03%. Both the PPV and NPV were 0.87.
Details of demographic data and clinical characteristics of MwTVD and MwoTVD are listed in Table 2. In general, MwTVD had worse clinical features than MwoTVD. The MwTVD group, compared with the MwoTVD group, had higher headache frequency, more severe headache-related disability, higher proportions of CM, MOH and psychiatric comorbidities, and were more likely to be photophobic.
TVD, photophobia and migrainous features
There is a trend for increasing TVD frequency with increasing migrainous features (Fig. 2A). While the TVD was reported in only 12.2% of MO patients with least migrainous features, its prevalence increased to 46.6% in those with all six migrainous features. To investigate whether patients with TVD were also more photophobic, we further compared the MPS between MwTVD and MwoTVD. Consistent with our speculation, MwTVD patients had higher MPS while comparing with MwoTVD (2.9 ± 2.0 vs. 1.6 ± 1.9, p < 0.001, Table 1). Moreover, we found MPS was a predictor of TVD (unadjusted OR = 1.37, 95% CI = 1.32‒1.42, p < 0.001), and the result remained after gender and age were adjusted (adjusted OR = 1.36, 95% CI = 1.31‒1.41, p < 0.001). Similar to TVD, there is also a trend for increasing photophobia frequency with increasing migrainous features, expanding from 15.6% to 38.4% (Fig. 2B).
Suicide risk in patients with migraine with visual disturbance and photophobia
Because the percentages of suicidal ideation and attempts were much higher in the MwTVD group than those in the MwoTVD group (31.9% vs. 18.1%, p < 0.001 for suicidal ideation; 8.2% vs. 3.5%, p < 0.001 for suicide attempts, Table 2), we further explored whether TVD could be an independent risk factor for suicidal risk in patients with migraine without typical visual aura. Univariate analysis showed that TVD, photophobia, headache frequency and headache-related disability, CM, MOH, along with traditional risk factors including depression, anxiety, and poor sleep quality, were associated with higher risks of suicidal ideation and attempts in patients with migraine, while marriage had a protective effect in suicide ideation (Table 3). In multivariable analysis, TVD remained an independent risk factor for suicidal ideation and attempts even after controlling for demographics, headache frequency, headache-related disability, CM, MOH and psychiatric comorbidities (Table 4).