Meteorological extremes and their impact on tinnitus-related emergency room visits: a time-series analysis

Study population and weather

From January 1st, 2015, to December 31st, 2018, 526 tinnitus-related EVs occurred at the Vienna General Hospital in Vienna, Austria. Patient characteristics for each EV are shown in Table 1. The study population had a higher proportion of male patients (58.4%) with a median age of 37 (range: 11–89). Most patients suffered from one-sided tinnitus (75.5%) and showed no indication of hearing loss, as determined by Weber’s and Rinne’s test (80.4%). In those with pathologic results, 15.7%, 1.3%, and 2.6% of patients showed signs of sensorineural loss, conductive loss and combined loss, respectively. The etiology was largely idiopathic (95%). However, some cases were attributable to recent acoustic trauma (2.5%) and cerumen (2.5%). From the small number of cases with available blood pressure measurements (n = 88), 55.7% of patients were hypertensive. On average, 2.5 tinnitus-related EVs occurred per week. Daily presentation rates throughout 2015–2018 are shown in Fig. 1. While the highest numbers of tinnitus-related EVs occurred during January (n = 55) and October (n = 54), no clear seasonal trends were observed. All meteorological variables in Vienna during the observational period are shown in Supplementary Fig. 1. Relative humidity and mean temperature showed clear differences between the seasons. Atmospheric pressure, precipitation and mean wind speed were less affected by seasonality.

Table 1 Patient characteristics of the study cohortFig. 1figure 1

Weekly tinnitus-related EVs are shown as the percentage of total EVs from any cause from 2015 to 2018 using smoothing splines (lines) and daily values (background)

Atmospheric pressure

Based on previous reports on the link between atmospheric pressure and tinnitus symptoms in patients suffering from MD [21, 29], we first aimed to investigate the effect of extreme atmospheric pressure events on tinnitus symptoms in patients without MD.

On the same day, extreme atmospheric pressure events did not significantly affect tinnitus-related EV. Over the subsequent lag period of single-day events, the earliest impact of extremely low atmospheric pressure occurred after 3 days at 983 hPa (P5) with an RR of 1.40 [1.04–1.89; p = 0.028]. For the same condition, cRR was elevated within 4 days to 1.88 [1.03–3.44; p = 0.038] and was significantly elevated for the remaining lag period with the highest cRR of 3.24 [1.04–10.15; p = 0.044] at day 14 (Fig. 2A). Similarly, extremely high atmospheric pressure at 1009 hPa (P95) increased RR to 1.12 [1.01–1.25; p = 0.028] starting at day 8 and elevated cRR to 2.31 [1.04–5.10; p = 0.038] on day 11.

Fig. 2figure 2

Line-plots of cRR for tinnitus-related EV are shown for single-day extreme weather events from lag0 to lag14 defined as the 1th, 5th, 95th, and 99th percentile of atmospheric pressure in hPa (A), relative humidity in % (B), mean temperature in °C (C), precipitation in mm [P95 & P99 only] (D) and mean wind speed in m/s (E). Confidence intervals (95%) are shown in grey. Significant decreases (green) and increases (red) in cRR (p ≤ 0.05) are highlighted on the lag-axis

Prolonged extremely low atmospheric pressure showed comparable results with the earliest elevation of cRR at 980 hPa (P1) to 1.7 [1.09–2.66; p = 0.020] within 4 days and at 985 hPa (P5) to 1.42 [1.09–1.85; p = 0.010] within 3 days, respectively (Fig. 3A). Following the single-day results, extremely high atmospheric pressure over 3 days at an average 1008 hPa (P95) raised the cRR to 1.36 [1.04–1.77; p = 0.026] within 3 days and led to a sustained cRR elevation over the remaining 14-day observational period with its maximum on day 13 at 2.23 [1.43–3.47; p < 0.001].

Fig. 3figure 3

Line-plots of cRR for tinnitus-related EV are shown for prolonged extreme weather events over three days from lag0 to lag14 defined as the 1th, 5th, 95th, and 99th percentile of atmospheric pressure in hPa (A), relative humidity in % (B), mean temperature in °C (C), precipitation in mm [P95 & P99 only] (D) and mean wind speed in m/s (E). Confidence intervals (95%) are shown in grey. Significant decreases (green) and increases (red) in cRR (p ≤ 0.05) are highlighted on the lag-axis

Taken together, atmospheric pressure increased tinnitus-related EV risk at low and high extremes as early as 3 days after the weather event.

Relative humidity

After we found extreme atmospheric pressure conditions to be a risk-increasing factor for tinnitus-related EV, we then analyzed the impact of relative humidity due to its correlation with Ménière’s attacks [21].

On the same day, extremely high relative humidity at 92% (P99) increased the risk for tinnitus-related EVs to 1.75 [1.01–3.03; p = 0.046]. Over the subsequent lag period for single-day events, extremely low relative humidity at 34% (P1) significantly decreased RR between 8 and 11 days after the event. cRR at 34% (P1) was reduced within 1 day to 0.40 [0.16–1.00; p = 0.050] and remained significantly decreased over the entire 14-day observational period with a cRR as low as 0.09 [0.02–0.39; p = 0.002] by day 12 (Fig. 2B). Extremely high relative humidity had no significant delayed effect.

Prolonged extremely low relative humidity over 3 days impacted the risk for tinnitus-related EV in a comparable way, with a significantly reduced RR between day 8 and day 12. cRR was decreased within 2 days to 0.44 [0.26–0.74, p = 0.002] at − 4 °C (P1) and within 3 days to 0.59 [0.43–0.79; p < 0.001] for 0 °C (P5). (Fig. 3B) An extremely high three-day average relative humidity of 89% (P99) decreased RR to 0.77 [0.62–0.95; p = 0.016] at day 4. However, cRR was significantly increased to 1.78 [1.11–2.85; p = 0.018] within 2 days after a prolonged extremely high humidity of 89% (P99).

In summary, extremely low relative humidity showed a pronounced reductive effect on tinnitus-related EV as early as one day after the event. Conversely, extremely high relative humidity significantly increased same-day risk and showed a bidirectional impact over the subsequent lag period.

Mean temperature

Next, we investigated mean daily temperature as a possible factor in tinnitus-related EV rates based on its described role in tinnitus severity in MD [21].

On the same day, extreme temperature events did not significantly affect EV risk. Over the subsequent lag period of single-day events, extremely low temperatures of 0 °C (P5) led to a decreased RR of 0.66 [0.47–0.92; p = 0.014] on day 4 and a decreased cRR from day 5 to day 7 with a low-point of 0.41 [0.19–0.91; p = 0.028] at day 6 (Fig. 2C). Extremely high temperatures of 27 °C (P95) significantly affected RR in both directions across different lag days, with a reduction to 0.33 [0.12–0.93; p = 0.036] on day 2 and an increase to 1.39 [1.01–1.91; p = 0.044] on day 4. cRR was not significantly impacted by extremely high temperatures.

Prolonged extreme conditions over three days in the form of cold waves exerted a risk-mitigating effect on tinnitus-related EVs (Fig. 3C). The highest reduction of RR was observed on day 4 to 0.68 [0.46–1.00, p = 0.048] after cold waves averaging 0 °C (P5). cRR was reduced after cold waves averaging both − 4 °C (P1) and 0 °C (P5) within 4 to 10 days. The lowest risk was observed at − 4 °C (P1) within 6 days at a cRR of 0.34 [0.18–0.64, p < 0.001]. Heat waves affected RR for tinnitus-related EV in both directions, although the risk-increasing effect was more pronounced. At day 3, RR was increased to 1.83 [1.02–3.30; p = 0.044] for 26 °C (P95) and to 2.61 [1.15–5.89; p = 0.022] for 30 °C (P99). cRR was not significantly affected by heatwaves.

In brief, extremely low temperatures reduced the risk for tinnitus-related EV as early as 4 days after the weather event. Extremely high temperatures showed a non-linear effect on RR starting 2 days after an event with a tendency towards increased risk.

Precipitation

Since the effect of precipitation on tinnitus has not yet been described, we chose to analyze it as an additional potential meteorological factor in tinnitus-related EV rates.

For single-day events, extremely high precipitation had no significant effect on same-day risk, nor did it show delayed effects in the subsequent 14-day observational period. (Fig. 2D) On the other hand, prolonged extremely high precipitation over 3 days had a risk-mitigating effect on tinnitus-related EV. At its earliest, RR was decreased to 0.62 [0.41–0.94; p = 0.026] 1 day after prolonged precipitation totaling 24 mm (P95). cRR for the same condition was decreased within 1 day to 0.53 [0.36–0.78, p = 0.002] and remained significantly reduced over the entire lag period (Fig. 3D). Late effects of prolonged precipitation totaling 40 mm (P99) moderately increased RR between day 8 and day 11 (p < 0.050). However, these effects did not remain significant after cumulation.

Therefore, the data indicates a risk-reducing effect of prolonged, but not single-day, extreme precipitation for tinnitus-related EV as early as 1 day after the weather event.

Mean wind speed

Finally, we investigated mean wind speed, which does not appear to affect tinnitus in MD [21], to dissect potential differences in weather impact in tinnitus unrelated to MD.

On the same day, extreme wind speeds did not significantly impact risk for tinnitus-related EV. Over the subsequent lag period of single-day events, the earliest effect of extremely high wind speeds was observed at 6 m/s (P95) on day 3 with a RR of 1.36 [1.01–1.84, p = 0.046]. However, these effects did not remain significant after risk cumulation. (Fig. 2E) Extremely low wind speeds showed no significant impact on RR or cRR.

Prolonged extremely low wind speeds at 2 m/s led to a decreased RR between day 8 and day 14, with a low-point of 0.88 [0.81–0.95, p < 0.001] by day 14. cRR was significantly reduced by day 12 to 0.67 [0.46–0.98, p = 0.040] (Fig. 3E). Extremely high wind speeds over 3 days only showed late, marginal effects on RR after day 9, with no significant results after cumulation.

To summarize, extremely high single-day wind speeds had minor, early effects on tinnitus-related EV risk by day 2. In contrast, extremely low prolonged wind speeds led to a moderate, late reduction of EV risk starting at day 8.

留言 (0)

沒有登入
gif