The Seasonality of Acute Attack of Primary Angle-Closure Glaucoma in Beijing, China

In this study, the seasonality of acute attack of primary angle-closure glaucoma (PACG) was analysed. This retrospective case series included 283 patients (200 women, 83 men; mean age, 68.2 ± 10.3 years; range, 37–96 years) with acute attack of PACG from a university-based clinic over 4 years. Patients’ age and sex, and the date and season of onset of PACG attack, were analysed. Descriptive analysis and von Mises distribution were used for statistical analysis. The highest incidence of acute attack of PACG was observed in those aged 60–69 years (34.6%). Descriptive analysis showed that the incidence was greater in June and July for men, November for women, and November for the entire sample. An angular plot (using von Mises distribution) of the individual dates of onset revealed the estimated peak onset on September 11, November 8, and October 28 for men, women, and both, respectively. Integration of the results from the two analyses revealed the incidence to be higher in the summer and winter for men, and in the winter for women and for the entire sample. More females than males were affected. Monthly and seasonal variations in onset were observed, which might be related to weather changes.

Glaucoma is one of the leading causes of blindness worldwide 1 . Acute attack of primary angle-closure glaucoma (PACG) is an ophthalmologic emergency. Over the past decades, there have been few published epidemiological studies investigating the incidence and seasonality of acute attack of PACG in the general population, and an association, though with a low level of significance, has been reported between PACG attack and seasonal variations [2][3][4] . A study from Taiwan found a significant association between relative humidity and monthly admission rates for PACG among males, and for the population in the age range of 60 to 69 years 5 . A higher incidence was also noted during months with less sunlight (November, December, and January), due to mydriasis 6 . A statistically significant predominance of attacks occurring during summer and winter was observed in the Negev region of Israel 7 . In contrast, a study from southern Croatia demonstrated absence of a statistically significant association between the incidence of acute angle-closure glaucoma and seasonal variation 2 . Another study from Croatia also reported no seasonal variations in the occurrence of acute PACG and no correlation of the incidence of PACG with the mean duration of light exposure per season 3 . Nevertheless, in Croatia, the correlation between acute glaucoma and sunshine in November cannot be directly attributed to meteorological factors, although the amount of sunshine in different seasons (p < 0.01 in winter) and the incidence of acute angle-closure glaucoma seem to be inversely correlated 4 . So far, there have been no studies investigating this issue in mainland China. Hence, in the present retrospective study, we examined the relationship between the incidence of acute attack of PACG and seasonal variations, sex, and age at onset.

Methods
The research protocol was approved by the ethics review board of the People's Hospital of Peking University. The study procedures were performed in accordance with institutional guidelines and the declaration of Helsinki. Informed consent was obtained from all patients after providing them a full explanation of the procedures. The study was a retrospective chart review of 283 hospital outpatients treated for PACG at the Eye Centre of People's Hospital of Peking University, Beijing, China, over a 4-year period from June 2011 through May 2015. The Monthly variation in incidence. Descriptive analysis: The incidence of acute attack of PACG was highest in June and July for men, November for women, and November for the combined sample (Table 2, Fig. 2A-C).
In the application of the von Mises distribution, data from each calendar year are standardized to 365 days and then converted to an angle between 0° and 360°. We illustrated the data on attack of PACG graphically in a rose diagram format (angular plot, Fig. 3A), where each petal represented a 'standard' month or an angle of 30° (360°/12). In Fig. 3B, R is the magnitude of the peak estimated by the length of the mean resultant, and μ stands for mean direction. As κ → 0, the distribution tends to become uniform. The larger the value of κ, the greater is the clustering around the mode 9 . The shapes of the corresponding (μ; κ) with μ = 0° for various values of κ are shown.  Graphical analysis and von Mises distribution: An angular plot of the 283 individual dates of onset of PACG is shown in Fig. 3A, and the corresponding data are presented in Table 3. For the combined sample (i.e. including both men and women), the estimated peak of onset was on October 28 (95% confidence interval [CI], August 28 to December 28), with μ 0 = 296.15°. In men, the estimated peak of onset was on September 11, with μ 0 = 250.40°. In women, the estimated peak of onset was on November 8 (95%CI, August 25 to January 22), with μ 0 = 307.10°. All three of these peaks were statistically significant (U2 Watson = 57.439 and p = 0.005 for the combined group; U2 Watson = 13.458 and p = 0.005 for men; U2 Watson = 43.635 and p = 0.005 for women) 11 , but were of modest magnitude (R = 0.08 and κ = 0.17 for the combined group; R = 0.06 and κ = 0.13 for men; R = 0.10 and κ = 0.20 for women).

Seasonal variation in incidence.
In Beijing, the spring season spans from February to April, summer from May to July, autumn from August to October, and winter from November to January. The data on seasonal variations in the number of PACG outpatients for each sex, and for the combined sample are summarised in Table 4.    In the descriptive analysis, the incidence of acute attack of PACG over the 4 years was highest in the summer and winter for men, and in winter for women as well as for the combined group (Table 4, Fig. 4).
According to the results of the von Mises distribution, the highest incidence of acute attack of PACG for females (October 28) and for the entire sample (November 8) appears to be in the winter (around November). Thus, there were statistically significant (p = 0.005, Table 3) seasonal variations in the occurrence of acute attack of PACG.

Discussion
In this retrospective study, we investigated the seasonality of acute attack of PACG in outpatients treated at a hospital in Beijing. The observation that acute attack of PACG affects women more frequently than men 12 has   been reported previously, and is in line with the results of our study; we found that acute attack of PACG affected women more than twice as often as men. Various authors have reported the highest risk for acute attack of PACG to be around the seventh decade (60 to 69 years of age) of life [5][6][7] . In our study, the highest incidence was observed in the same age group. The prevalence of acute attack of PACG has been found to increase proportionately with age for different racial groups 12 . This age-related increase may be due to the anterior chamber becoming shallower with increasing age 13 . Other possible reasons for increasing predisposition to an acute attack of PACG include the following: use of mydriatic drugs 6,14 ; low background illumination 14 ; emotional disturbances (antidepressant use) 15, 16 ; increased antero-posterior lens diameter due to prolonged reading 6 ; performing sustained Yoga postures 17 ; meteorological factors such as sunspot activity 18 , and increased relative humidity 5 . A previous study from Romania reported the following: "the increase of antero-posterior lens diameter (intumescent lens) is just a starting factor for a potential primary closed-angle glaucoma" 6 . However, from the circular plot in our study, it is clear that there are some evidences of more than a single mode which causing acute attack of PACG, but these evidences do not appear to be strongly related to, for example, sex (Table 3); the peak onset for women occurred 58 days later than that for men. Nevertheless, the von Mises distribution provides a reasonable description of the data, with a stronger peak onset of PACG indicated for different age group with p = 0.005 (Table 3). Hence, there were statistically significant monthly and age-dependent variations in the incidence of acute attack of PACG.
There are limited data in the existing literature regarding the seasonality of acute attack of PACG in mainland China, though similar studies have been conducted in other parts of the world. November and December have been described as the months of the peak incidence in Finland 19 . In the United Kingdom also, an increased incidence has been reported in the winter months, especially in December 18 . While in Singapore, the incidence was found to be higher on hotter days (i.e., days with a higher number of attacks appeared to be hotter and drier, with more direct sunshine) 20 . A study by David et al. also found that in Israel, a significant number of attacks occurred during the summer and winter 7 . In our study, we observed a significant predominance of attacks during the summer and winter (for men) and during the winter (for women and for the combined sample), which is in line with the observation of a higher incidence of acute attacks of PACG coinciding with periods of extreme temperatures in the above-mentioned regions.
Meteorological conditions have previously been reported to be associated with the attack of PACG, and early significant results were only considered to be associated with shortened duration of sunshine. In Finland, an increase in the incidence of acute angle closure glaucoma was noted whenever the number of hours without sunshine increased 21 . Teikari et al. reported that the incidence of acute attack of PACG was higher in winter and autumn compared with spring and summer 21 . However, in Taiwan, hospital admission rates of PACG were significantly higher in March and with increased relative humidity 5 . Therefore, we believe that cold weather, especially in case of women, and hot dry summers especially in case of men, tend to drive everyone 22 , especially the elderly 23 , indoors (where poorer light acts as a surrogate for the darkroom test and makes patients susceptible to glaucoma).
In our study, we have investigated the difference between males and females in mainland China while previous studies were conducted on a combined sample and did not look into sex-related differences in the seasonality of acute attack of PACG. We used more advanced and robust statistical methods to analyse the trends in the incidence of acute attack of PACG, in contrast to previous studies, which used only the chi-square test; angular analysis based on von Mises distribution is more appropriate for seasonal studies 8,9 . This is a major strength of our study over previous studies. However, in our study, results on the peak onset in males obtained by descriptive analysis (June and July) were different from those obtained by Von Mises analysis (September 9th); therefore, the association between seasonal variations and acute attack of PACG is not as clear in men as that in women. Thus, we acknowledge that our statistical methods may need further improvement. Better methods for detecting and treating this disease are still needed, as is optimization of the appropriate statistical approach. In conclusion, our results indicate that seasonal variations during the year seem to be an important factor affecting the occurrence of acute attack of PACG. Findings from the present study will contribute to an improved understanding of the epidemiology of acute attack of PACG.