Prevalence of smoking and its associations with ocular parameters in adult Chinese: The Beijing Eye Study

Abstract

Editor, Recent population-based studies on Caucasians have shown that smokers have an increased risk for age-related macular degeneration (Thornton et al. 2005; de Jong 2006). Because the smoking habits vary between different cultures and because there have been no population-based data on the associations between smoking and ocular normative parameters and ophthalmic diseases for the Chinese population of mainland China, we conducted the present study to assess the associations between smoking and ophthalmic parameters and selected systemic parameters in China. The Beijing Eye Study is a population-based cross-sectional study in Northern China. It has been described in detail recently (Jonas et al. 2009). First performed in the year 2001, the study was repeated in 2006, with 3251 subjects participating. The study participants underwent a detailed ophthalmic examination and answered a questionnaire including questions on current and former smoking and the mean number of packages smoked per day. Information on smoking was obtained from 3214 (98.9%) subjects. There were 725 (22.6%) current smokers, 1042 (32.4%) current or former smokers and 2172 (67.6%) never smokers. In univariate analysis, the prevalence of current smoking/former smoking was significantly (p < 0.01) associated with male gender, rural region, hyperopic refractive error, lower serum levels of high-density lipoproteins and low-density lipoproteins, a lower level of education and income, higher diastolic and systolic blood pressure, a higher mortality, a more shallow anterior chamber, a lower prevalence of dry eye symptoms, a higher prevalence of pterygium, a higher prevalence of trachoma, and thinner retinal arteries and veins. Smokers and nonsmokers did not vary significantly in age, the prevalences of open-angle glaucoma, angle-closure glaucoma, nonglaucomatous optic nerve atrophy, age-related macular degeneration, retinitis pigmentosa, diabetic retinopathy, myelinated retinal nerve fibers, alpha zone and beta zone of parapapillary atrophy and retinal vein occlusions. In binary logistic regression analysis, current/former smoking versus never smoking was significantly associated with male gender, hyperopic refractive error, rural region, lower level of education and higher intraocular pressure (Table 1). It was marginally significant with mean blood pressure. If the analysis additionally included the serum levels of cholesterol, high-density lipoproteins and low-density lipoproteins, only an elevated serum concentration of cholesterol was additionally associated with smoking (p = 0.05). Interestingly, the multivariate analysis showed that smokers were significantly more hyperopic than the never smokers. This may have clinical importance for ophthalmic disorders, because some ocular diseases are associated with the refractive error. To cite examples, previous studies such as the Rotterdam Study and the Beijing Eye Study have revealed that hyperopia is associated with an increased prevalence of age-related macular degeneration, diabetic retinopathy and angle-closure glaucoma (Ikram et al. 2003; Xu et al. 2006). If in the assessment of the relationships between smoking and these ocular diseases the refractive error as a potentially confounding factor was not taken into account, the correlation between smoking and these diseases may become artificially statistically significant. Smoking was associated with a lower socioeconomic background. Because the socioeconomic background plays a role in the development and progression of diseases, the association between smoking and a low socioeconomic background may have to be taken into account if the role of smoking for the development of progression of ocular diseases is examined. In the multivariate analysis, smoking was not significantly associated with alpha zone and beta zone of parapapillary atrophy. It agrees with previous studies in which parapapillary atrophy was not related to other vascular parameters such as arterial hypertension (Jonas & Gründler 1996). The prevalence of early age-related macular degeneration was not markedly associated with smoking. It agrees with the Copenhagen City Eye Study (Buch et al. 2005), and it contradicts previous population-based studies examining other ethnic groups (Thornton et al. 2005; de Jong 2006). Reason for the discrepancies between these studies may potentially be differences in the statistical analysis, if the potentially confounding associations between smoking, hyperopia and age-related macular degeneration were not sufficiently taken into account. In the present investigation, smoking was not associated with the prevalence of glaucoma. It is in agreement with the recent Los Angeles Latino Eye Study (Doshi et al. 2008). In conclusion, smoking was significantly associated with male gender, rural region, lower level of education, elevated serum concentration of cholesterol, hyperopia and higher intraocular pressure in the adult population of mainland China. In contrast to studies on Western ethnic groups, smoking was not markedly associated with the prevalence of age-related macular degeneration, nor of glaucoma or retinal vein occlusions. The association between smoking as dependent parameter and hyperopia and lower level of education as independent factors may have to be considered when the role of smoking for ocular diseases, particularly for age-related macular degeneration, is examined. Supported by the Beijing Natural Science Foundation.