We found that smoking is associated with self-reported visual impairment among older adults with ARED, especially those with cataract and AMD. Smoking may be linked to ocular diseases and conditions including cataract, AMD, Graves' ophthalmopathy, ocular irritation, and ocular ischemia (11
). However, few studies have addressed the association between smoking and visual impairment. A New Zealand study suggested that the blindness of 26.8% of registered blind people aged 55 years or older with AMD and 9.5% of those with cataract was attributable to smoking (12
). Results from the Canadian Study of Health and Aging suggested that the odds of self-reported visual impairment were 2.8 times as high among smokers than nonsmokers (13
). Although we did not find odds this large, our findings suggested that smoking was independently associated with visual impairment among older adults with ARED (current smokers vs never smokers).
Smoking is strongly associated with the development of cataract, particularly nuclear cataract. Weintraub et al found that, compared with current smokers, former and nonsmokers were less likely to have cataract extraction (14
). Recent findings from the Visual Impairment Project and the Blue Mountains Eye Study confirmed a higher risk and dose response of smoking and having nuclear cataract (15
). We also found an independent significant association between smoking (current smokers vs never smokers) and visual impairment among those with cataract. However, we did not find a significant difference between former smokers and never smokers.
Our findings suggested a strong association between smoking and visual impairment among people with AMD that is consistent with findings in the British population. After controlling for potentially confounding factors, current smokers were twice as likely to have AMD that caused visual impairment compared with nonsmokers in a British study; for former smokers, the association was not significant (17
). We found that visual impairment among BRFSS respondents with AMD was higher among current smokers than never smokers. Smoking seemed to be associated with more severe AMD cases. Similarly, the Beaver Dam Eye study suggested that, after controlling for age, sex, and baseline AMD severity, current smokers were at higher risk of progression of AMD than nonsmokers (18
). The Blue Mountains Eye Study found that current smokers had an increased risk of 5-year incidence of late age-related maculopathy lesions and developed maculopathy at a significantly earlier age (19
). These findings suggest a possible effect of smoking on progression of AMD.
Although the 2004 Surgeon General's report found insufficient evidence to infer a causal relationship between smoking and AMD, more recent studies indicate a strong association between current smoking and AMD. Smoking was associated with an increased frequency of recurrence of exudative AMD after laser photocoagulation, the only proven treatment for AMD (20
). According to a recent systematic review, the evidence meets the criteria for causality (21
). Recent findings from large population-based studies such as the Beaver Dam Eye Study also indicated an increased risk of incident early AMD during a 15-year follow-up (18
). Furthermore, the number of pack years of smoking was found to be strongly associated with AMD (22
Our findings differ from those of the Beaver Dam Eye Study, which suggested that smoking plays a minor role in primary open-angle glaucoma (POAG) (23
). A meta-analysis of 7 reports (4 cross-sectional studies and 3 case-control studies) before 2003 suggested a pooled odds of POAG of 1.37 for current smokers and 1.03 for former smokers (24
). However, a recent systematic review of case-control and cohort studies by Edwards et al concluded that little evidence exists for a causal association between smoking and development of POAG (25
Some findings suggest a role of smoking in poorer glycemic control (elevated hemoglobin A1c levels), escalating insulin resistance, and an increase in microvascular complications (eg, microalbuminuria) (26
). Furthermore, current smokers with type 1 diabetes have higher odds of severe hypoglycemia than patients with type 1 diabetes who do not smoke (29
). A prospective cohort study of CARDIA (coronary artery risk development in young adults) indicated that both active and passive smoke exposure play a role in developing glucose intolerance (including impaired fasting glucose and diabetes) (30
). Although the literature in general suggested little or no association between cigarette smoking and the incidence or progression of DR (10
), Mouton and Gill found that smoking influences the severity of DR (31
). In this study, we found no significant relationship between smoking and visual impairment among BRFSS respondents with DR.
In summary, smoking is a major modifiable risk factor for AMD (11
), and smoking is the most important modifiable risk factor for primary and secondary prevention of cataract (10
). A lack of awareness may exist among health care providers and patients about the risks of developing eye diseases and vision loss from smoking. Of an estimated 61 million adults in the United States who are at high risk for serious vision loss due to aging, diabetes, or vision or eye problems, only half visit an eye care provider annually, making the situation even worse (32
). Given the effect of smoking on a person's overall health, and especially on sight, comprehensive tobacco control interventions, as recommended by the Centers for Disease Control and Prevention (CDC), are needed (10
). These interventions include health care provider counseling, telephone quit lines, insurance coverage for cessation therapies, legislation of clean indoor air, and increased tobacco taxes (33
This research is subject to several limitations. First, the prevalences of visual impairment and eye diseases are self-reported and may differ from objective clinical measurements. Data on family history of eye diseases were unavailable. Social desirability bias may have caused some current or former smokers to identify themselves as former or never smokers. Accordingly, the actual effect of smoking may be larger than what we found. Second, the data were collected by telephone survey and may not be representative of people without landline telephones. Although low, the BRFSS response rate is in the normal range for telephone surveys. BRFSS data are valid and reliable when compared with other household surveys (34
). Third, institutionalized populations (eg, nursing home residents) are not included in the BRFSS. Fourth, only a few states used the BRFSS vision module, so our findings may not be nationally representative. Fifth, data on frequency and quantity of tobacco use were not obtained and could not be analyzed. Finally, data were cross-sectional; therefore, we were unable to identify causal relationships.
In conclusion, self-reported smoking is linked to self-reported visual impairment among older adults with ARED. The associations between smoking and visual impairment were mostly observed among BRFSS respondents with cataract and AMD but not among those with glaucoma and DR. Further longitudinal evaluation is warranted to explore how smoking cessation or prevention might benefit vision preservation.