The average percentage of front-windshield UV-A blockage was 96% (range, 95%-98% [95% CI, 95.7%-96.3%]) and was higher than the average percentage of side-window blockage, which was 71% (range, 44%-96% [95% CI, 66.4%-75.6%]) (P < .001). The percentages of UV-A blockage for each automobile’s front windshield and driver’s side window were calculated by subtracting the UV-A energy measured behind the window from the outside UV-A energy. That figure was divided by the amount of outside UV-A energy. This fraction was converted to a percentage by multiplying by 100.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Boxer Wachler BS. Assessment of Levels of Ultraviolet A Light Protection in Automobile Windshields and Side Windows. JAMA Ophthalmol. 2016;134(7):772–775. doi:10.1001/jamaophthalmol.2016.1139
Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Ultraviolet A (UV-A) light is associated with the risks of cataract and skin cancer.
To assess the level of UV-A light protection in the front windshields and side windows of automobiles.
In this cross-sectional study, 29 automobiles from 15 automobile manufacturers were analyzed. The outside ambient UV-A radiation, along with UV-A radiation behind the front windshield and behind the driver’s side window of all automobiles, was measured. The years of the automobiles ranged from 1990 to 2014, with an average year of 2010. The automobile dealerships were located in Los Angeles, California.
Main Outcomes and Measures
Amount of UV-A blockage from windshields and side windows. The average percentage of front-windshield UV-A blockage was 96% (range, 95%-98% [95% CI, 95.7%-96.3%]) and was higher than the average percentage of side-window blockage, which was 71% (range, 44%-96% [95% CI, 66.4%-75.6%]). The difference between these average percentages is 25% (95% CI, 21%-30% [P < .001]). A high level of side-window UV-A blockage (>90%) was found in 4 of 29 automobiles (13.8%).
Conclusions and Relevance
The level of front-windshield UV-A protection was consistently high among automobiles. The level of side-window UV-A protection was lower and highly variable. These results may in part explain the reported increased rates of cataract in left eyes and left-sided facial skin cancer. Automakers may wish to consider increasing the degree of UV-A protection in the side windows of automobiles.
The risks of sun exposure to the skin and eyes are well known. People who drive cars and other motor vehicles expose the left sides of their faces, including their left eyes, to sunlight and solar radiation because that side is near the side windows, which increases the risk of various diseases in the left eye and on the left side of the face. Ultraviolet A (UV-A) is linked to increased risks of cataract formation1 and skin cancer.2 Studies have found that skin cancers are more common on the left side of the face.3,4 It has been reported that left eyes have higher rates of cataract because drivers are seated on the left side of the automobile (and drive on the right side of the road).5 UV-A light that passes through glass without UV-A protection has led to in vitro cytotoxic effects on cells, whereas glass with UV-A protection dramatically reduced the cytotoxic effect of UV-A light.6 Therefore, auto glass with UV-A protection would be expected to reduce the risks of disorders related to sun damage. In the United States, the level of auto glass UV-A protection for drivers of different makes and models of vehicles is unknown.
Question What is the difference in the level of ultraviolet A (UV-A) light protection between the front windshield and the side windows of an automobile?
Findings In this cross-sectional study, the average percentage of front-windshield UV-A blockage was 96%, which was 25% higher than the side-window blockage. A high level of side-window UV-A blockage (>90%) was found in 4 of 29 automobiles (13.8%).
Meaning These results may in part explain the reported increased rates of cataract in left eyes and left-sided facial skin cancer.
I undertook this study to evaluate the levels of UV-A penetration of a variety of vehicles’ front windshields and side windows to better understand the levels of UV-A protection and the risks for drivers’ left eyes and the left sides of their faces. This study was undertaken in the late morning and afternoon on May 4, 2014 (a cloudless day), when I visited a number of automobile dealerships in Los Angeles, California, with my daughter Micaela Boxer Wachler, who assisted me in the collection of data and is a student. Dealerships were selected for their proximity because they were all clustered within approximately a half a mile from each other. A variety of makes, models, and years of vehicle production were assessed. A handheld UV-A light meter (Omega) was used to assess the amount UV-A radiation. Test-retest reliability was subsequently evaluated, and no measurement with this device differed by more than ±0.02 mW/cm2. We measured external ambient UV-A levels (with the meter pointed in the direction of the sun for each measurement) for each selected automobile; we did not rely on a single measurement for all automobiles because ambient UV-A changes during the day. Immediately following this measurement for a given selected automobile, the UV-A meter was then held on the inside of the driver’s side window for measurement and oriented in the direction of the sun. The UV-A meter was then held immediately behind the driver’s side of the front windshield for measurement and pointed in the direction of the sun.
The percentages of UV-A blockage for each automobile’s front windshield and driver’s side window were calculated by subtracting the UV-A energy measured behind the window from the outside UV-A energy. That figure was divided by the amount of outside UV-A energy. This fraction was converted to a percentage by multiplying by 100. A paired t test was used to compare the UV-A blockage of front windshields with that of side windows, and simple linear regression analysis was performed using 2 variables: side-window protection (dependent variable) and vehicle year (StatView; SAS Institute Inc), in order to assess whether older cars had lower levels of UV-A protection as a possible result of aging UV-A window films. Statistical significance was set at P = .05.
We analyzed 29 automobiles from 15 automobile manufacturers. The years of the automobiles ranged from 1990 to 2014, with an average year of 2010. The average percentage of front-windshield UV-A blockage was 96% (range, 95%-98% [95% CI, 95.7%-96.3%]) and was higher than the average percentage of side-window blockage, which was 71% (range, 44%-96% [95% CI, 66.4%-75.6%]). The difference between these average percentages is 25% (95% CI, 21%-30% [P < .001]). A high level of side-window UV-A blockage (>90%) was found in only 4 of 29 automobiles (13.8%). Regression analysis of the association between side-window protection and vehicle year of manufacturer resulted in the correlation coefficient r2 = 0.04 (P = .75).
The Table provides data on the automobiles tested, along with the UV-A measurements and percentages of UV-A blockage. The Figure shows the distribution of percentages of front-windshield UV-A blockage and the more variable distribution of percentages of side-window UV-A blockage for the same vehicles.
Cortical cataracts have been reported to occur more in left eyes than right eyes.5 We discovered that, after visual inspection of summary photographs of the crystalline lenses in another study,7 there was increased cortical cataract formation in left eyes than right eyes. These results seem to indicate that left-sided sun exposure for drivers plays a role in the development of cataracts.
Cumulative UV-A exposure is a significant risk factor for skin cancer. Multiple studies have found that such cancers are more common on the left side of the face in countries where automobiles are driven on the right side of the road.3,4 The faces and arms of these drivers are preferentially exposed to the sunlight. A study from Australia, where automobiles are driven on the opposite (left) side of the road, found that drivers had more skin cancer on the opposite (right) side of the face.8
The present study found that front windshields, not side windows, offer consistent protection for drivers’ eyes and faces against the harmful effects of UV-A light. The reason lies in the design of the front windshields. These windshields are constructed from 2 planes of glass with a clear layer of plastic “wedged” in between them. This plastic makes the windshield shatterproof in the event of an accident. The plastic layer largely contains the UV-A protection. In contrast, the side windows of cars are a single plane of glass that may or may not have a high level of UV-A–blocking polymers and other protective components. In addition, a number of automobiles had a factory window tint on the side windows, but it did not correlate with increased UV-A blockage.
To the best of my knowledge, I believe that this is the first US study to analyze and compare the percentages of UV-A blockage between front windshields and side windows. A PubMed search revealed one other study9 that assessed the UV-A radiation passing through automobile windows. This study by Castanedo-Cazares et al9 was from Mexico and had similar findings of consistent and high levels of UV-A blockage from all vehicle front windshields. They found that side windows had statistically lower levels of UV-A protection than front windshields. The average percentage of side-window protection in the Mexico study9 was 84%, which was higher than the average percentage of side-window protection of 71% in the present study. In addition, the Mexico study9 did not find any significant differences in levels of side-window UV-A protection among the vehicles tested, whereas there was a large range of differences in levels of side-window protection in the present study. The reason for this difference between the Mexico study9 and the present study is not well understood. The Mexico study9 did not specify which vehicle makes and models were tested.
The Table demonstrates that the highest level of UV-A side-window protection was found in newer Lexus models (92%-96%) and a newer Mercedes automobile (95%). Porsche and BMW models had among the lowest levels of side-window protection (56% and 55%, respectively), which were lower than a 1990 Buick (65%). The high level of variability may be explained by the type of glass used by the automaker. The level of UV-A protection varies depending on glass type, glass color, and coating, among other factors.10 There was reasonable consistency regarding side-window UV-A protection among automobiles of a similar year from the same manufacturer (Lexus and Honda). Some cars of different years from the same manufacturer showed higher levels of side-window UV-A protection in more recent year models (Lexus, VW, and Mercedes). I would expect UV-A measurements to be similar across different ambient lighting conditions; however, we did not assess that variable.
Some patients in my practice have assumed that an automobile with tinted windows will provide a high level of UV-A protection. Tint provides blockage of visible light. Films that purely block UV-A light are virtually clear. Indeed, some tints provide a high level of UV-A protection, whereas others do not. Our observation was that some of the automobiles with tinted windows (eg, BMW) had the lowest levels of UV-A blockage. Therefore, one cannot assume that tinted side windows are providing a high level of protection against UV-A light.
There are limitations to the present study. For example, we evaluated a specific number of makes and models. There are other types of vehicles that were not part of the present study, and the levels of side-window protection for those automobiles are unknown. It is possible that side windows with a lot of scratches could reduce the efficiency of side-window blockage. We did not observe any side windows with noticeable scratches. Aging of the windows could be a factor in reducing UV-A protection, but that is unlikely because we failed to find a correlation between side-window UV-A blockage and year of vehicle. Interestingly, the 1990 Buick Riviera had a higher level of UV-A side-window protection than some of the newer luxury automobiles. Another limitation is that the present study was a correlational study of reduced UV-A protection in side windows and increased cataracts and skin cancer on the left side. If this association is shown to be a cause-and-effect relationship, then automakers may wish to consider increasing the level of UV-A protection in vehicle side windows, although the clinical relevance of these differences remains unknown at this time.
The present study found that the front windshields of the automobiles tested provided a high and consistent level of UV-A protection. In contrast, the side windows of the automobiles tested provided variable levels of UV-A protection, which may expose drivers’ left eyes and left sides of faces to greater cumulative UV-A light. This exposure may increase the risk of cataract and skin cancer. It remains unknown whether this risk can be modified for each individual based on vehicle profile (whether or not sunglasses are worn) or other factors, such as time spent driving in a certain environment.
The present study may have a relevance beyond that of the risk of UV damage from driving automobiles. It brings awareness of the other types of windows that make people vulnerable to the effects of the sun. These findings may also provide support for the recommendation that residential, commercial, and school glass windows have UV protection for adults and children alike.11
Submitted for Publication: June 13, 2015; final revision received March 24, 2016; accepted March 25, 2016.
Corresponding Author: Brian S. Boxer Wachler, MD, Boxer Wachler Vision Institute, 465 N Roxbury Dr, Ste 902, Beverly Hills, CA 90210 (firstname.lastname@example.org).
Published Online: May 12, 2016. doi:10.1001/jamaophthalmol.2016.1139.
Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Additional Contributions: I would like to acknowledge the contributions of Micaela Boxer Wachler, who assisted with data collection for this study.