Purpose: To determine the prevalence of cataract and its association with sun exposure and other environmental risk factors in three different geographically diverse populations of India.

Design: Population based cross sectional study during 2010-2016

Participants: People aged > 40 years residing in randomly sampled villages were enumerated (12021) and 9735 (81%) underwent ophthalmic evaluation from plains, hilly and coastal regions (3595, 3231, 2909 respectively)

Methods: A detailed questionnaire-based interview about outdoor activity in present, past and remote past, usage of sun protective measures, exposure to smoke, and detailed ophthalmic examination including assessment of uncorrected and best corrected visual acuity, measurement of intraocular pressure, slit lamp examination, lens opacities categorization using LOCS III and posterior segment evaluation was done. Lifetime effective sun exposure was calculated using Melbourne formula and expressed as quantiles. These were supplemented with physical environmental measurements.

Main outcome measures: Lifetime sun exposure hours, smoking, indoor kitchen smoke exposure and their association with cataract and subtypes. Prevalence of cataract calculated based on lens opacities or evidence of cataract surgery.

Results: Cataract was identified in 3231 (33.3%) participants. Prevalence of cataract in males (32.3%) and females (34.1%) was similar. Nuclear cataract was the commonest sub-type identified in 94.7% of affected eyes. Sun exposure had a significant association with cataract with odds ratio (OR) increasing from 1.6 (95% Confidence Intervals [CI]: 1.4, 1.9) in 3^rd quantile, to 2.6 (CI: 2.2, 3.1) in 4^th quantile and 9.4 (CI: 7.9, 11.2) in 5^th quantile (p<0.0001). Cataract also showed a significant association with smoking (OR: 1.4, CI: 1.2, 1.6) and indoor kitchen smoke exposure (OR: 1.2, CI: 1.0-1.4). Nuclear cataract showed a positive association with increasing sun exposure in 3^rd (β coefficient 0.5, CI:0.2-0.7), 4^th (β: 0.9, CI: 0.7-1.1) and 5^th (β: 2.1, CI:1.8-2.4) quantiles of sun exposure, smoking (β: 0.4, CI: 0.2-0.6) and indoor kitchen smoke exposure (β: 0.3, CI: 01-0.5) while cortical cataract showed a positive association with sun exposure only in 5^th quantile (β: 2.6, CI:1.0-4.2). Posterior subcapsular cataract was not associated with any of the risk factors.

Conclusion: Cataract is associated with increasing level of sun exposure, smoking and exposure to indoor kitchen smoke.

All analysis was doen using Stata Version 13

Participants with incomplete information on sun exposure or ocular examination were excluded.

All study participants were distributed across quintiles, based lifetime effective sun exposure.

The prevalence of cataract was assessed among persons whose clinical lens evaluation was complete.

Prevalence of specific subtype of cataract was calculated among participants whose LOCS III evaluation in either eye was done.

Participants with traumatic or developmental cataracts were excluded.

The combination patterns of subtypes of cataracts in the LOCS III assessed eyes were tabulated.

In case one eye had cataract while other eye was pseudophakic or apahakic, we classified the participant based on the cataractous eye.

Multivariable logistic regression analysis was done to estimate the association of sun exposure, smoking, and gender and other risk factors with any cataract.

Analysis were repeated for each site and for pooled data.

Patients with pure cortical, nuclear and PSC based on LOCS III score were identified and the association with sun exposure and other risk factors was evaluated using multinomial logistic regression analysis, keeping person with both eyes normal as controls.

The main person-wise data file  is "UVR-cataract-analysis.dta" and to reproduce the analysis, use the "UVR-cataract-analysis.do" file in Stata version 13. The "UVR-cataract-analysis - Codebook.pdf" file lists the variables in the dataset. Its also describes the number of missing values in the dataset for various variables. 

The "LOCSeyePattern.dta" file is a longform data of each eye, where eye = 1 means right eye and eye=2 means left eye. The "LOCSeyePattern - Codebook.pdf" describes the variables used in this dataset.

The "UVR-cataract-analysis.do" file runs the analysis and saves the results as "Results.pdf" and "date_combinedUVR40+_6.smcl" files. As an intermediate step, "combinedUVR40+_7.dta" and "UVR-cataract-analysis_7.dta" files are generaed and these may be ignored.

The Key Variables are - 

1. unique_id - person unique ID
2. anyCataract - Person with any eye having cataract - prevalence of any cataract
3. cort_prev nuc_prev psc_prev - prevalence of any specific type of cataract
4. cort_assoc nuc_assoc psc_assoc pure_assoc - Distribution of pure sub-types of cataracts
5. age - Age of participant - continuous
6. gender - Gender
7. educat2 - Educational status
8. workTypePrCat - Occupation (Present
9. smoke20 - Smoking
10. indoor16_kitchen - Indoor Smoke
11. tExposure - Lifetime cumulative effective sun exposure - continuous
12. tExposureQuantilesAll  - Lifetime cumulative effective sun exposure expressed as quintiles
13. study_location 
14. sugar200 - Diabetes
15. pvaBe - Vision Status based on presenting visual acuity in better eye
16. bvaBe - Vision Status based on best corrected visual acuity in better eye
17. beOp - Both eyes operated for cataract
18. oneEyeOp - One Eye operated for cataract
19. locsRe*, locsLe* - LOCS III grading results of eirther eye for cortical opacity, nuclear opacity, nuclear opalescence and PSC