Example: Confounding Bias - Night Lite

Lead Author(s): Jeff Martin, MD

Example of Confounding: Night Lite

Confounding can be found in Quinn's study of myopia and ambient lightning.

Quinn studied children, who were exposed to ambient night lightning before two years of age.

In the figure below you can see that according to this study these children, tested at 2 - 16 years age,

0330_2night.JPG

2x2 Table of Study Data

Looking at the data in our standard notation of a 2x2 table below we see a prevalence of 3.4. 0326_night2.JPG

Responses to Study

This study prompted responses from two other investigators: Zadnik and Gwiazda, who showed that was no association in their studies and attributed the original finding to confounding.

As seen in the figure below Zadnik and Gwiazda believed that the

0330_1myopia.JPG
These authors provided direct data that nearsightedness in parents is associated with the use of night lights,

Confounder: Parents' Myopia

So, this is an example, where parenteral myopia confounds the relationship between night lights and child's myopia and there is no real direct, or what we call independent, association between use of night lights and child's myopia.

References

Gwiazda, J., Ong, E., Held, R., & Thorn, F. (2000). Myopia and ambient night-time lighting. Nature, 404(6774), 144

Quinn, G. E., Shin, C. H., Maguire, M. G., & Stone, R. A. (1999). Myopia and ambient lighting at night. Nature, 399(6732), 113-114.

Zadnik, K., Jones, L. A., Irvin, B. C., Kleinstein, R. N., Manny, R. E., Shin, J. A., et al. (2000). Myopia and ambient night-time lighting. CLEERE Study Group. Collaborative Longitudinal Evaluation of Ethnicity and Refractive Error. Nature, 404(6774), 143-144.