Customize your JAMA Network experience by selecting one or more topics from the list below.
Loprinzi PD. Association of Accelerometer-Assessed Sedentary Behavior With Diabetic Retinopathy in the United States. JAMA Ophthalmol. 2016;134(10):1197–1198. doi:10.1001/jamaophthalmol.2016.2400
Physical activity (PA) is inversely associated with diabetic retinopathy (DR) prevalence among adults in the United States.1,2 Sedentary behavior (SB), independent of PA, is also detrimentally associated with risk factors (eg, hemoglobin A1c [HbA1c] level)3 linked with DR incidence.4 To my knowledge, no study has evaluated the association of SB with DR, which was this study’s purpose.
The study design, methods, and covariates used in this cohort study are identical to those of a previous study,2 with the exception that the current study evaluates the association between SB and DR, independent of PA. Data from the 2005 to 2006 National Health and Nutrition Examination Survey were used. All study procedures were approved by the National Center for Health Statistics review board, and written informed consent was obtained from all participants.
Diabetes was defined by self-report of a previous diabetes diagnosis, use of insulin or diabetes medications, HbA1c level of 6.5% or higher (to convert to proportion of total hemoglobin, multiply by 0.01), or a fasting glucose level of 126 mg/dL or higher (to convert to millimoles per liter, multiply by 0.0555).
Retinal imaging was performed using the Canon Nonmydriatic Retinal Camera CR6-45NM. The presence of nonproliferative DR (none, mild, or moderate to severe) was determined using the Early Treatment Diabetic Retinopathy Study grading criteria.5
Sedentary behavior was measured during waking hours in participants with 4 or more days of at least 10 h/d of accelerometer wear time using the Actigraph 7164 accelerometer. Activity counts of less than 100/min were used to define SB; activity counts of 100/min or greater were classified as total PA. Both SB and total PA are expressed as average minutes per day.
Statistical analyses were performed with Stata version 12 statistical software (StataCorp LP) using survey data procedures. Multivariate logistic regression was used to examine the association between SB (independent variable) and presence of nonproliferative DR (no DR as reference group). Analyses were adjusted for accelerometer-assessed total PA, age, sex, race/ethnicity, comorbid illness, smoking status, visual acuity, mean arterial pressure, serum cholesterol level, HbA1c level, homocysteine level, and functional disability.2 There was no evidence of multicollinearity (highest variance inflation factor, 1.86).
Characteristics of the 282 analyzed participants with diabetes can be found in a previous article.2 Briefly, the mean (SE) age was 62.2 years, 29.2% had mild or worse DR, and participants engaged in a mean (SE) of 522 (8.8) min/d of SB. After adjustment for variables described in the Methods, for a 60-min/d increase in SB, participants had 16% increased odds of having mild or worse DR (odds ratio [OR] = 1.16; 95% CI, 1.00-1.35; P = .04); total PA was not associated with DR (OR = 1.00; 95% CI, 0.99-1.01; P = .36). When both SB and total PA were categorized into tertiles, results were not significant when associating with DR (sedentary tertile 3 [vs 1], OR = 1.09; 95% CI, 0.45-2.64; P = .83; and total PA tertile 3 [vs 1], OR = 0.81; 95% CI, 0.28-2.35; P = .68).
When adding duration of diabetes (available only among those with physician-diagnosed diabetes [n = 217]) as a covariate to this adjusted model, SB (when expressed as a continuous variable) remained associated with DR (OR = 1.29; 95% CI, 1.04-1.60; P = .02). Also, there was no evidence of a multiplicative interaction effect of SB and PA on DR (OR = 1.00; 95% CI, 0.99-1.01; P = .89) or of interaction effects of SB and sex (OR = 1.18; 95% CI, 0.86-1.64; P = .26), SB and age (OR = 1.01; 95% CI, 0.99-1.04; P = .19), or SB and HbA1c level (OR = 1.04; 95% CI, 0.91-1.20; P = .48) on DR.
The plausibility of this positive association between SB and DR may in part be a result of the increased cardiovascular disease risks associated with SB,3 which in turn may increase the risk of DR.4 This association does not prove a cause and effect of SB and increased chance of worsening DR. To know whether this observed association had a cause-and-effect relationship, interventional trials would be needed in which individuals were assigned randomly to increase PA and decreased prolonged SB had a decreased chance of worsening DR.
Corresponding Author: Paul D. Loprinzi, PhD, Jackson Heart Study Vanguard Center of Oxford, Physical Activity Epidemiology Laboratory, School of Applied Sciences, Department of Health, Exercise Science, and Recreation Management, University of Mississippi, 229 Turner Center, University, MS 38677 (email@example.com).
Published Online: August 4, 2016. doi:10.1001/jamaophthalmol.2016.2400.
Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.