Certain confounders were not measured, most notably contraindications to aspirin (eg, genetic platelet abnormalities). Such findings could explain some patterns of aspirin use in both strata, as up to 10% of any given population has a contraindication to aspirin due to allergy, intolerance, gastrointestinal ulcer, concomitant anticoagulant medication, or other high bleeding risk.18,31 Few of these variables were known about our sample.
TABLE 4W (available below) provides a breakdown of some possible aspirin contraindications, as well as possible reasons other than primary CVD prevention for regular aspirin use. Because clinical judgment is often required to assess the degree of severity of a given health condition in order to deem it an aspirin contraindication, these findings could not reliably be used to reclassify participants. We present them simply for hypothesis generation.
Some data collection predates the current USPSTF guidelines,10 which could have resulted in a misclassification of participants’ aspirin indication. However, sensitivity analyses restricted to the 2010 sample alone—the only one with data collection after the newer guidelines were released—did not reveal any meaningful differences.
Other methodological limitations include the less racially diverse population of Wisconsin compared with other parts of the country and the sample size, which did not permit testing for statistical interactions and perhaps resulted in larger confidence intervals for some associations (eg, race/ethnicity) relative to the population as a whole.
TABLE 4W
Possible reasons for aspirin use—or contraindication— by aspirin indication*
| Has a doctor or other health professional ever told you that you had … | Aspirin indicated (n=268) | Aspirin not indicated (n=563) | ||
|---|---|---|---|---|
| Regular aspirin user (n=83) | Nonregular aspirin user (n=185) | Regular aspirin user (n=102) | Nonregular aspirin user (n=461) | |
| Migraine headache Yes No | 20 (24%) 63 (76%) | 28 (15%) 157 (85%) | 24 (24%) 78 (76%) | 76 (16%) 385 (84%) |
| Arthritis† Yes No | 2 (2%) 81 (98%) | 1 (1%) 184 (99%) | 12 (12%) 90 (88%) | 26 (6%) 435 (94%) |
| Stomach or intestinal ulcer Yes No | 5 (6%) 78 (94%) | 6 (3%) 179 (97%) | 7 (7%) 95 (93%) | 10 (2%) 451 (98%) |
| Reflux or GERD Yes No | 8 (10%) 75 (90%) | 14 (8%) 171 (92%) | 11 (11%) 91 (89%) | 32 (7%) 429 (93%) |
| Values presented as n (%). *Data not included in study analysis. †Osteoarthritis or rheumatoid arthritis. GERD, gastric esophageal reflux disease. | ||||
Acknowledgement
The authors thank Matt Walsh, PhD, for his assistance in creating the analytical dataset, as well as Sally Steward-Townsend, Susan Wright, Bri Deyo, Bethany Varley, and the rest of the Survey of the Health of Wisconsin staff.
CORRESPONDENCE Jeffrey J. VanWormer, PhD, Epidemiology Research Center, Marshfield Clinic Research Foundation, 1000 North Oak Avenue, Marshfield, WI 54449; vanwormer.jeffrey@mcrf.mfldclin.edu