Table. Characteristics of the Study Population at Baselinea.
| Study period | Patients with MS | Controls | P value | |||
|---|---|---|---|---|---|---|
| 1987-2018 | 2002-2018 | 1987-2018 | 2002-2018 | 1987-2018 | 2002-2018 | |
| No. | 12 251 | 7957 | 72 572 | 47 175 | ||
| Follow-up time, mean (SD), y | 10.3 (6.3) | 8.3 (4.6) | 11.5 (6.5) | 9.1 (4.6) | <.001 | <.001 |
| Female sex, % | 69.9 | 70.3 | 69.8 | 70.1 | .75 | .95 |
| Age, mean (SD), y | 44.9 (13.3) | 44.8 (13.6) | 44.9 (13.3) | 44.9 (13.6) | .73 | .75 |
| White ethnicity, % | 93.9 | 93.6 | 91.2 | 91.0 | <.001 | <.001 |
| Smoking status, % | ||||||
| Nonsmoker | 47.1 | 46.4 | 58.1 | 60.1 | <.001 | <.001 |
| Ex-smoker | 15.0 | 16.9 | 12.5 | 13.1 | ||
| Current | 37.9 | 36.7 | 29.4 | 26.8 | ||
| Diabetes, % | 7.2 | 6.8 | 5.0 | 4.6 | <.001 | <.001 |
| Depression, % | 20.9 | 23.0 | 8.7 | 8.8 | <.001 | <.001 |
| Medication, % | ||||||
| Antidiabetic | 1.2 | 1.4 | 1.2 | 1.4 | .97 | .79 |
| Antihypertensive | 5.6 | 5.6 | 5.7 | 6.0 | .97 | .13 |
| Lipid lowering | 2.7 | 3.7 | 2.9 | 3.9 | .35 | .44 |
| Antiplatelet | 2.7 | 3.1 | 2.0 | 2.3 | <.001 | <.001 |
| Anticoagulant | 0.5 | 0.5 | 0.5 | 0.5 | .43 | >.99 |
| No. of primary care visits in previous year, mean (SD) | 7.8 (11.2) | 8.8 (12.5) | 2.9 (5.9) | 3.2 (6.5) | <.001 | <.001 |
| Index of multiple deprivation, % | ||||||
| 1st Quintile (least deprived) | 14.4 | 14.6 | 14.4 | 14.6 | >.99 | >.99 |
| 2nd Quintile | 18.5 | 19.4 | 18.5 | 19.3 | ||
| 3rd Quintile | 17.8 | 18.2 | 17.8 | 18.2 | ||
| 4th Quintile | 19.2 | 19.0 | 19.2 | 19.0 | ||
| 5th Quintile (most deprived) | 20.4 | 20.0 | 20.4 | 20.1 | ||
| Missing data | 9.7 | 8.9 | 9.7 | 8.9 | ||
Abbreviation: MS, multiple sclerosis.
a
Study baseline was defined as the year of MS diagnosis (index year). For matched controls, a matched index year was assigned. All patients with MS and matched controls from 1987 to 2018 were included in the main analysis, while a sensitivity analysis restricted the sample to only diagnoses from 2002 to 2018. Univariate baseline differences in study variables between patients with MS and controls were assessed using χ2 tests, t tests, and Kruskal-Wallis tests, as appropriate.