Supplementary Material for: Daily Habit of Water Intake in Patients with Cerebral Infarction before its Onset; Comparison with a Healthy Population: A Cross-Sectional Study

Background: While water intake is frequently recommended to prevent cerebral infarction (CI), only few studies have been published on this topic. Objectives: This study retrospectively estimated the daily non-alcohol drink (NAD) intake in CI patients before CI onset and compared it with NAD in healthy subjects. Methods: We performed a cross-sectional study on CI patients in 3 hospitals and healthy subjects in the Kobe Orthopedic and Biomedical Epidemiologic (KOBE) study. Data from 1,287 subjects (274 CI patients and 1,013 healthy subjects) were used for the analyses. By dividing the CI patients into “increased”, “unchanged”, and “decreased” groups according to their current NAD intake, we compared the NAD intake between these 3 groups and healthy subjects by analyses of covariance and the post hoc test, adjusting for sex, age, surveyed month, body mass index, alcohol drinking history, and smoking history. Under the assumption that the NAD intake in the “unchanged” group was equal to the NAD intake before CI onset, the OR of less NAD intake for CI adjusting for the relevant variables in the “unchanged” group and the healthy subjects was calculated; the cut-off point was chosen using Youden’s index. Results: The mean age (mean ± SD) of the participants was 62.8 ± 9.3 years. One hundred and fifty-one patients (36 women) were included in the “increased” group; 105 (30 women), in the “unchanged” group; 18 (2 women), in the “decreased” group; and 1,013 (706 women), in the “healthy” group. The mean NAD intake was 1,702.5 ± 670.2 mL in the “increased” group, 1,494.2 ± 611.2 mL in the “unchanged” group, 1,268.0 ± 596.1 mL in the “decreased” group, and 1,720.6 ± 686.0 mL in the “healthy” group. After adjusting for the relevant variables, a significant difference in NAD intake between the groups was observed (F = 6.1, p < 0.001), and a post hoc test demonstrated significant differences (p < 0.05) in NAD intake between the “healthy” and “unchanged” groups, and between the “increased” and “unchanged” groups. The OR of less NAD intake (<1,570 mL/day, chosen using Youden’s index) for CI was 2.48 (95% CI 1.52–4.07). Conclusion: This study showed that daily NAD intake before CI onset in CI patients was less than that in healthy persons, indicating that sufficient intake of NAD may be protective for CI.