Skip to main content
. 2022 Jul 26;45(8):1772–1778. doi: 10.2337/dc21-2563

Table 1.

Summary statistics for main outcome variables (per 100,000) by year, HOLC grade grouping, and death cause

HOLC grades Diabetes All cause
Mortality YLL Mortality YLL
Year 1990
 A or B 16.41 (4.08) 300.92 (74.61) 914.34 (113.38) 16,048.45 (2,330.08)
 C or D 26.78 (10.47) 523.75 (214.45) 1,141.19 (222.61) 23,019.26 (6,387.03)
Year 2000
 A or B 17.11 (4.4) 294.78 (76.7) 751.07 (96.00) 12,262.23 (1,795.94)
 C or D 28.94 (10.27) 531.95 (199.43) 935.09 (158.61) 17,550.35 (4,119.66)
Year 2010
 A or B 12.7 (3.56) 221.51 (63) 577.43 (79.16) 9,556.78 (1,521.81)
 C or D 20.91 (7.49) 397.78 (149.98) 721.5 (142.1) 13,640.63 (3,258.12)
Year 2014
 A or B 12.68 (3.87) 217.57 (66.79) 547.17 (75.25) 9,033.6 (1,467.63)
 C or D 21.47 (8.51) 403.39 (168.13) 695.53 (148.24) 13,089.29 (3,329.08)

Data are means (SD). Two-way t tests indicate that mean differences across A/B and C/D HOLC grade areas are significant at P < 0.001 level across all years and causes. Mean difference tests performed with use of spatial autoregressive models that allow for spatial correlation across the outcomes and HOLC grade score measures also resulted in significant mean differences (P < 0.001 for all). (Details on these estimates can be found in Supplementary Appendix B2.) Total number of observations is 109 census tracts; of these, 53 are contained in the A or B HOLC grade group and 56 in the C or D HOLC grade group.