Department of Health Policy and Management, Texas A&M University School of Public Health, College Station.
Department of Medicine, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, New York.
JAMA Intern Med. 2024 Nov 1;184(11):1324-1328. doi: 10.1001/jamainternmed.2024.4998.
Historic redlining, the practice by the Home Owners' Loan Corporation (HOLC) of systematically denying credit to borrowers in neighborhoods that were inhabited by primarily African American individuals, has been associated with poor community outcomes, but the association with individual risk of death is not clear.
To examine if exposure to residential redlining practices by HOLC in 1940 is associated with increased risk of death later in life.
DESIGN, SETTING, AND PARTICIPANTS: The study linked individuals who resided within HOLC-graded neighborhoods (defined as Census Enumeration Districts) in 1940 with administrative death records data. The study estimated hazard ratios as well as age-specific life expectancy gaps (at age 55, 65, and 75 years) for HOLC grading exposure. This was done using methods that adapted standard parametric survival analysis to data with limited mortality coverage windows and incomplete observations of survivors. The analysis sample consisted of 961 719 individual-level observations across 13 912 enumeration districts within 30 of the largest US cities (based on 1940 population counts) across 23 states. Data were analyzed between December 1, 2023, and September 4, 2024.
The exposure was HOLC grade based on historic HOLC maps, with A representing "best" or creditworthy areas; B, "still desirable"; C, "definitely declining"; and D, "hazardous" areas not worthy of credit (ie, redlined), and the main outcome was age at death from the Social Security Numident file.
The 961 719-person individual sample had a mean (SD) age of 19.26 (9.26) years in 1940 and a mean (SD) age at death of 76.83 (9.22) years. In a model adjusted for sex (52.48% female; 47.52% male), race and ethnicity (7.36% African American; 92.64% White), and latent place effects, a 1-unit lower HOLC grade was associated with an 8% (hazard ratio, 1.08 [95% CI, 1.07-1.09]) increased risk of death. At age 65 years, these hazard differentials translated into an estimated life expectancy gap of -0.49 (95% CI, -0.56 to -0.43) years for each 1-unit decrease of the HOLC grade.
This study found that individuals who resided within redlined neighborhoods in 1940 had lower life expectancy later in life than individuals who resided within other HOLC-graded areas.
历史上的红线政策（由房主贷款公司（HOLC）实施的系统地拒绝为主要由非裔美国人居住的社区的借款人提供信贷的做法）与较差的社区结果有关，但与个体死亡风险的关联尚不清楚。
研究 1940 年 HOLC 的住宅红线政策暴露是否与以后的死亡风险增加有关。
设计、地点和参与者：该研究将居住在 1940 年 HOLC 评级社区（定义为人口普查区）内的个人与行政死亡记录数据相关联。该研究使用了一种方法，该方法将标准参数生存分析适应于具有有限死亡率覆盖窗口和幸存者观察不完全的数据。分析样本包括来自 23 个州的 30 个美国最大城市（基于 1940 年人口数）的 13912 个普查区中的 961719 个个体水平观察值。数据于 2023 年 12 月 1 日至 2024 年 9 月 4 日之间进行分析。
暴露是基于历史 HOLC 地图的 HOLC 等级，A 代表“最佳”或有信用的区域；B，“仍然理想”；C，“明显下降”；D，“危险”不值得信贷（即红线）区域，主要结果是来自社会安全号码文件的死亡年龄。
961719 人的个体样本在 1940 年的平均（SD）年龄为 19.26（9.26）岁，平均（SD）死亡年龄为 76.83（9.22）岁。在调整了性别（52.48%为女性；47.52%为男性）、种族和民族（7.36%为非裔美国人；92.64%为白人）和潜在地点效应的模型中，HOLC 等级每降低 1 个单位，死亡风险就会增加 8%（风险比，1.08 [95%CI，1.07-1.09]）。在 65 岁时，这些危险差异转化为 HOLC 等级每降低 1 个单位，预期寿命差距估计为-0.49（95%CI，-0.56 至-0.43）年。
本研究发现，1940 年居住在红线社区内的个体与居住在其他 HOLC 评级区域内的个体相比，晚年的预期寿命较低。
