Longitudinal analysis of health outcomes after exposure to toxics, Willits California, 1991–2012: application of the cohort-period (cross-sequential) design

BACKGROUND About 1963, a factory in Willits, Mendocino County (County), California added chrome plating to the manufacture of steel products. After years of residents reporting high illness rates, the State undertook a series of investigations. They found exposures had been high and warranted further research into possible health effects. Applying the seldom-used cross-sequential design, we tested if Willits had an excess rate of adverse health conditions, compared to people of the same sex and cohort living in the rest of county (ROC). This is the first report on long-term health outcomes for Willits. METHODS Hospital discharge data for 1991-2012 were searched to find admissions for people born between 1940 and 1989 who ever gave the County as their residence. Diagnoses and procedures were classified to Level 1 (body systems) of the Multi-level Clinical Classification Software (CCS). We analyzed 796,917 diagnoses and 289,980 procedures found on 117,799 admissions of 43,234 patients who lived in the County at some time between 1991 and 2012. Of these, 7,564 lived in Willits. We summarized data to the person-level then the group level over cohort-period (cross-sequential) to control the age by time relationship, then calculated incidence rates, relative risk, and excess case statistics, each with confidence limits. A secondary analysis focused on whether Willits differed markedly from the rest of County (ROC). Specifically, other than the presence of the Plant, did Willits differ so much that those differences could plausibly explain outcome differences? RESULTS Illness was excessive in the exposed group (Willits) compared to the unexposed (ROC). Overall number of excess cases attributable to living in Willits was estimated: Men, 301 (95% confidence limit (CL) 200-398), women: 696 (CL 569-820). CONCLUSIONS This study demonstrates the strength of the cross-sequential design. Willits and ROC had comparable disadvantages relative to the State. Yet, when stratified by cohort, Willits had more illness per population. Little is known about the health effect of chemicals used at Willits on a non-occupationally exposed population. We recommend a follow-up study to evaluate the long-term health of people who lived in Willits during childhood and the reproductive age.