Risk Factors for Elder Abuse Severity

Background and Purpose: Approximately 1 in 10 older adults experience elder abuse (EA) each year. EA is independently associated with significant morbidity and premature mortality and represents a major challenge facing a population with increased longevity. To date, EA risk factor research has focused on predicting dichotomous (yes/no) incidence outcomes. However, substantial variation exists in the extent and frequency of EA; mistreatment tends to fall along a continuum of severity. Older adults enduring more frequent and varied forms of mistreatment are more likely to experience poor health, hospitalization, and all-cause mortality. The National Research Council recommended that, in addition to understanding incidence risk factors in the general population, identifying factors associated with problem severity among victims is a research priority. This paper sought to identify factors, from several ecological levels, associated with the severity of elder emotional abuse, physical abuse and neglect among mistreated elders. We hypothesized that older adults with higher levels of physical vulnerability, victim-perpetrator relationship imbalance, cohabitation, and socio-cultural disadvantage would experience more severe EA.

Methods: Data came from the New York State Elder Abuse Prevalence Study (NYSEAPS). The NYSEAPS used a random digit-dial sampling strategy to conduct direct telephone interviews with a representative sample (n = 4156) of NYS older adults. Eligible participants were aged 60 years or above, living in the community, cognitively intact, and English/Spanish-speaking. Elder emotional and physical abuse were assessed with mistreatment items from the Conflict Tactics Scale (CTS). Elder neglect was assessed using the Duke Older Americans Resources and Services (OARS) scale. EA severity was measured as the summation of past-year mistreatment behavior frequencies across endorsed CTS and DUKE OARS items and converted into an ordinal variable (e.g., low/medium/high severity). Multivariate ordinal regression was used to model the severity of each EA type.

Results: Increasingly severe emotional abuse was predicted by younger age, functional impairment, shared living, Hispanic ethnicity, a spousal/partner perpetrator, and living alone with the perpetrator. Higher levels of physical abuse severity were associated with younger age, a grandchild perpetrator, and living alone with the perpetrator. Increasingly severe neglect was predicted by younger age, functional impairment, and low household income.

Conclusions and Implications: This paper extends existing EA risk factor knowledge as the first known study to examine factors predicting EA severity. A focus on severity begins to frame EA phenomena as they exist in reality along a spectrum. Elder protective service agencies respond to tens of thousands of EA cases each year with limited research guidance to inform best practices. Clinical decisions related to triaging incoming referrals, formulating safety plan interventions, and/or closing cases depend upon an understanding of EA severity. EA severity risk factor findings from this paper carry direct implications for the development of targeted interventions to alleviate the magnitude of existing EA cases and protect victims from heightened risks of mortality and morbidity. Different risk profiles underlying the severity of elder emotional abuse, physical abuse and neglect, respectively, indicate a need for EA-type-specific response protocols and intervention approaches.