Department of Sociology, Florida State University , Tallahassee , FL , USA.
Department of Psychology, State University of New York , Fredonia , NY , USA.
Aging Ment Health. 2019 Nov;23(11):1452-1466. doi: 10.1080/13607863.2018.1503999. Epub 2018 Oct 31.
Research on the influence of companion animals (CA) on the health of older adults has yielded contradictory results. Selection factors, leading to heterogeneity both between and within groups of CA owners and non-owners, likely bias results. We conduct analyses to identify typologies of owners and non-owners. Using data on older adults (60+) from the 2012 Health and Retirement Study (HRS), and the HRS companion animal module, (owners = 478) and (non-owners = 624), we conducted latent class analyses (LCA). We used key demographic, health, daily engagement, and pet characteristic variables to complete our analyses. Analyses revealed five clusters of CA owners and four clusters of non-owners. Health and CA related characteristics distinguishing clusters suggest important sources of variability and reflect qualitatively different profiles of owners and non-owners. We also found CA owners were more likely than non-owners to be high on neuroticism and to be less extroverted than non-owners-but again there was considerable within group variability. Factors that select people into pet ownership not only work individually to characterize ownership, they create distinct typologies of CA owners and non-owners that likely contribute to subsequent health outcomes. In order to determine if having a CA is beneficial to health in later life and for whom, future research should consider selection factors like those identified in the typologies. Statistical analyses, such as LCA, that can adequately account for these selection factors is necessary to avoid biases in the interpretation of results.
伴侣动物（CA）对老年人健康的影响研究得出了相互矛盾的结果。选择因素，导致 CA 所有者和非所有者群体之间以及群体内部存在异质性，可能会产生偏差。我们进行了分析，以确定所有者和非所有者的类型。
使用来自 2012 年健康与退休研究（HRS）和 HRS 伴侣动物模块的数据（所有者=478 人，非所有者=624 人），我们进行了潜在类别分析（LCA）。我们使用了关键的人口统计学、健康、日常活动和宠物特征变量来完成我们的分析。
分析结果揭示了五个 CA 所有者群体和四个非所有者群体。区分群体的健康和 CA 相关特征表明存在重要的可变性来源，并反映了所有者和非所有者的定性不同特征。我们还发现，与非所有者相比，CA 所有者更有可能神经质较高，且比非所有者更内向，但同样存在相当大的群体内变异性。
选择人们养宠物的因素不仅单独作用于特征描述，还会创建不同类型的 CA 所有者和非所有者，这可能会对后续的健康结果产生影响。为了确定在晚年养宠物是否对健康有益，以及对谁有益，未来的研究应该考虑到这些类型学中确定的选择因素。能够充分考虑这些选择因素的统计分析，如潜在类别分析，对于避免结果解释中的偏差是必要的。
