Digital Health, IBM Research, Tokyo, Japan.
AI and Cognitive Software, IBM Research-Almaden, San Jose, CA, USA.
Aging Ment Health. 2023 Jun;27(6):1127-1134. doi: 10.1080/13607863.2022.2087207. Epub 2022 Jun 23.
To investigate whether latent subgroups with distinct patterns of factors associated with self-rated successful aging can be identified in community-dwelling adults, and how such patterns obtained from analysis of quantitative data are associated with lay perspectives on successful aging obtained from qualitative responses.
Cross-sectional data were collected from 1,510 community-dwelling Americans aged 21-99 years. Latent class regression was used to identify subgroups that explained the associations of self-rated successful aging with measures of physical, cognitive, and mental health as well as psychological measures related to resilience and wisdom. Natural language processing was used to extract important themes from qualitative responses to open-ended questions, including the participants' definitions of successful aging.
Two latent subgroups were identified, and their main difference was that the wisdom scale was positively associated with self-rated successful aging in only one subgroup. This subgroup had significantly lower self-rated successful aging and worse scores for all health and psychological measures. In the subgroup's qualitative responses, the theme of wisdom was only mentioned by 10.6%; this proportion was not statistically different from the other subgroup, for which the wisdom scale was not statistically associated with the self-rated successful aging.
Our results showed heterogeneous patterns in the factors underpinning successful aging even in community-dwelling adults. We found the existence of a latent subgroup with lower self-rated successful aging as well as worse health and psychological scores, and we suggest a potential role of wisdom in promoting successful aging for this subgroup, even though individuals may not explicitly recognize wisdom as important for successful aging.
探究在社区居住的成年人中,是否可以识别出与自我评定成功老龄化相关的因素存在明显模式的潜在亚组,以及从定量数据分析中获得的此类模式与从定性反应中获得的关于成功老龄化的通俗观点之间存在何种关联。
从年龄在 21-99 岁的 1510 名社区居住的美国人中收集了横断面数据。采用潜在类别回归来识别亚组,这些亚组解释了自我评定成功老龄化与身体、认知和心理健康测量以及与韧性和智慧相关的心理测量之间的关联。自然语言处理被用来从对开放式问题的定性反应中提取重要主题,包括参与者对成功老龄化的定义。
确定了两个潜在亚组,它们的主要区别在于智慧量表仅与其中一个亚组的自我评定成功老龄化呈正相关。该亚组的自我评定成功老龄化明显较低,所有健康和心理测量指标的得分也较差。在该亚组的定性反应中,只有 10.6%的人提到了智慧这一主题;这一比例与另一个亚组没有统计学差异,对于后者,智慧量表与自我评定的成功老龄化没有统计学关联。
我们的研究结果表明,即使在社区居住的成年人中,成功老龄化的基础因素也存在异质性模式。我们发现存在一个自我评定成功老龄化较低、健康和心理评分较差的潜在亚组,我们提出智慧在促进该亚组成功老龄化方面可能具有潜在作用,尽管个体可能没有明确将智慧视为成功老龄化的重要因素。