Abstract
The cognitive problems are prominent in the context of global aging, and the traditional Mendelian randomization method is not applicable to ordered multi-categorical exposures. Therefore, we aimed to address this issue through the development of a method and to investigate the causal inference of cognitive-related lifestyle factors. The study sample was derived from the Chinese Longitudinal Healthy Longevity Survey, which included 897 older adults aged 65 + . This study used genome-wide association analysis to screen genetic loci as instrumental variables and innovatively combined maximum likelihood estimation to infer causal associations between ordered multi-categorical exposures (diet, exercise, etc.) and continuous outcomes (cognitive level). The causal inference method for ordered multi-categorical exposures developed in this study was simple, easy to implement, and able to effectively and reliably discover the potential causal associations between variables. Through this method, we found a potential positive causal association between exercise status and cognitive level in Chinese older adults ( β^ = 1.883, 95%CI 0.182-3.512), in which there was no horizontal pleiotropy (p = 0.370). The study provided a causal inference method applicable to ordered multi-categorical exposures, that addressed the limitations of the traditional Mendelian randomization method.