In this study, MCR and slow gait were both significantly associated with frailty in older adults in geriatric department, but not subjective cognitive complaints. These associations remained unchanged after adjusting for the potentially confounding variables. These results revealed the association between MCR and frailty was significantly observed with slow gait, and the combination of subjective cognitive complaints and slow gait may have more predictive value than the single component in order to explain their common mechanisms with frailty.
The primary finding of this study is that MCR was associated with frailty in elder population. This finding is in accordance with a cross-sectional study from France , which showed that physical frailty and in particular slow gait speed were associated with cognitive impairment, indirectly reflecting slow gait and cognitive impairment often coexist with MCR, and suggesting that MCR and frailty interact with each other with aging. Subjective cognitive complaints and slow gait are common condition in aging process, and have been considered as early clinical indicators of cognitive impairment and dementia during the preclinical stages. Cognitive impairment is a longitudinal process that begins with minor alterations in certain domains and slowly progresses to multiple changes in many domains . The accumulation of cerebral amyloid deposition and neurofibrillary tangles plays a critical role in progression of Alzheimer’s disease neuropathology [43,44]. In addition, brain regions alterations including motor cortices, striatum and substantia nigra are often involved [43,44]. Studies have shown that MCR had a wider involvement of brain regions not limited to the hippocampus, which is a key brain region for memorization and spatial navigation processes [45–47]. It was characterized by smaller volumes of total gray matter, total cortical gray matter, premotor cortex, prefrontal cortex especially dorsolateral segment, but no significant differences were obtained in terms of the volumes of hippocampal and white matter [48,49], which are in concordance with these results as the two components of MCR controlling cognitive complaints and slow gait. MCR related brain regions reduction contributes to predict cortical neurodegenerative dementia more than subcortical dementia such as vascular dementia . A study by Wang and colleagues found that frontal lacunar infarcts was associated with slow gait, poor memory function and MCR, but not with cognitive complaints , revealing that this brain area might result in MCR by disrupting frontally based neural networks facilitating memory and gait functions [51,52]. On the other hand, the accumulation of common brain pathologies and related brain areas alterations may also contribute to progressive frailty [53,54]. Cerebral amyloid-beta deposition and neurofibrillary tangles were associated with the symptoms of frailty including slower gait speed and lower body mass index in dementia free elders [55,56]. Thus, the association between MCR and frailty could be explained by the aboved mechanisms. Slow gait as a common manifestation of frailty in the elderly, which is why slow gait is linked with frailty in our samples. However, the lack of longitudinal prospective studies about the casual associations between MCR and frailty prevents from drawing conclusions.
Association between subjective cognitive complaints and frailty failed to observe in this study sample. It was inconsistent with previous studies such as the Hellenic Longitudinal Investigation of Aging and Diet study  and the Healthy Aging Longitudinal Study in Taiwan  revealing that subjective cognitive decline was associated with increased likelihood for frailty in cognitively unimpaired elderly individuals. Self-reported subjective cognitive decline complaints were adopted, but the discrepancy results were obtained due to the differences in frailty assessment criteria to quantify frailty status. In addition, the characteristics of the enrolled study sample were also inconsistent. The potential neurobiological explanations for the association between subjective cognitive decline and frailty lies in that both two factors increased risk of future cognitive decline [16,17,29,30] and Alzheimer’s disease pathology [57–60].
An important advantage of this study lies in all the enrolled older participants who underwent a comprehensive geriatric assessment focusing on the functional status of the elderly, such as daily activities of life, cognitive and frailty. Indeed, in our geriatric department, we have a trained team for comprehensive geriatric assessment, and all the assessments are designed to be a professional software system. Thus, the validity and consistency of the evaluations are guaranteed. Another advantage is that MCR and CFS-defined frailty screening in this study are relatively simple and easy to be implemented in other clinical settings, such as community health service centers and other primary medical centers. CFS was reported to reflect the dynamic and reversible changes of frailty, and well associate with frailty index. Thus, it can similarly predict adverse outcomes for elder patients [35,61]. On the contrary, this study also has some limitations. First, although this study is first to report the association between MCR and frailty in Chinese older population, the cross-sectional study precluded us from elucidating the causal relationships. Second, due to a convenience sample recruited inpatients in a monocentric study and the relatively small sample size of MCR cases, there are a selection bias and a representativeness bias in this study. Third, although these study participants were at normal MMSE range, some participants with mild hidden dementia might be undiagnosed. Finally, the standard definition of MCR is also a limitation in this study. The definition of slow gait is already standardized, but subjective cognitive complaints have different choices. Subjective cognitive complaints are defined as by one item of GDS–15 in this study according to the standard reported in most studies. Furthermore, the accuracy of self-reported cognitive complaints may be affected by the disease, emotional and environmental factors. Thus, the generalization of the study results could not be ignored because it is less sensitive for non-demented older adults.