The results of the current study showed that 40.8% of the participants had a high FoF. The factors associated with a high FoF varied with the perceived level of social support. Among the participants with perceived low and moderate levels of social support, cognitive impairment was associated with a high FoF, whereas in the participants with a high perceived level of social support, this relationship of the FoF with cognitive function was not observed.
Prevalence of high FoF
This study conducted in a population-representative sample of 725 Vietnamese elderly living in rural and urban areas revealed a high prevalence of FoF. Overall, 40.8% of the elderly were identified as having a high FoF. This considerably high prevalence as compared to previous reports [9, 30, 31] might need to be interpreted from the perspective of the study location [18, 32], age of the participants (mean age: 72 ± 8.8) and standardized measurement of the FoF . While safe environment could have effects on the FoF [18, 32], another possible reason is Thua Thien Hue’s long rainy season, which normally lasts for 4 months, from September to December, with a large amount of rainfall , which leaves surfaces slippery and unsafe for the elderly to perform their daily activities. This finding suggests the high level of concern for falls among Vietnamese elderly living in the community. Therefore, attention should be paid to this population in terms of falls and fall-related issues.
Cognitive impairment and FoF among different perceived social support levels
The univariate analyses in the present study indicated that elderly persons with impaired cognition had a higher odd of experiencing a high FoF. With advancing age, attention and memory are the two aspects of cognition that are the most severely affected . Impairment of cognitive function in the elderly, characterized by a poorer performance in memory, executive functions, attention and information processing, has been reported as a predictor of the onset of FoF  and to also be associated with higher levels of anxiety , which, in turn, was shown to be associated with a FoF . Furthermore, in the logistic regression model with adjustments for confounding factors performed to determine the association between the participant’s characteristics and the likelihood of a high FoF (Additional file 1), elderly with cognitive impairment were approximately 2 times more likely to have a high FoF than those with normal cognition. This finding confirmed the effect of cognitive impairment on the high FoF among the study participants.
An even stronger association between cognition and FoF was observed among the elderly with low and moderate perceived levels of social support in the fully adjusted model: elderly with cognitive impairment in this group had a nearly 3 times higher likelihood of having high a FoF. This result can be explained by the observed characteristics of the study participants with two different perceived levels of social support, as well as by the two-way relationship between the perceived social support level and the fear of falling. A lower perceived social support level among the Vietnamese elderly might be one of the consequences of social modernization, which is currently spreading quickly in Vietnam; the younger generation tends to migrate to bigger cities and leave older persons on their own . Moreover, Vietnamese elderly are exposed to less social interactions due to the limited social activities held by the local authority as compared to other developed countries . Elderly with lower levels of social support are more likely to be physically inactive , depressed [36, 39, 40], which are reported to be closely related to FoF [18, 41]. In addition, we found that the mean MMSE score of the participants with low-moderate perceived social support levels was 25.5, which was significantly lower than that of those with a high perceived level of social support (26.6) (Additional file 2). This was also in line with a prior finding in an Asia setting of a positive relationship between better cognitive function and a higher social support level . Accordingly, an ageing-friendly environment, such as the availability of wide sidewalks, a proper rainwater drainage system, and establishment of more social activities for the elderly in the community may represent some solutions to prevent FoF, especially in the elderly with limited social support and impaired-cognition.
However, this association between cognitive decline and the FoF was no longer seen in participants with perceived high social support levels. Considering the close-knit community culture in Thua Thien Hue province, older adults with a high perceived level of social support are more likely to be surrounded by their families, neighbors and/or friends, which is beneficial in enhancing their wellbeing and physical health [42, 43], consequently maintaining their cognitive functions  and reducing the odds of FoF [17,18,19]. Thus, it can be inferred that good social support is valuable for eliminating the FoF in the elderly, irrespective of whether they suffer from cognitive impairment or not.
Strengths and limitations
To the best of our knowledge, the current study is the first study conducted in Vietnam to assess the association between FoF and cognitive function in the elderly according to their social support status. The study was conducted on a large representative sample of Vietnamese elderly, with a high response rate and an appropriate sampling method. FoF, perceived social support and cognitive function were measured using standardized and validated scales. However, being a cross-sectional study, no causal inference could be drawn and the results should therefore be interpreted with caution. A further longitudinal study is needed to investigate the two-way associations between the FoF and other factors. In addition, it is possible that the number of previous falls was underreported, because there was no direct question addressing this, and a history of fall was explored based on the types of bodily injury experienced by the participants during the last 12 months. However, we still found a significant association between a history of fall and a high FoF.
Given the fact that the coefficient of determination Pseudo R2 indicated that the independent variables explained 31 and 36% of variance of FoF in model 1 and model 2 respectively, results from this study should be interpreted with caution. However, the Goodness of Fit test suggested that these models were a good fit to the data as p = 0.6 (> 0.05, model 1) and p = 0.7 (> 0.05, model 2). Future studies are encouraged to consider other factors such as living conditions, psychology or physiological function in the research components.
Our results highlight the important implications of the FoF for the vulnerable elderly living in the community. Although cognitive or functional impairment in older adults is inevitable with advancing age, factors such as the social support status are modifiable. Despite the fact that from a cultural perspective, Vietnamese elderly are respected and well supported, more than a half of the elderly (423/725) in this study perceived low or moderate level of social support in this study. With the recognition of this emerging issue, identification of isolated elderly in the community might be one of effective approaches for addressing individuals with a high risk of FoF. This could be done in practice by keeping close communication with local organizations such as Associations of the Elderly at community level, community health centers or provincial People’s Committees. Therefore, involvement of caregivers and stakeholders are important in improving the social support status of the elderly, by developing public programs focusing on social relationships and interactions, along with strengthening of the infrastructure in community to encourage more participation from the elderly.