Dementia is a progressive neurodegenerative syndrome characterized by acquired deterioration in cognition, function, and behavior, which interferes with independence in daily activities (1). It has been estimated that approximately 57.4 million people worldwide were affected by dementia in 2019 and this number is projected to reach 152.8 million by 2050 (2). With the unprecedented aging of the population, dementia has imposed a tremendous burden on health care resources and significantly impaired the quality of life among affected individuals and their families (3). Despite advances in symptomatic treatments, curative or effective disease-modifying therapies for dementia are lacking (4), thereby making prevention of this disease a major public health priority (5). As a preclinical, transitional state between healthy cognitive aging and dementia (6), mild cognitive impairment (MCI) represents a window of opportunity for intervention in which the impairments of one or more cognitive domains occur with specific neuropathological and biomarker changes (7). Therefore, a better understanding of the potentially modifiable risk factors for MCI is urgently needed to prevent or postpone the onset of dementia (8).

Hypoperfusion and hypometabolism in temporoparietal cortices have been recognized as pivotal neurobiological features of MCI (7). As the primary cation in extracellular fluid, sodium ion regulates the water-electrolyte balance, blood volume, and osmotic pressure (9). Abnormalities of serum sodium concentrations can trigger an imbalance of osmotic pressure between the outside and inside of the cell, potentially leading to disturbances in cerebral blood volume and neuronal activities (10, 11). Additionally, the underlying pathophysiological mechanisms of neurodegeneration involve oxidative stress and inflammation (12), both of which can be activated by altered sodium homeostasis (13, 14). Several epidemiological studies have revealed that lower serum sodium was associated with cognitive decline among older persons (15, 16); however, research conducted in the general population found null associations between lower circulating sodium levels and worse overall cognitive function (17).

The homeostasis of serum sodium can be profoundly disturbed by chronic diseases (18, 19), which also constitute significant risk factors for frailty, particularly in older adults (20). As a geriatric syndrome characterized by the accumulation of deficits (20), frailty also encompasses the domain of nutrition. It has been reported that older adults with nutritional frailty, the co-occurrence of physical frailty and nutritional imbalance which includes high dietary sodium intake (21), are at higher risk for all-cause mortality (22). Moreover, frailty shares common biological pathways with cognitive impairment, including elevations in biomarkers of oxidative stress and inflammatory cytokines (23). Numerous studies have shown a significant association between frailty and the risk of cognitive decline (24–26). Individuals with frailty experience more severe health consequences owing to declines in physiological reserves across multiple organ systems (20) and may exhibit impaired compensatory responses to sodium imbalances. Nevertheless, the joint associations of frailty and circulating sodium levels with MCI risks remain unexplored.

Therefore, this study aimed to investigate the associations of serum sodium concentrations as well as frailty with the risk of MCI, among Chinese hospitalized older adults with chronic diseases. Additionally, this study sought to determine whether the association between serum sodium levels and MCI risk varies according to frailty status. It was hypothesized that serum sodium concentrations and frailty are associated with MCI risk, and furthermore, the association between serum sodium and MCI risk depends on the frailty status of the individuals.

The characteristics of the study participants are presented in Table 1. Compared with participants in the middle tertile of serum sodium, those in the lowest and highest tertile groups were less likely to be married and to have a history of stroke. Lower serum sodium levels were positively associated with smoking and alcohol consumption. In addition, participants with frailty were older and tended to be non-drinkers, in contrast to those who were robust or prefrail.

Both the lowest and highest tertiles of serum sodium were positively associated with the risk of MCI, with OR of 1.91 and 1.14, respectively; however, only the association between the lowest tertile of serum sodium and MCI risk reached statistical significance (p = 0.01). Adjustments for gender and age somewhat attenuated these associations. After further controlling for BMI, smoking status, and alcohol consumption, participants in the lowest tertile of serum sodium had a 1.77-fold increased risk of MCI (95% CI: 1.02–3.06) compared with those in the middle tertile, whereas no statistically significant associations were observed between the highest tertile of serum sodium and MCI risk (OR = 0.99, 95% CI: 0.55–1.77) (Table 2). Restricted cubic spline analysis demonstrated a U-shaped relationship between serum sodium concentrations and MCI risk, with a nadir at approximately 143 mmol/L (Figure 2). Below 143 mmol/L, the risk of MCI was 1.38 (95% CI: 1.03–1.84) for per 1 SD decrease in serum sodium.

Regarding frailty status, a significantly positive association was observed between frailty and the risk of MCI (OR = 5.28, 95% CI: 2.70–10.30); however, no significant associations were found for prefrailty. The gender- and age-adjusted model yielded a similar pattern of associations. When BMI, smoking status, and alcohol consumption were further adjusted for, participants with frailty still demonstrated an increased risk of MCI compared with those who were robust (OR = 3.99, 95% CI: 1.95–8.17). Additionally, a significantly increasing linear trend was observed between frailty status and MCI risks (P for trend<0.01) (Table 2).

To estimate the joint association of serum sodium and frailty status with MCI risk, frailty status was dichotomized as robust/prefrailty and frailty, as prefrailty was not significantly associated with MCI risk. For participants who were robust or prefrail, both the lowest and highest tertile of serum sodium were not significantly associated with the risk of MCI. However, compared with the reference group of robust/prefrail participants in the middle tertile of serum sodium, those with frailty in the lowest and highest tertile of serum sodium demonstrated 5.52 (95% CI: 2.09–14.52) and 3.48 (95% CI: 1.20–10.07) times the risk of MCI, respectively (Figure 3).

In sensitivity analyses excluding 13 individuals with cancer, the associations of serum sodium and frailty status with MCI risk were largely similar to the results from primary analyses (Supplementary materials).

The present study indicates that lower serum sodium concentration and frailty are associated with an elevated risk of MCI. A significantly positive association was also observed between higher serum sodium levels and the risk of MCI among individuals with frailty. These findings extended the existing evidence by demonstrating that the relationship between serum sodium and the risk of MCI varies according to frailty status.

Previous studies have examined the association between lower serum sodium and risks of cognitive deterioration. A cross-sectional study conducted in the elderly US population revealed that hyponatremia was significantly associated with cognitive change, particularly in memory and executive function (15). In a cohort study of older community-dwelling men, individuals with lower serum sodium levels had 1.30 times the odds of cognitive impairment compared to those with sodium levels of 141–142 mmol/L (16). Despite the heterogeneity in study samples and the various assessment instruments for cognitive function, the consistent findings across studies are that a reduction in serum sodium concentration can adversely impact cognitive functioning. Our study similarly suggests that lower serum sodium is significantly associated with an elevated risk of MCI among hospitalized older adults with chronic conditions. The neurological detriment of lower serum sodium can primarily be attributed to the abnormality of osmotic pressure in the brain. In response to hypotonicity, astrocyte swelling can induce the release of electrolytes and organic osmolytes (35), some of which act as excitatory neurotransmitters (e.g., glutamate) and play a crucial role in neuronal damage (10). Another potential mechanism linking lower serum sodium to MCI involves impaired neuronal mitochondrial distribution and decreased ATP production (36). In primary neurons cultured in a hyponatremic medium, Fujisawa et al. found that mitochondria disappeared from neurites, suggesting that chronic hyponatremia impairs mitochondrial function (36), which may deleteriously affect neuritic transport by regulating cellular energy production and oxidative stress control (37). In older adults, chronic hyponatremia is common but often rather mild and asymptomatic (38). Our results indicate that reduced serum sodium could constitute a modifiable target for preventing cognitive dysfunction, especially in older adults with chronic diseases.

A significant association between frailty and MCI risk was observed, aligning with previous evidence (23, 39, 40). For example, a cross-sectional study of Chinese middle-aged and elderly people found that frailty was significantly associated with lower Mini-mental State Examination scores and increased MCI risk (39). Similarly, cohort studies of community-dwelling older adults have demonstrated that frailty is associated with a subsequent decline in cognitive function within 12 months (23) or over 2 years (40). The possible mechanisms underlying the association of frailty and cognitive impairment may be diverse, including inflammation and oxidative stress. Chronic proinflammatory activation (e.g., IL-6, TNF-α) is a cardinal pathophysiologic process contributing to frailty (41), and inflammation has been implicated in the neuropathological cascade through modulating the amyloid protein precursor processing and increasing the production of amyloid-β42 peptide (42). Moreover, frailty is associated with oxidative imbalance due to overproduction or insufficient elimination of reactive oxygen species (43, 44). Compared to robust older adults, individuals with frailty have higher levels of oxidative stress (45), which may trigger synaptic dysfunction and cognitive decline (46, 47). It has been proposed that the detrimental association between frailty and cognitive impairment could be bidirectional (48). Therefore, interventions aimed at improving components of frailty may mitigate the risk of MCI in hospitalized older adults with chronic diseases.

It should be noted that higher serum sodium was significantly associated with an elevated risk of MCI in older adults with frailty, while no associations were found for individuals who were robust or prefrail. Similarly, Wei and colleagues found that functional disability attributed to malnutrition were more likely to be associated with frailty (49). One potential mechanism of action for these divergent associations is that frailty could aggravate hyperosmotic stress and cell dehydration, severely impacting the structure and function of intracellular proteins (50). Our study provides complementary evidence supporting the notion that frailty represents a state of increased vulnerability to poor resolution of homeostasis following stress (20). Sodium is one of the most important extracellular ions for regulating osmotic pressure, which governs the exchange of water between intracellular and extracellular spaces (51). Once the body systems lose their in-built reserves, frailty could make cells more sensitive to elevated serum sodium, potentially contributing to dehydration (52). Suboptimal hydration may subsequently increase cortisol levels (53), which can affect memory function and speed of information processing (54). Another potential mechanism is pertinent to sodium-induced inflammation (55). Higher serum sodium may upregulate systemic inflammatory responses (14) and amplify the already elevated inflammatory responses in older adults with frailty (56), thereby synergistically exacerbating the neurotoxic effects of inflammation. The results of our study indicate that higher serum sodium can be a surrogate marker for MCI in older adults with frailty. Nonetheless, further studies are needed to elucidate the mechanism underlying the observed associations of serum sodium with MCI risk, particularly as modified by frailty status.

This study has several limitations that should be acknowledged when interpreting the findings. Firstly, the cross-sectional design precludes the establishment of causal relationships between exposures and health outcomes. Future research should employ longitudinal designs or randomized controlled trials to determine the temporal sequence of serum sodium levels, frailty, and MCI. Secondly, although several potential confounders were collected and controlled for in the multivariate models, we could not rule out the possibility that unmeasured confounders may have influenced the results. Thirdly, almost all the participants had a serum sodium level not exceeding the criteria of hyponatremia and hypernatremia, which may underestimate the detrimental effects of altered sodium homeostasis on cognitive functioning. Fourthly, because natremia is a highly variable value, and cerebral cells can respond to changes in serum osmolality to maintain homeostasis, a single, punctual value may not be sufficiently representative. Therefore, serum sodium variance over time should be considered when examining the effects of serum sodium on MCI risks. Fifthly, the MCI of participants was measured by the MoCA scale only, a brief screening tool for global cognitive function, rather than a combination with comprehensive neuropsychological test assessments. However, this tool has indicated high sensitivity and specificity for the detection of MCI in various populations (57, 58). Finally, the findings of this study, conducted among hospitalized older adults with chronic diseases, may not be generalizable to other populations due to the specific characteristics and conditions of the study sample.

In conclusion, participants with lower serum sodium or frailty are more likely to have MCI. If meeting the criteria of frailty, higher serum sodium also imposes an increased risk of MCI to hospitalized older adults with chronic diseases. Additional studies should be conducted to confirm whether the correction of lower and higher serum sodium in older adults with frailty improves their cognitive function. In addition, future studies are needed to examine if these associations are similar in other aging populations and to elucidate the mechanism underlying such relationships.