Serum lipidomic analysis identifies VLSFAs as inversely associated with NfL and positively with cognitive performance.

As population aging accelerates, age-related neurodegenerative diseases present a growing challenge to healthcare systems worldwide (Yang et al., 2013). Although these diseases can be highly complex and etiologically diverse, neurodegeneration—a progressive process encompassing both structural damage to the nervous system, such as neuroaxonal injury, and ensuing functional decline, including cognitive impairment—stands as a common and pervasive pathological feature (Katsnelson et al., 2016). Key risk factors that contribute to neurodegeneration include chronic hypertension (Kruyer et al., 2015; Santisteban and Iadecola, 2018; Jeon et al., 2019), high inflammatory state (Glass et al., 2010), hypoxia (Peers et al., 2009), and lipid and immune abnormalities (Hallett et al., 2019). Among these, hypertension stands out due to its global prevalence and has been recognized as a major cause of age-related cognitive impairment (Iadecola and Gottesman, 2019). To easily monitor neurodegeneration, serum neurofilament light chain (NfL) has emerged as an instrumental biomarker. NfL is a component of the neuron-specific type IV intermediate filament and released after neuroaxonal damage. NfL shows high specificity and has been validated in several neurological diseases, including Parkinson’s disease (Aamodt et al., 2021), Alzheimer’s disease (Quiroz et al., 2020), multiple system atrophy (Chelban et al., 2022), and multiple sclerosis (Benkert et al., 2022). Consequently, NfL has become an essential tool for assessing the efficacy of interventions targeting neurodegeneration. Meanwhile, standardized cognitive function tests provide direct functional assessment of neurodegeneration, offering clinical relevance that complements molecular biomarkers (Gavett et al., 2018). Together, the measurement of NfL and cognitive assessments provide a comprehensive evaluation of neurodegenerative processes, combining molecular insights with functional outcomes.

Very-long-chain saturated fatty acids (VLSFAs), characterized by a carbon chain length greater than 20 without double bonds, hold potential as neuroprotective interventions in therapeutic applications (Lemaitre and King, 2022). Traditionally, the general public has been advised to reduce their intake of saturated fatty acids (FAs), as they are considered to be detrimental to health (Helgadottir et al., 2022). However, it is inappropriate to generalize the harm of all saturated FAs, as their health effects vary depending on the length of the carbon chain (Perna and Hewlings, 2022; Fretts et al., 2016). Extensive studies have demonstrated the cardiovascular benefits of VLSFAs in conditions such as hypertension (Li et al., 2020), heart failure (Lemaitre et al., 2018; Flock and Kris-Etherton, 2013), atrial fibrillation (Fretts et al., 2014), coronary artery disease (Malik et al., 2015; Papandreou et al., 2019), and sudden cardiac arrest (Lemaitre et al., 2014). Interestingly, the health benefits of VLSFAs may extend beyond the cardiovascular system. In the Atherosclerosis Risk in Communities (ARIC) cohort, researchers found that higher levels of circulating VLSFAs in midlife were associated with less cognitive decline 20 years later, suggesting a potential neuroprotective role for VLSFAs (Li et al., 2020). More recent cross-sectional analyses of National Health and Nutrition Examination Survey (NHANES) data from older adult populations have further explored these associations, albeit with some specificity for individual VLSFA species. For instance, higher circulating concentrations of docosanoic acid (C22:0) and lignoceric acid (C24:0) were found to be associated with better global cognitive function and delayed recall (Shen et al., 2024). Separately, higher serum tricosanoic acid (C23:0) levels have been linked to improved cognitive performance, with complementary evidence indicating reduced C23:0 expression in the frontal cortex of Alzheimer’s disease patients (Yang et al., 2025). Despite these emerging associations between specific circulating VLSFAs and cognitive outcomes, the underlying mechanisms and potential involvement of molecular pathways remain unclear. Further validation of the neurological benefits of VLSFAs and exploration of their mechanisms are warranted.

Therefore, building upon the longitudinal cognitive findings from the ARIC cohort which highlighted a gap in understanding the underlying mechanisms, this cross-sectional study aims to advance the field by investigating the association between VLSFAs (both independently and jointly) and neurodegeneration within the NHANES 2013–2014 dataset. Specifically, we extend prior work by: (1) Utilizing serum NfL, a sensitive molecular biomarker of neuroaxonal damage, in conjunction with functional cognitive performance measures to provide a multi-faceted assessment of neurodegeneration, thereby complementing previous cognitive-only studies. (2) Explicitly exploring the potential mediating role of hypertension, a key modifiable cardiovascular risk factor linked to both VLSFAs and neurodegeneration, to elucidate potential biological pathways underlying the observed associations. (3) Employing sophisticated statistical mixture models (WQS, BKMR, Qgcomp) to assess the combined impact of VLSFAs, offering insights beyond single-lipid analyses.

In this cross-sectional study, we investigated the association between circulating fatty acids (FA) profiles and serum neurofilament light chain (NfL), a biomarker of neurodegeneration. Our findings revealed that lower levels of circulating VLSFAs were associated with higher serum NfL concentrations, with C22:0 and C23:0 contributing substantial negative weights. In this association, hypertension mediated approximately 15–20% of the indirect effects. Subgroup analyses and correlation analyses suggested that diet quality played an important role in this association and may be a potential intervention avenue against neurodegeneration.

The observed neuroprotective associations of very-long-chain saturated fatty acids (VLSFAs) likely arise from a combination of mechanisms. Our finding that hypertension partially mediates the VLSFA-NfL relationship (accounting for approximately 15–20% of the indirect effect) supports an indirect neuroprotective pathway via improved cardiovascular health. Contextualizing this magnitude is informative; while direct comparisons across studies are complex due to methodological heterogeneity, the substantial impact of hypertension on neurodegenerative risk is well-established. For instance, a large UK cohort study (Zheng et al., 2022) quantified this risk, reporting a 10% increase in incident dementia associated with hypertension in patients with type 2 diabetes (HR 1.10, 95% CI 1.03–1.18). Although representing a direct risk estimate rather than a mediation proportion, this finding underscores the clinical relevance of blood pressure control in neurodegeneration. Our estimate that 15–20% of the VLSFA-NfL association operates via hypertension aligns conceptually with this perspective, indicating that the vascular pathway is a meaningful, albeit partial, contributor to the overall observed VLSFA effect. By potentially contributing to lower blood pressure, VLSFAs may mitigate cerebrovascular damage and associated neuronal injury.

Beyond these vascular benefits, direct neurobiological roles for VLSFAs are also strongly indicated. Specific VLSFAs, particularly odd-chain species like C23:0, are crucial structural components of neuronal membranes and myelin, notably within glycosphingolipids (Svennerholm and Stallberg-Stenhagen, 1968). Adequate levels of these fatty acids are essential for maintaining myelin integrity, neuronal membrane stability, and the function of specialized membrane domains. The pronounced protective effect of C23:0 observed at lower concentrations in our study may reflect the fulfillment of such fundamental structural or functional requirements within the nervous system, consistent with its significant presence in brain lipids. This aligns with longitudinal findings from the Atherosclerosis Risk in Communities (ARIC) cohort, where higher midlife VLSFAs were linked to attenuated cognitive decline decades later (Li et al., 2020), suggesting long-term direct effects on brain resilience.

Collectively, these lines of evidence suggest that VLSFAs may confer neuroprotection through both indirect (cardiovascular-mediated) and direct (neural-structural/functional) pathways. The interplay and relative importance of these mechanisms likely vary and warrant further investigation to fully elucidate the neuroprotective profile of VLSFAs.

The health-promoting benefits of polyunsaturated fatty acids, particularly omega-3 and omega-6 FAs, have been widely recognized in the relation to cardiovascular diseases, cancer, and autoimmune diseases (Kapoor et al., 2021). Instead, it is recommended to reduce the intake of saturated FAs, as they are regarded to be detrimental to health (Helgadottir et al., 2022). However, this does not seem to be the case with circulating VLSFAs (more than 20 carbons, without double bonds) (Lemaitre and King, 2022). Extensive cardiovascular benefits of VLSFAs have been demonstrated in heart failure (Lemaitre et al., 2018; Flock and Kris-Etherton, 2013), atrial fibrillation (Fretts et al., 2014), coronary artery disease (Malik et al., 2015; Papandreou et al., 2019), and sudden cardiac arrest (Lemaitre et al., 2014). Particularly, Li et al. found that circulating VLSFAs was negatively associated with the risk of pregnancy-induced hypertension and BP levels (Li et al., 2020). Consistently, our study reveals a negative joint association between VLSFAs and hypertension in the FA profiles. This association remains between the independent FAs (C22:0, C23:0, and C24:0) and hypertension. These findings verify a potential antihypertensive effect of VLSFAs.

In addition to cardiovascular benefits, VLSFAs also exhibit potential neuroprotective effects. In the ARIC study, Li et al. were the first to report that midlife circulating VLSFAs were negatively associated with cognitive decline 20 years later, as assessed by three cognitive tests: the Delayed Word Recall Test (DWRT), the Digit-Symbol Substitution Test (DSST), and the Word Fluency Test (WFT). While foundational, this longitudinal work primarily focused on functional outcomes and did not incorporate molecular biomarkers of neurodegeneration or explore specific mediating pathways. Our study significantly extends these findings in several key ways. First, using a cross-sectional design in the NHANES cohort, we corroborate the association between higher VLSFA levels and better cognitive performance, particularly in processing speed (DSST) and memory (delayed recall) domains (Supplementary Table S4). Second, and critically, we provide direct molecular evidence by demonstrating an inverse association between VLSFAs and serum NfL (Table 2), a validated biomarker of neuroaxonal damage. This integration of both molecular (NfL) and functional (cognitive performance) assessments provides a more comprehensive picture of VLSFAs’ potential neuroprotective role than previously available. Third, as a key mechanistic advancement, our mediation analysis suggests that the protective association of VLSFAs may be partially channeled through pathways involving lower hypertension risk (Figure 6), offering a potential biological explanation for the observed benefits. These results, demonstrating associations with both NfL concentrations and cognitive performance, underscore the potential importance of VLSFAs in maintaining neurological integrity.

Given that cognitive decline is often accompanied by structural neurodegeneration (Sokolov et al., 2020), our dual-outcome approach offers a more comprehensive understanding of this relationship (Table 2; Supplementary Tables S4, S5).

Hypertension has been recognized as a major cause of age-related cognitive impairment and accelerates the process of neurodegeneration (Li et al., 2020). In the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study (Levine et al., 2019), Levine et al. found that elevated systolic blood pressure (SBP), but not diastolic blood pressure (DBP), was associated with an accelerated cognitive decline over an 8-year period, suggesting a more detrimental effect of elevated SBP on neurodegeneration. Consistently, our study confirmed a positive association between hypertension and NfL, which was more pronounced with increasing SBP levels (Table 2).

Several potential genetic and behavioral factors may account for the association between VLSFAs and neurodegeneration. First, the circulating levels of VLSFAs have been found to be partially determined by genetics (Lemaitre et al., 2015), which further influences the development of diseases such as type 2 diabetes (Luo et al., 2021). Therefore, the association between VLSFAs and neurodegeneration may also be influenced by genetic background. Second, unhealthy dietary patterns have been identified as critical factors contributing to neurodegeneration (Wieckowska-Gacek et al., 2021) and cognitive aging (McGrattan et al., 2019). Additionally, circulating levels of VLSFA are closely related to dietary patterns (Pfeuffer and Jaudszus, 2016). Therefore, low levels of VLSFA are indicative of an unhealthy diet, which may exacerbate the risk of neurodegeneration.

The consistent findings between NfL (Table 2) and cognitive function assessments (Supplementary Tables S4, S5) in our study have significant clinical implications. While NfL serves as a sensitive molecular marker for neurodegeneration, as demonstrated by its association with cognitive decline (Supplementary Table S5), cognitive function tests directly reflect the functional impact of neurological changes. The observation that very-long-chain saturated fatty acids (VLSFAs) show beneficial associations with both outcomes suggests that these fatty acids may influence neurological health through multiple pathways. This comprehensive evidence highlights the potential value of VLSFA-focused interventions in preserving both the structural and functional aspects of neurological health.

There are notable findings in our study that warrant attention. First, dietary behaviors may play a crucial role in association between VLSFAs and NfL. On the one hand, we found that the majority of circulating FA concentrations exhibit a significant positive correlation with the dietary intake of corresponding FAs (Supplementary Table S8). On the other hand, compared to individuals with good diet quality, individuals with poor diet quality obtained more pronounced benefits from elevated concentrates of C23:0 (P for interaction = 0.001) and C24:0 (P for interaction = 0.035) (Figure 4). In support of this, Lai et al. found that dietary pattern with higher saturated fat intake and nut consumption contributed to a higher concentration of circulating C22:0 and C24:0 (Lai et al., 2023). Besides, higher dietary intake of VLSFAs is also associated lower risk of diseases such as metabolic syndrome (Lee et al., 2015). Therefore, improving dietary patterns is expected to be an avenue to prevent neurodegeneration. Regarding potential clinical implications, our findings suggest avenues for targeted strategies, although definitive recommendations require further investigation, particularly through intervention trials. The observed non-linear association between VLSFAs and NfL, characterized by a more pronounced protective effect at lower circulating concentrations (Figures 2, 3A), suggests a potential threshold effect. This implies that interventions aimed at increasing VLSFA levels might be most beneficial for individuals with initially low circulating concentrations. While our study did not determine optimal intake levels, the data suggest that achieving levels above the lower range (e.g., moving individuals from the lowest quartile or below the median toward higher concentrations) could be associated with reduced markers of neuroaxonal damage. Furthermore, the subgroup analysis indicating potentially greater benefits from higher C23:0 and C24:0 levels among individuals with poorer diet quality (low Healthy Eating Index-2015 score, Figure 4) tentatively identifies this group as a potential target population for future dietary interventions focusing on VLSFA-rich food sources. However, translating these circulating levels into specific dietary recommendations necessitates further research to understand the dose–response relationship between dietary VLSFA intake and resulting circulating concentrations, as well as their ultimate impact on neurological outcomes in prospective and interventional settings.

Second, compared with C22:0 and C24:0, we found a more pronounced and notably non-linear association between C23:0 and NfL. Specifically, when circulating C23:0 levels were relatively low (e.g., below the median), NfL concentrations exhibited a significant increase with further decreases in C23:0. This observation suggests a potential threshold effect, implying that C23:0 may be an essential component for maintaining the integrity of neural structures, with its protective influence being particularly critical at lower physiological concentrations. Several plausible biological mechanisms could underlie this non-linear relationship. One possibility relates to the efficiency of C23:0 incorporation into vital neural lipids or its role in specific cellular processes that become saturated or achieve optimal function at a certain threshold. For instance, as an odd-chain fatty acid, C23:0 is incorporated into complex lipids such as sphingolipids and glycerophospholipids, which are integral to neuronal membranes and myelin sheaths. At very low C23:0 availability, the synthesis or maintenance of these crucial structures might be compromised, leading to increased neuronal vulnerability and higher NfL levels. Initial supplementation or an increase in C23:0 up to a certain physiological requirement could therefore lead to a substantial restoration of membrane integrity or function, resulting in a marked decrease in NfL. However, once these structural or functional demands are met, further increases in C23:0 might yield diminishing returns in terms of NfL reduction via these specific pathways, or its effects might plateau. Svennerholm et al. highlighted the physiological relevance of odd-chain FAs within the neurological system, finding that C23:0 and C25:0 constitute a significant portion (10% of total non-hydroxylated FAs and 25% of total hydroxylated FAs) of adult human brain glycosphingolipids (Svennerholm and Stallberg-Stenhagen, 1968). This underscores the quantitative importance of C23:0 in brain lipid architecture. Given that odd-chain FAs like C23:0 are primarily obtained through dietary intake or gut microbial metabolism, our findings suggest that dietary interventions aimed at ensuring adequate C23:0 levels, particularly for individuals with lower baseline concentrations, could hold promising prospects for preventing or mitigating neurodegeneration. The pronounced effect observed at lower C23:0 levels may indeed point toward a fundamental requirement that, once fulfilled, provides a significant degree of neuroprotection.

This study has several limitations. First, the cross-sectional design precludes establishing temporal sequences between VLSFAs, hypertension, NfL, and cognition. Consequently, causal inferences cannot be drawn, and the possibility of reverse causality cannot be excluded. Specifically, the observed inverse association between VLSFA levels and NfL could potentially arise if the neurodegenerative process itself leads to alterations in systemic lipid metabolism that result in lower circulating VLSFA concentrations. If such a pathway exists, our findings might overestimate the protective effect of VLSFAs, or the association could, in part, reflect the metabolic consequences of neurodegeneration rather than a purely protective role of VLSFAs against it. Similarly, while less likely given that hypertension is often considered a precursor, it’s conceivable that advanced neurodegenerative states could influence autonomic regulation affecting both blood pressure and lipid profiles concurrently, further complicating the interpretation of the VLSFA-hypertension-NfL nexus. This lack of temporality significantly impacts the interpretation of the mediation analysis. While exploratory, the mediation findings rely on strong, untestable assumptions: specifically, the assumed temporal order (VLSFAs → hypertension → NfL/cognition) and, critically, the absence of unmeasured confounding factors that could influence all three components (e.g., specific genetic predispositions affecting both lipid metabolism and neurological health, detailed dietary patterns beyond the Healthy Eating Index-2015 (HEI-2015) score, or low-grade systemic inflammation). Therefore, the mediation results (15–20% effect) should be viewed cautiously as hypothesis-generating regarding potential pathways, rather than conclusive evidence. Second, while we focused on age-related neurodegeneration markers, NHANES data does not allow for the complete exclusion of specific underlying neurological conditions that might affect NfL levels. Third, the use of an unweighted sample, chosen to facilitate the complex mixture modeling, may limit the direct generalizability of the quantitative estimates to the entire US population, although the NHANES sampling design itself aims for national representation. Fourth, while NfL is a valuable marker of neuroaxonal damage, it lacks pathological specificity and cannot localize damage without complementary data like neuroimaging.

Our serum lipidomic analysis identified very-long-chain saturated fatty acids (VLSFAs: C22:0, C23:0, and C24:0) as potential therapeutic targets for neurodegeneration, demonstrating significant inverse associations with serum neurofilament light chain (NfL) levels and positive correlations with cognitive performance. The stronger protective association at lower VLSFA concentrations suggests a clinically relevant threshold effect. Mechanistic investigation revealed hypertension as a partial mediator, accounting for approximately 15–20% of the relationship between VLSFAs and neurodegeneration biomarkers. These findings highlight dual neuroprotective pathways through which these lipid species may confer benefits. Further interventional studies are warranted to determine whether targeted strategies (e.g., dietary modifications) aimed at modulating circulating VLSFA levels, particularly in individuals with lower baseline concentrations or poorer diet quality, can effectively mitigate neurodegeneration and enhance cognitive function.