Association of Serum Neurofilament Light Chain with Depressive Symptoms: a population-based study

doi:10.21203/rs.3.rs-2337172/v1

Download PDF

Research Article

Association of Serum Neurofilament Light Chain with Depressive Symptoms: a population-based study

https://doi.org/10.21203/rs.3.rs-2337172/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Neurofilament light chain proteins (NfL) are widely accepted biomarkers of neuronal injury, and depressive symptoms are related to neuronal injury. Nevertheless, evidences of the association between NfL and depressive symptoms are still limited. This is the first population-based study to examine the association between NfL and depressive symptoms in general population.

Methods

We included 1907 participants with measurement of serum NfL from the 2013–2014 U.S. National Health and Nutrition Examination Survey. Depressive symptoms were measured by the 9-item Patient Health Questionnaire. Logistic regression models were conducted to examine associations between serum NfL levels and depressive symptoms. Restricted cubic spline regressions were applied to estimate the possible nonlinear relationships between them.

Results

Increased serum NfL were associated with higher risk of depressive symptoms after adjusted for confounder factors (per SD: odds ratios [OR] = 1.16, 95% confidence intervals [CI]: 1.02–1.34). Individuals with higher serum NfL (> 21.8 pg/ml, the upper 20% quantile) had a higher likelihood of depressive symptoms compared to normal counterparts (OR = 1.50, 95% CI: 1.01–2.22), and sensitivity analysis using different grouping criteria provided similar results. Furthermore, restricted cubic spline regression analysis demonstrated that a near-linear association occurred between serum NfL and the risk of depressive symptoms (P-nonlinear = 0.681, P-overall < 0.001).

Conclusions

This study found linear association between serum NfL levels and depressive symptoms in general population. Our findings support that serum NfL levels may be a novel biomarker for depressive symptoms, further studies are needed to validate our findings underlying this association.

Epidemiology

Psychiatry

Neurofilament light chain proteins

Depressive symptoms

Epidemiology

Cohort

Depression is a common mental disorder that affects over 322 million people in the general population, the number of persons with depressive disorders account for 4.4% and is still going up in the world(WHO, 2017). The Global Burden of Disease Study reported that depressive disorders are the leading cause of mental health-related disability burden worldwide(GBD.2019.Diseases.and.Injuries.Collaborators, 2022), and associated with all-cause and cardiovascular disease mortality among adults(Cuijpers et al., 2014; Leone et al., 2021; Meng et al., 2020). Additionally, depressive symptoms are related to many cytoarchitectural changes and injury in neurons(Banasr, Sanacora, & Esterlis, 2021; Csabai, Wiborg, & Czéh, 2018; Holmes et al., 2019; Williams et al., 2019).

Recently, neurofilament light chain proteins (NfL), a neuron-specific component of the axonal cytoskeleton, has been proposed as a potential indicator of neuronal damage(Khalil et al., 2018). It has been reported to associated with many neurological disorders, such as multiple sclerosis, dementia, stroke, Parkinson disease (PD) and traumatic brain injury(Gaetani et al., 2019; Khalil et al., 2018). Meanwhile, previous preliminary studies also investigated the association of serum NfL with depressive disorder and reported conflicting findings(Travica, Berk, & Marx, 2022). Several studies found that there were no significant associations between NfL and depressive symptoms(Besse et al., 2020; Isgren et al., 2017; Tauil et al., 2021), while increased evidences suggested significantly higher NfL levels in patients with major depression among case-control studies(Bavato et al., 2021; Chen et al., 2022). Furthermore, prospective studies also reported that elevated levels of NfL were associated with higher risk of depressive symptoms among specific populations, such as patients with ischemic stroke(Zhao et al., 2020), PD(Yin et al., 2022), and veterans with history of mild traumatic brain injury(Guedes et al., 2020). However, due to either the restricted sample size and settings, or specific populations, the relationship between NfL and depression was still controversial. Therefore, the association of NfL with depression remains to be further investigated.

In this study, we aimed to investigate the association of serum NfL with depressive symptoms in general population. We first hypothesized that serum NfL is a novel biomarker for depressive symptoms, and elevated levels of serum NfL would be associated with higher risk of depressive symptoms in general population, and then determined the cross-sectional association between serum NfL and depressive symptoms in the 2013–2014 National Health and Nutrition Examination Survey (NHANES).

Study population

We used data from the 2013–2014 NHANES, which is a nationally representative cross-sectional survey of civilian, noninstitutionalized persons living in the U.S.(Statistics). A total of 10175 participants were recruited. Of them, 2071 individuals were measured serum NfL and 5372 individuals were assessed depressive status, respectively. At last, 1907 participants with completing both serum NfL and depressive symptoms assessments were analyzed in this study. Due to the publicly available and secondary analysis of the data, no additional ethical review was necessary.

Depressive Symptoms

Depressive symptoms were assessed by the 9-item Patient Health Questionnaire (PHQ-9) among participants aged 20 years or older (Kroenke, Spitzer, & Williams, 2001; Spitzer, Kroenke, & Williams, 1999), which assesses frequencies at certain symptoms experienced in the last two weeks, ranging from 0 (‘not at all’) to 3 (‘nearly every day’). These questionnaires were self-administered, and the total scores ranged from 0 to 27. A cut-off score of 10 or above was used to define probable cases of depressive symptoms, with a sensitivity of 88% and a specificity of 88%(Kroenke et al., 2001).

Serum Neurofilament Light Chain Proteins

The NfL would release into the cerebrospinal fluid (CSF) after axonal injury and then into the peripheral blood(Khalil et al., 2018). Serum NfL was measured by a novel, high throughput, automated platform (Attelica) developed by Siemens Healthineers. A full description of the assay′s development and validation has been reported elsewhere(Lee et al., 2022). Briefly, the light reagent uses a monoclonal anti-NfL antibody labeled with a proprietary acridinium ester for chemiluminescent detection(Lee et al., 2022). In NHANES, serum NfL was measured in participants aged 20 years or older. In this analyses, serum NfL were evaluated as both continuous and categorized variables. Due to serum NfL exhibited a nonnormal distribution, serum NfL was log-transformed prior to performing analyses. Additionally, the serum NfL levels were dichotomized as high (upper 20% quantile ≥ 21.8 pg/ml vs lower 80% quantile < 21.8 pg/ml) and normal groups. Especially, in order to obtain robust results, we used different cut-off level to define high serum NfL. In brief, high serum NfL was defined as the upper 25% (≥ 19.3 pg/ml), 10% (≥ 30.2 pg/ml) and 5% (≥ 40.4 pg/ml) quantiles of serum NfL in sensitivity analyses, respectively.

Covariates

Covariates included sociodemographic information, lifestyle behaviors and clinical characteristics. In detail, sociodemographic information included age, gender, marital status (married or living with partners vs. single), education (less than college vs. college or more), and race/ethnicity (Mexican Hispanic, other Hispanic, Non-Hispanic White, Non-Hispanic black and other race). Lifestyles included smoking status (never, former, and current smokers), alcohol consumption, physical activity (PA) and body mass index. PA was assessed by a questionnaire, which included the frequency (time per week), duration (minutes per time) and intensity (vigorous or moderate) of PA in one typical week. The vigorous intensity of PA was doubled and added to minutes of the moderate intensity(Divney et al., 2019). According to the 2018 PA Guidelines, which recommend that adults should engage in at least 150–300 minutes per week of moderate-intensity PA, 75–150 min/week of vigorous-intensity PA, or an equivalent combination(Piercy et al., 2018). Meanwhile, the participants were also categorized into two groups: 1) participants who met the 2018 PA Guidelines (≥ 150 min/week) and 2) participants who did not meet the 2018 PA Guidelines (< 150 min/week). Clinical characteristics included cardiovascular disease (CVD), diabetes, cancer, and arthritis. CVD included angina angina/angina pectoris, heart attack, coronary heart disease and stroke. Especially, aforementioned diseases were defined as being told by a doctor or health professional that they have the disease. Additionally, systolic blood pressure (SBP) and diastolic blood pressure (DBP) were also measured.

Statistical Analysis

Continuous and categorical variables are presented as the means with standard deviation (SD) or frequency (%), respectively. Group differences were analyzed by Student’s t test, ANOVA, or Wilcoxon rank-sum test. We used logistic regression models to evaluate the association between NfL and the risk of depressive symptom using three models. Model 1: unadjusted; Model 2: adjusted for age, gender, race, educational status, smoking, alcohol assumption, married status, body mass index, and physical activity; Model 3: additional adjusted for SBP, DBP, cardiovascular disease, cancer, diabetes mellitus and arthritis. Firstly, when examining serum NfL as continuous variable, we analyzed the association of NfL with depressive symptom using logNfL as independent variable, and investigated the association of each per SD increase in serum NfL with depressive symptom. Secondly, for categorical variable, we used different cut-off level to define high serum NfL, and analyzed the relationships between different NfL groups and depressive symptom with three models. Thirdly, in further analyses, we investigated the association of logNfL with depressive symptom across various strata defined by age, gender, body mass index, race, marital status, smoking status, alcohol consumption and PA by using Model 3. Due to the number of specific diseases (such as CVD, diabetes, arthritis, and cancer) was relatively few, we did not examine the association of serum NfL with depressive symptom in these subgroups. Lastly, we applied restricted cubic spline regression, with four knots located at the 5th, 35th, 65th, and 95th percentiles of logNfL to estimate the possible nonlinear association of logNfL with depressive symptom by adjusting for the potential confounders factors(Durrleman & Simon, 1989). The test of potential nonlinearity was assessed by the chi-squared test, with compared the model with only the linear term to the model that included the cubic spline terms(Palmer et al., 2021). In these analyses, logNfL was analyzed as continuous variable and the odds ratios (OR) and 95% confidence intervals (CI) were calculated using the median value of logNfL as the reference value. All tests were considered significant at a P-value < 0.05 (2-tailed). All analyses were conducted using R statistical software (version 4.1.3; www.r-project.org).

In this study, we included 1907 participants (48.40% males and 51.60% females) from the 2013–2014 NHANES (Table 1). The mean age was 47.12 (15.49) years. Of these participants, 894 (46.88%) and 1013 (53.12%) participants were middle (aged 45–64 years) and older adults (aged ≥ 65 years), respectively. Overall, 179 (9.39%) participants were defined as having depressive symptoms. The more detail characteristics were presented in Table 1.

Table 1

The characteristics of study population.
		Serum NfL
	Overall	Normal*	High*	P-value
Sample, n (%)	1907	1525 (79.97%)	382 (20.03%)	-
Serum NfL, Median (IQR)	12.40 (8.30, 19.30)	10.60 (7.50, 14.60)	30.35 (24.90, 40.27)	< 0.001
Age, years, Mean (SD)	47.12 (15.49)	44.53 (14.81)	57.48 (13.73)	< 0.001
45–64, years, n (%)	894 (46.88)	808 (52.98)	86 (22.51)	< 0.001
≥ 65, years, n (%)	1013 (53.12)	717 (47.02)	296 (77.49)
Gender				0.050
Male, n (%)	923 (48.40)	721 (47.28)	202 (52.88)
Female, n (%)	984 (51.60)	804 (52.72)	180 (47.12)
Educational status				0.012
Less than college, n (%)	801 (42.00)	622 (40.79)	179 (46.86)
College or more, n (%)	1105 (57.94)	903 (59.21)	202 (52.88)
BMI, kg/m², Mean (SD)	29.30 (7.36)	29.14 (7.23)	29.96 (7.83)	0.050
Non-Obese (< 30), n (%)	1194 (62.61)	970 (63.61)	224 (58.64)	0.073
Obese (≥ 30), n (%)	713 (37.39)	555 (36.39)	158 (41.36)
Race/Ethnicity
Mexican American, n (%)	268 (14.05)	224 (14.69)	44 (11.52)	0.040
Other Hispanic, n (%)	185 (9.70)	154 (10.10)	31 (8.12)
Non-Hispanic White, n (%)	859 (45.04)	665 (43.61)	194 (50.79)
Non-Hispanic Black, n (%)	337 (17.67)	265 (17.38)	72 (18.85)
Other Race, n (%)	258 (13.53)	217 (14.23)	41 (10.73)
Marital status
Married^#, n (%)	1185 (62.14)	944 (61.90)	241 (63.09)	0.669
Single, n (%)	722 (37.86)	581 (38.10)	141 (36.91)
Smoking status
Nonsmoker, n (%)	1062 (55.69)	890 (58.36)	172 (45.03)	< 0.001
Former smoker, n (%)	433 (22.71)	322 (21.11)	111 (29.06)
Current smoker, n (%)	411 (21.55)	312 (20.46)	99 (25.92)
Alcohol consumption
Yes, n (%)	1415 (74.20)	1134 (74.36)	281 (73.56)	0.639
No, n (%)	489 (25.64)	388 (25.44)	101 (26.44)
PA, min/week, Median (IQR)	240 (20, 840)	270 (30, 900)	120 (0, 480)	< 0.001
< 150 min/week, n (%)	811 (42.53)	605 (39.67)	206 (53.93)	< 0.001
≥ 150 min/week, n (%)	1096 (57.47)	920 (60.33)	176 (46.07)
SBP, mmHg, Mean (SD)	121.52 (17.16)	119.78 (15.58)	128.49 (20.96)	< 0.001
DBP, mmHg, Mean (SD)	68.97 (12.33)	68.91 (11.73)	69.21 (14.48)	0.671
Cardiovascular disease				< 0.001
Yes, n (%)	141 (7.39)	83 (5.44)	58 (15.18)
No, n (%)	1766 (92.61)	1442 (94.56)	324 (84.82)
Diabetes
Yes, n (%)	209 (10.96)	120 (7.87)	89 (23.30)	< 0.001
No, n (%)	1698 (89.04)	1405 (92.13)	293 (76.70)
Arthritis				< 0.001
Yes, n (%)	472 (24.75)	316 (20.72)	156 (40.84)
No, n (%)	1432 (7.51)	1207 (79.15)	225 (58.90)
Cancer
Yes, n (%)	144 (7.55)	102 (6.69)	42 (10.99)	0.004
No, n (%)	1763 (92.45)	1423 (93.31)	102 (26.70)
Depressive symptoms				< 0.001
Yes, n (%)	179 (9.39)	123 (8.07)	56 (14.66)
No, n (%)	1728 (90.61)	1402 (91.93)	326 (85.34)
NfL: neurofilament light chain, SBP: Systolic blood pressure, DBP: Diastolic blood pressure, PA: Physical activity, IQR: interquartile range. High NfL and normal NfL groups were defined if serum NfL levels were ≥ 21.9 pg/ml (the upper 20% quintile) and < 21.9 pg/ml (the lower 80% quintile), respectively. Unknown: Educational status, 1 (0.05%); Smoking status, 1 (0.05%); Alcohol consumption, 3 (0.16%);
Arthritis, 3 (0.16%); ^# Included married or living with partner.

Table 2 shows the results of logistic regression analyses between NfL and risk of depressive symptoms. When examining serum NfL as continuous variable, higher levels of logNfL were associated greater risk of depressive symptom in unadjusted model (OR = 3.04, 95% CI = 1.84-5.00), and the association persisted after considering potential confounder factors (OR = 2.09, 95% CI = 1.09–3.96). In further analyses, per SD increase of serum NfL were associated with higher risk of depressive symptom in both unadjusted (OR = 1.25, 95% CI = 1.11–1.43) and adjusted model (OR = 1.16, 95% CI = 1.02–1.34). For categorical variable (Table 2), when dichotomized serum NfL as high (upper 20% quantile) and normal (lower 80% quantile) groups, high serum NfL groups were also associated with higher risk of depressive symptom both in unadjusted (OR = 1.96, 95% CI = 1.39–2.73) and adjusted model (OR = 1.50, 95% CI = 1.01–2.22). Sensitivity analysis using different grouping criteria provided similar results after adjusted for potential confounder factors (upper 25% vs lower 75% quantile: OR = 1.47, 95% CI = 1.00-2.15; upper 10% vs lower 90% quantile: OR = 1.66, 95% CI = 1.02–2.64; upper 5% vs lower 95% quantile: OR = 2.38, 95% CI = 1.30–4.21).

Table 2

Association of serum NfL with depressive symptoms.
		Model 1		Model 2		Model 3
	Event/N (%)	OR (95% CI)	P-value	OR (95% CI)	P-value	OR (95% CI)	P-value
Contiguous serum NfL
Per logNfL increased	179/1907 (9.39%)	3.04 (1.84, 5.00)	1.26e-5	2.21 (1.21, 3.99)	0.009	2.09 (1.09, 3.96)	0.025
Pre SD increased	179/1907 (9.39%)	1.25 (1.11, 1.43)	6.28E-4	1.19 (1.05, 1.36)	0.009	1.16 (1.02, 1.34)	0.035
Categorical serum NfL
Upper 25% quantile ^a	67/475 (14.11%)	1.94 (1.39, 2.66)	5.97e-05	1.50 (1.04, 2.14)	0.027	1.47 (1.00, 2.15)	0.050
Upper 20% quantile ^b	56/382 (14.66%)	1.96 (1.39, 2.73)	9.84e-05	1.50 (1.03, 2.17)	0.034	1.50 (1.01, 2.22)	0.046
Upper 10% quantile ^c	33/192 (17.19%)	2.23 (1.45, 3.33)	1.33e-4	1.68 (1.06, 2.61)	0.022	1.66 (1.02, 2.64)	0.038
Upper 5% quantile ^d	21/95 (22.11%)	2.97 (1.74, 4.87)	3.00e-5	2.30 (1.30, 3.90)	0.003	2.38 (1.30, 4.21)	0.004
OR: odds ratio, CI: confidence interval, SD: standard deviation, NfL: neurofilament light chain. ^{a, b, c, d} the reference is the lower 75% quantile (112/1432, 7.82%), 80% quantile (123/1525, 8.07%), 90% quantile (146/1715, 8.51%) and 95% quantile (158/1812, 8.72%), respectively. Model 1: Unadjusted, Model 2: adjusted for age, gender, race, educational status, smoking, alcohol assumption, married status, body mass index, and physical activity; Model 3: additional adjusted for SBP, DBP, cardiovascular disease, cancer, diabetes mellitus and arthritis.

Furthermore, we conducted stratified analysis across various strata defined by age, gender, educational status, body mass index, race, marital status, smoking status, alcohol consumption, and physical activity using multivariable Model 3 (Table 3). We found significant associations of logNfL with depressive symptoms in each stratum, and the interaction tests comparing the ORs across the strata were not significant (P-value > 0.05), suggesting that increased logNfL was associated with a higher risk of depressive symptoms, regardless of age, gender, educational status, body mass index, race, marital status, smoking status, alcohol consumption, and physical activity. Additionally, to visualize the potential dose-response relationship, multivariable restricted cubic spline regression was performed to predict the ORs for depressive symptoms between the 5th and 95th percentiles of the logNfL (Fig. 1). Interestingly, the association of serum NfL with the risk of depressive symptoms was nearly linear (P-overall < 0.001, P-nonlinear = 0.681).

Table 3

Association of serum NfL with depressive symptoms in subgroups.
	Event/N (%)	OR (95% CI)	P-value	P^φ
Age, years	179/1907 (9.39%)			0.567
45–64	60/894 (6.71%)	1.54 (0.53, 4.38)	0.421
≥ 65	119/1013 (11.75%)	2.27 (1.00, 5.03)	0.046
Gender				0.408
Male	64/923 (6.93%)	2.96 (1.09, 7.86)	0.031
Female	115/984 (11.69%)	1.69 (0.69, 4.08)	0.245
Educational status				0.982
Less than college	94/801 (11.74%)	2.08 (0.80, 5.22)	0.126
College or more	85/1105 (7.69%)	2.11 (0.83, 5.22)	0.109
BMI, kg/m²				0.358
< 30	77/1194 (6.45%)	3.04 (1.09, 8.17)	0.029
≥ 30	102/713 (14.31%)	1.63 (0.68, 3.87)	0.267
Race/Ethnicity				0.325
Mexican American	25/268 (9.33%)	0.70 (0.07, 5.91)	0.758
Other Hispanic	20/185 (10.81%)	0.50 (0.02, 8.73)	0.655
Non-Hispanic White	91/859 (10.59%)	2.67 (1.11, 6.31)	0.026
Non-Hispanic Black	31/337 (9.20%)	3.23 (0.62, 16.32)	0.157
Other Race	12/258 (4.65%)	0.09 (0.01, 3.39)	0.224
Marital status				0.270
Married	91/1185 (7.68%)	3.14 (1.28, 7.53)	0.011
Single	88/722 (12.19%)	1.49 (0.55, 3.95)	0.427
Smoking				0.075
Non-smoker	78/1062 (7.34%)	0.79 (0.26, 2.32)	0.680
Ever smoker	47/433 (10.85%)	3.94 (1.23, 12.70)	0.020
Current smoker	54/411 (13.14%)	4.08 (1.09, 14.86)	0.034
Alcohol consumption				0.452
Yes	130/1415 (9.19%)	2.45 (1.13, 5.21)	0.021
No	49/489 (10.02%)	1.37 (0.36, 4.91)	0.639
Achieve PA				0.271
Yes	63/1096 (5.75%)	3.26 (1.12, 9.22)	0.027
No	116/811 (14.30%)	1.54 (0.67, 3.47)	0.305
OR: odds ratio, CI: confidence interval, SD: standard deviation, NfL: neurofilament light chain. The OR was calculated based on the per unit increase in logNfL. All models were adjusted for age, gender, race, educational status, smoking alcohol assumption, married status, body mass index, and physical activity, SBP, DBP, cardiovascular disease, cancer, diabetes mellitus and arthritis. ^φ P represents the heterogeneity between subgroups based on the meta-regression analysis.

In this cross-sectional design of the 2013–2014 NHANES, we found that elevated levels of serum NfL were robustly and independently associated with higher risk of depressive symptoms in general population. Additionally, these associations were nearly linear. In summary, our findings support the hypothesis that serum NfL is a novel biomarker for depressive symptoms.

Previous studies have release the conflicting relationship between levels of NfL and depression(Travica et al., 2022). Several studies reported that there were no significant associations of NfL with depression(Besse et al., 2020; Isgren et al., 2017; Tauil et al., 2021). For example, Besse et.al recruited 15 patients with major depressive disorder receiving electroconvulsive therapy and 15 sex- and age-matched healthy controls, and found that serum NfL concentrations did not differ between patients and healthy controls (patients: 16.04 pg/ml, control: 15.57 pg/ml, P-value > 0.05)(Besse et al., 2020). Tauil et.al also did not found significant association between level of NfL in the cerebrospinal fluid (CSF) and depressive score (r = -0.20, P-value = 0.38), which included 10 controls, 14 untreated patients with relapse remitting multiple sclerosis (RRMS) and 16 RRMS patients treated with fingolimod(Tauil et al., 2021). Isgren et.al recruited 77 patients with bipolar disorder, and found that baseline CSF NfL was associated with depressive episodes at 6–7 years follow-up (OR = 1.19, 95% CI: 0.44–3.18)(Isgren et al., 2017). In these case-control studies, researchers did not find significant relationships between NfL and depressive symptoms.

In contrast, other studies demonstrated significant relationships between NfL and depression(Bavato et al., 2021; Chen et al., 2022; Guedes et al., 2020; Yin et al., 2022; Zhao et al., 2020). Chen et.al enrolled 40 patients with major depressive disorder, and 40 age- and sex-matched healthy controls, they found that patients with major depressive disorder exhibited significantly higher plasma NfL levels than controls (28.76 ± 22.53 vs 16.65 ± 8.07, P-value = 0.007)(Chen et al., 2022). Similarly, Bavato et.al included 41 patients with major depressive disorders and 485 healthy controls, and reported patients showed elevated serum levels of NfL than healthy controls(Bavato et al., 2021). Additionally, prospective studies also released that elevated levels of NfL were associated higher risk of depressive symptoms among several specific populations, such as patients with ischemic stroke(Zhao et al., 2020), PD(Yin et al., 2022), and veterans with history of mild traumatic brain injury(Guedes et al., 2020). In detail, Zhao et.al included 236 ischemic stroke cases in a single-center prospective cohort, and 55 patients were defined as post-stroke depression during 3-month follow-up period(Zhao et al., 2020). They showed that per IQR increase of serum NfL, the OR of incident depression was 2.65 (95% CI: 1.59–4.04) after adjusted potential confounder factors(Zhao et al., 2020). Yin et.al conducted a prospectively study with enrolling 116 patients with PD, and found plasma NfL levels were higher in patients with depression than in those without these symptoms (PD with no depression: 16.2 ± 5.9, PD with moderate depression 32.1 ± 20.5, PD with severe depression 31.2 ± 19.7) with all P-value less than 0.01(Yin et al., 2022). In addition, Guedes et.al enrolled 45 controls and 150 veterans with history of mild traumatic brain injury, they found that increase plasma levels of NfL were associated with higher depressive score (r = 0.204, P-value = 0.016)(Guedes et al., 2020). In summary, these studies demonstrated a significant association between NfL and depression, which was consistent with our findings.

Taken together, these studies were all case-control settings with small sample size, findings of the conflicting relationship between NfL levels and depression would be misinterpret due to either the restricted sample size, or settings, or specific populations. While, we investigated the association in a population-based cohort study with larger sample size, our findings would provide more compelling evidence that elevated levels of serum NfL were associated with higher risk of depressive symptoms in general population.

The mechanisms that increased levels of serum NfL were associated with depressive symptoms were not entirely understood. The possible mechanisms maybe proposed based on published literatures. NfL is a neuron-specific component of the axonal cytoskeleton, and released into the CSF and peripheral blood after neuroaxonal injury caused by neuroinflammatory, neurodegenerative, traumatic, or vascular injuries(Khalil et al., 2018). The levels of NfL can be quantified in serum and plasma(Ashton et al., 2021; Gisslén et al., 2016; Kuhle et al., 2016; Simrén, Ashton, Blennow, & Zetterberg, 2021). Depressive disorder is closely related to immunological and inflammatory dysfunctions(Beurel, Toups, & Nemeroff, 2020; Kappelmann et al., 2021; Miller & Raison, 2016), whereby activated neuroimmune and related oxidative pathways would induce damage in neurons, leading to the depressive symptoms(Hodes, Kana, Menard, Merad, & Russo, 2015; Wang et al., 2019; Wohleb, Franklin, Iwata, & Duman, 2016). In brief, serum NfL was a marker of axonal injury, and depression would lead to damage in neurons via immunological and inflammatory dysfunctions.

Our study has several strengths. The study was conducted based on the NHANES, which is a nationally representative cross-sectional survey of civilian, noninstitutionalized persons. This population-based study could generalize our findings to general population, regardless of age, gender, educational status, body mass index, race, marital status, smoking status, alcohol consumption, and physical activity. Secondly, until now, this is the first and largest study conducted among general rather than specific population. Hence, our findings of the relationship between NfL and depressive symptoms could be more reliable and robust. Additionally, to our knowledge, this study first reported a nonlinear association between serum NfL and depressive symptoms.

However, several limitations also presented in our study. First, the NHANES data are cross-sectional, and causal relationship between serum NfL and depressive symptoms was not investigated. Therefore, it would be interesting for future research to get more insight on the role of serum NfL on depressive symptoms to validate our findings in longitudinal and Mendelian randomization studies. Second, depressive symptoms were assessed by the PHQ-9, which is a self-report questionnaire. The accurate assessment of depression by clinical diagnosis is needed in further studies. Last, we just included participants from the USA, whether our findings may be generalized to other population would need further investigation.

In conclusion, serum NfL, a neuronal marker of neuronal injury, was robustly and independently associated with depressive symptoms in general population. Additionally, these associations were nearly linear. Our findings support the hypothesis that serum NfL is a novel biomarker for depressive symptoms. Longitudinal studies and objective measurements of depression are needed to validate our findings in the future.

Acknowledgements

The data and samples used for this research were obtained from the U.S. NHANES. We would like to thank the U.S. Centers for Disease Control and Prevention.

Conflict of interest

The authors declare that they have no relevant financial interests.

Funding

This work was supported by grants from the National Natural Science Foundation of China (82000070), the National Natural Science Foundation of China-Youth Science Fund (3210040426, 32200536) and the Shanghai Rising-Star Program (21QB1400900).

Compliance with ethical standards

The data in our study were publicly available and used in a secondary analysis; thus, no additional ethical review was necessary.

Author Contributions

Zhang H and Jiang S designed and conducted the research. He F, Li Y, Hao M and Hu Z analyzed the data and performed the statistical analyses, Zhang H wrote the paper, Zhang H and Jiang S had primary responsibility for the final content, and all authors read and approved the final manuscript.

Data available

The datasets analyzed in the study are publicly available at https://www.cdc.gov/nchs/nhanes/index.htm. The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.

Sponsor’s Role

None.

Ashton, N. J., Suárez-Calvet, M., Karikari, T. K., Lantero-Rodriguez, J., Snellman, A., Sauer, M.,.. . Zetterberg, H. (2021). Effects of pre-analytical procedures on blood biomarkers for Alzheimer's pathophysiology, glial activation, and neurodegeneration. Alzheimers Dement (Amst), 13(1), e12168. doi:10.1002/dad2.12168
Banasr, M., Sanacora, G., & Esterlis, I. (2021). Macro- and Microscale Stress-Associated Alterations in Brain Structure: Translational Link With Depression. Biol Psychiatry, 90(2), 118–127. doi:10.1016/j.biopsych.2021.04.004
Bavato, F., Cathomas, F., Klaus, F., Gütter, K., Barro, C., Maceski, A.,.. . Quednow, B. B. (2021). Altered neuroaxonal integrity in schizophrenia and major depressive disorder assessed with neurofilament light chain in serum. J Psychiatr Res, 140, 141–148. doi:10.1016/j.jpsychires.2021.05.072
Besse, M., Belz, M., Folsche, T., Vogelgsang, J., Methfessel, I., Steinacker, P.,.. . Zilles, D. (2020). Serum neurofilament light chain (NFL) remains unchanged during electroconvulsive therapy. World J Biol Psychiatry, 21(2), 148–154. doi:10.1080/15622975.2019.1702717
Beurel, E., Toups, M., & Nemeroff, C. B. (2020). The Bidirectional Relationship of Depression and Inflammation: Double Trouble. Neuron, 107(2), 234–256. doi:10.1016/j.neuron.2020.06.002
Chen, M. H., Liu, Y. L., Kuo, H. W., Tsai, S. J., Hsu, J. W., Huang, K. L.,.. . Bai, Y. M. (2022). Neurofilament Light Chain Is a Novel Biomarker for Major Depression and Related Executive Dysfunction. Int J Neuropsychopharmacol, 25(2), 99–105. doi:10.1093/ijnp/pyab068
Csabai, D., Wiborg, O., & Czéh, B. (2018). Reduced Synapse and Axon Numbers in the Prefrontal Cortex of Rats Subjected to a Chronic Stress Model for Depression. Front Cell Neurosci, 12, 24. doi:10.3389/fncel.2018.00024
Cuijpers, P., Vogelzangs, N., Twisk, J., Kleiboer, A., Li, J., & Penninx, B. W. (2014). Comprehensive meta-analysis of excess mortality in depression in the general community versus patients with specific illnesses. Am J Psychiatry, 171(4), 453–462. doi:10.1176/appi.ajp.2013.13030325
Divney, A. A., Murillo, R., Rodriguez, F., Mirzayi, C. A., Tsui, E. K., & Echeverria, S. E. (2019). Diabetes Prevalence by Leisure-, Transportation-, and Occupation-Based Physical Activity Among Racially/Ethnically Diverse U.S. Adults. Diabetes Care, 42(7), 1241–1247. doi:10.2337/dc18-2432
Durrleman, S., & Simon, R. (1989). Flexible regression models with cubic splines. Stat Med, 8(5), 551–561. doi:10.1002/sim.4780080504
Gaetani, L., Blennow, K., Calabresi, P., Di Filippo, M., Parnetti, L., & Zetterberg, H. (2019). Neurofilament light chain as a biomarker in neurological disorders. J Neurol Neurosurg Psychiatry, 90(8), 870–881. doi:10.1136/jnnp-2018-320106
GBD.2019.Diseases.and.Injuries.Collaborators. (2022). Global, regional, and national burden of 12 mental disorders in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Psychiatry, 9(2), 137–150. doi:10.1016/s2215-0366(21)00395-3
Gisslén, M., Price, R. W., Andreasson, U., Norgren, N., Nilsson, S., Hagberg, L.,.. . Zetterberg, H. (2016). Plasma Concentration of the Neurofilament Light Protein (NFL) is a Biomarker of CNS Injury in HIV Infection: A Cross-Sectional Study. EBioMedicine, 3, 135–140. doi:10.1016/j.ebiom.2015.11.036
Guedes, V. A., Kenney, K., Shahim, P., Qu, B. X., Lai, C., Devoto, C.,.. . Gill, J. M. (2020). Exosomal neurofilament light: A prognostic biomarker for remote symptoms after mild traumatic brain injury? Neurology, 94(23), e2412-e2423. doi:10.1212/wnl.0000000000009577
Hodes, G. E., Kana, V., Menard, C., Merad, M., & Russo, S. J. (2015). Neuroimmune mechanisms of depression. Nat Neurosci, 18(10), 1386–1393. doi:10.1038/nn.4113
Holmes, S. E., Scheinost, D., Finnema, S. J., Naganawa, M., Davis, M. T., DellaGioia, N.,.. . Esterlis, I. (2019). Lower synaptic density is associated with depression severity and network alterations. Nat Commun, 10(1), 1529. doi:10.1038/s41467-019-09562-7
Isgren, A., Sellgren, C., Ekman, C. J., Holmén-Larsson, J., Blennow, K., Zetterberg, H.,.. . Landén, M. (2017). Markers of neuroinflammation and neuronal injury in bipolar disorder: Relation to prospective clinical outcomes. Brain Behav Immun, 65, 195–201. doi:10.1016/j.bbi.2017.05.002
Kappelmann, N., Czamara, D., Rost, N., Moser, S., Schmoll, V., Trastulla, L.,.. . Arloth, J. (2021). Polygenic risk for immuno-metabolic markers and specific depressive symptoms: A multi-sample network analysis study. Brain Behav Immun, 95, 256–268. doi:10.1016/j.bbi.2021.03.024
Khalil, M., Teunissen, C. E., Otto, M., Piehl, F., Sormani, M. P., Gattringer, T.,.. . Kuhle, J. (2018). Neurofilaments as biomarkers in neurological disorders. Nat Rev Neurol, 14(10), 577–589. doi:10.1038/s41582-018-0058-z
Kroenke, K., Spitzer, R. L., & Williams, J. B. (2001). The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med, 16(9), 606–613. doi:10.1046/j.1525-1497.2001.016009606.x
Kuhle, J., Barro, C., Andreasson, U., Derfuss, T., Lindberg, R., Sandelius, Å.,.. . Zetterberg, H. (2016). Comparison of three analytical platforms for quantification of the neurofilament light chain in blood samples: ELISA, electrochemiluminescence immunoassay and Simoa. Clin Chem Lab Med, 54(10), 1655–1661. doi:10.1515/cclm-2015-1195
Lee, S., Plavina, T., Singh, C. M., Xiong, K., Qiu, X., Rudick, R. A.,.. . Uzgiris, A. J. (2022). Development of a Highly Sensitive Neurofilament Light Chain Assay on an Automated Immunoassay Platform. Front Neurol, 13, 935382. doi:10.3389/fneur.2022.935382
Leone, M., Kuja-Halkola, R., Leval, A., D'Onofrio, B. M., Larsson, H., Lichtenstein, P., & Bergen, S. E. (2021). Association of Youth Depression With Subsequent Somatic Diseases and Premature Death. JAMA Psychiatry, 78(3), 302–310. doi:10.1001/jamapsychiatry.2020.3786
Meng, R., Yu, C., Liu, N., He, M., Lv, J., Guo, Y.,.. . Li, L. (2020). Association of Depression With All-Cause and Cardiovascular Disease Mortality Among Adults in China. JAMA Netw Open, 3(2), e1921043. doi:10.1001/jamanetworkopen.2019.21043
Miller, A. H., & Raison, C. L. (2016). The role of inflammation in depression: from evolutionary imperative to modern treatment target. Nat Rev Immunol, 16(1), 22–34. doi:10.1038/nri.2015.5
Palmer, C. R., Bellinge, J. W., Dalgaard, F., Sim, M., Murray, K., Connolly, E.,.. . Bondonno, N. P. (2021). Association between vitamin K(1) intake and mortality in the Danish Diet, Cancer, and Health cohort. Eur J Epidemiol, 36(10), 1005–1014. doi:10.1007/s10654-021-00806-9
Piercy, K. L., Troiano, R. P., Ballard, R. M., Carlson, S. A., Fulton, J. E., Galuska, D. A.,.. . Olson, R. D. (2018). The Physical Activity Guidelines for Americans. Jama, 320(19), 2020–2028. doi:10.1001/jama.2018.14854
Simrén, J., Ashton, N. J., Blennow, K., & Zetterberg, H. (2021). Blood neurofilament light in remote settings: Alternative protocols to support sample collection in challenging pre-analytical conditions. Alzheimers Dement (Amst), 13(1), e12145. doi:10.1002/dad2.12145
Spitzer, R. L., Kroenke, K., & Williams, J. B. (1999). Validation and utility of a self-report version of PRIME-MD: the PHQ primary care study. Primary Care Evaluation of Mental Disorders. Patient Health Questionnaire. Jama, 282(18), 1737–1744. doi:10.1001/jama.282.18.1737
Statistics, N. C. f. H. Centers for Disease Control and Prevention National Health and Nutrition Examination Survey. https://www.cdc.gov/nchs/nhanes/about_nhanes.htm.
Tauil, C. B., Rocha-Lima, A. D., Ferrari, B. B., Silva, F. M. D., Machado, L. A., Ramari, C.,.. . Santos-Neto, L. L. D. (2021). Depression and anxiety disorders in patients with multiple sclerosis: association with neurodegeneration and neurofilaments. Braz J Med Biol Res, 54(3), e10428. doi:10.1590/1414-431x202010428
Travica, N., Berk, M., & Marx, W. (2022). Neurofilament light protein as a biomarker in depression and cognitive function. Curr Opin Psychiatry, 35(1), 30–37. doi:10.1097/yco.0000000000000756
Wang, P., Feng, Y. B., Wang, L., Li, Y., Fan, C., Song, Q., & Yu, S. Y. (2019). Interleukin-6: Its role and mechanisms in rescuing depression-like behaviors in rat models of depression. Brain Behav Immun, 82, 106–121. doi:10.1016/j.bbi.2019.08.002
WHO. (2017). Depression and Other common mental disorders: global health estimates. Geneva: World Health Organization, 24.
Williams, M. R., Sharma, P., Macdonald, C., Pearce, R. K. B., Hirsch, S. R., & Maier, M. (2019). Axonal myelin decrease in the splenium in major depressive disorder. Eur Arch Psychiatry Clin Neurosci, 269(4), 387–395. doi:10.1007/s00406-018-0904-4
Wohleb, E. S., Franklin, T., Iwata, M., & Duman, R. S. (2016). Integrating neuroimmune systems in the neurobiology of depression. Nat Rev Neurosci, 17(8), 497–511. doi:10.1038/nrn.2016.69
Yin, W., Zhu, Y., Yang, B., Wang, F., Yin, K., Zhou, C.,.. . Yang, X. (2022). Plasma neurofilament light chain levels are associated with depressive and anxiety symptoms in Parkinson's disease. Neurol Sci, 43(4), 2839–2843. doi:10.1007/s10072-022-05914-2
Zhao, H., Mo, M., Miao, C., Li, L., Yang, H., Liu, Y., & Yang, G. (2020). Association of serum biomarker neurofilament light concentration with post-stroke depression: A preliminary study. Gen Hosp Psychiatry, 64, 17–25. doi:10.1016/j.genhosppsych.2020.01.006

Download PDF

Version 1

posted

You are reading this latest preprint version

Association of Serum Neurofilament Light Chain with Depressive Symptoms: a population-based study

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Introduction

Materials And Methods

Study population

Depressive Symptoms

Serum Neurofilament Light Chain Proteins

Covariates

Statistical Analysis

Results

Discussion

Conclusions

Declarations

References

Status:

Version 1