Association of Polyunsaturated Fatty Acids and Clinical Progression in Patients With ALS

Abstract

Background and Objectives

Polyunsaturated fatty acids (PUFAs) have neuroprotective and anti-inflammatory effects and could be beneficial in amyotrophic lateral sclerosis (ALS). Higher dietary intake and plasma levels of PUFAs, in particular alpha-linolenic acid (ALA), have been associated with a lower risk of ALS in large epidemiologic cohort studies, but data on disease progression in patients with ALS are sparse. We examined whether plasma levels of ALA and other PUFAs contributed to predicting survival time and functional decline in patients with ALS.

Methods

We conducted a study among participants in the EMPOWER clinical trial who had plasma samples collected at the time of randomization that were available for fatty acid analyses. Plasma fatty acids were measured using gas chromatography. We used Cox proportional hazards models and linear regression to evaluate the association of individual fatty acids with risk of death and joint rank test score of functional decline and survival.

Results

Fatty acid analyses were conducted in 449 participants. The mean (SD) age of these participants at baseline was 57.5 (10.7) years, and 293 (65.3%) were men; 126 (28.1%) died during follow-up. Higher ALA levels were associated with lower risk of death (age-adjusted and sex-adjusted hazard ratio comparing highest vs lowest quartile 0.50, 95% CI 0.29–0.86, p-trend = 0.041) and higher joint rank test score (difference in score according to 1 SD increase 10.7, 95% CI 0.2–21.1, p = 0.045), consistent with a slower functional decline. The estimates remained similar in analyses adjusted for body mass index, race/ethnicity, symptom duration, site of onset, riluzole use, family history of ALS, predicted upright slow vital capacity, and treatment group. Higher levels of the n-3 fatty acid eicosapentaenoic acid and the n-6 fatty acid linoleic acid were associated with a lower risk of death during follow-up.

Discussion

Higher levels of ALA were associated with longer survival and slower functional decline in patients with ALS. These results suggest that ALA may have a favorable effect on disease progression in patients with ALS.

Introduction

Amyotrophic lateral sclerosis (ALS) is a progressive adult-onset neurodegenerative disease whose etiology is unknown and for which there is no effective treatment that can halt or reverse progression.¹ Polyunsaturated fatty acids (PUFAs) can regulate the structure and function of neurons² and play a role in neuronal survival and neuroinflammation,^3,4 all of which are mechanisms relevant for the etiology of ALS.¹ The n-3 family of PUFAs has been of particular interest for neurodegenerative diseases because it exhibits neuroprotective properties³ and has in some,^5-7 but not all,⁸ preclinical models demonstrated beneficial disease-modifying effects.

In a large prospective study documenting nearly 1000 ALS cases during follow-up, higher dietary intake of n-3 PUFAs was associated with a markedly lower ALS risk, in particular, the 18-carbon, plant-derived alpha-linolenic acid (ALA), but also the long-chained n-3 PUFAs,⁹ which was consistent with the results from 2 previously conducted case-control studies.^10,11 In a follow-up study in the same prospective cohorts, higher prediagnostic plasma levels of ALA were associated with a lower risk of ALS.¹² Although these results suggest that n-3 PUFAs play a role in the etiology of ALS, data on disease progression in patients with ALS are sparse. Higher dietary intake of n-3 PUFAs was associated with a higher percentage of predicted forced vital capacity (FVC) in a study of 302 patients with ALS,¹³ although because of the cross-sectional study design, the direction of association remains unclear. Furthermore, the study did not include analyses on individual n-3 fatty acids.

We conducted therefore a study to investigate whether plasma levels of ALA and other fatty acids were associated with survival time and functional decline during follow-up in 449 individuals with ALS who participated in the EMPOWER clinical trial.

Methods

Study Population

The EMPOWER trial has previously been described in detail.¹⁴ In short, it was a randomized, double-blind, placebo-controlled phase III trial designed to evaluate the treatment effect of dexpramipexole in participants with ALS. The participants were recruited from medical centers in Australia, Belgium, Canada, France, Germany, Ireland, the Netherlands, Spain, Sweden, the United Kingdom, and the United States. Individuals aged 18–80 years with symptom onset within 24 months of baseline and an upright slow vital capacity at screening of at least 65% of the predicted value for their age were eligible for the study. The exclusion criteria in the study included several comorbidities: significant cognitive impairment, clinical dementia, psychiatric illness, or other neurodegenerative diseases; clinically significant history of unstable or severe cardiac, oncologic, hepatic, or renal disease or other medically significant illnesses; and pulmonary disorder not attributed to ALS. In total, 942 individuals were included in the EMPOWER trial. There were no treatment differences observed in the active and placebo groups of the trial; samples from both arms were included in this study. Samples were provided by Biogen, the sponsor of the EMPOWER study.

Standard Protocol Approvals, Registrations, and Patient Consents

The study was reviewed and approved by the institutional review board at the Brigham and Women's Hospital. All participants provided written informed consent at the time of enrollment in the EMPOWER study.

Assessment of Fatty Acids

Blood samples were collected from all participants at the time of randomization and shipped by courier on the day of collection to the central laboratory where they were processed and stored at −70°C. For this study, samples were shipped to Harvard T.H. Chan School of Public Health on dry ice using overnight shipping and were immediately stored at −80°C on arrival. The samples were shipped in 2 batches: the first in 2019 (samples from 300 participants) and the second in 2021 (samples from 150 individuals).

The plasma samples were analyzed at the Nutritional Biomarker Laboratory at the Harvard T.H. Chan School of Public Health. Fatty acids in plasma were extracted into isopropanol and hexane and subsequently transmethylated. The resultant fatty acid methyl esters were then evaporated and redissolved in isooctane. Gas chromatography was used to measure known peaks by comparison with accepted standards and quantified using the area under the peak. The percentage of total area under the peak was used to determine the concentration of each individual fatty acid.¹⁵ Individual fatty acids were described as a percentage of total fatty acids measured.

The fatty acid panel included measurements of 45 fatty acids. Of these, 8 fatty acids had very low levels in most samples (mean percentage <0.01 of total fatty acids measured) and were therefore excluded from further analyses. One individual was excluded from the main analyses because of extreme ALA levels (13.2% of total fatty acids measured), which was considerably higher than the levels of the other participants with samples in the same batch (median 0.69%, interquartile range 0.54%–0.88%).

Endpoints

The endpoints included in the analyses were (1) death up to 18 months and (2) a joint rank test that considers both functional decline as change from baseline to month 12 in ALS Functional Rating Scale–Revised (ALSFRS-R) score and survival up to 12 months. The joint rank test allows for a statistical test of the treatment effect while accounting for truncation of data due to deaths.^14,16 Participants who died during follow-up were ranked according to time to death (shorter time to death ranked worse), while those who survived until 12 months were ranked according to change in ALSFRS-R scores (larger decline in ALSFRS-R score ranked worse). Those who survived ranked above those who died. ALSFRS-R at 12 month was used because estimation of change in ALSFRS-R was not available for all participants after month 12, and this was also the primary endpoint in EMPOWER.¹⁴ Participants who withdrew from the study before month 12, and thus were missing ALSFRS-R score at month 12, were excluded from the joint rank test.

Statistical Analysis

We log transformed fatty acid levels to improve normality and standardized the levels (mean = 0, SD = 1) within batch. This was performed to account for potential differences in the measurements of the fatty acids for each batch. Fatty acid concentrations below the level of detection were imputed with half of the minimum observed value for each fatty acid before log transformation.

We used Cox proportional hazards models to estimate hazard ratios (HRs) and 95% CIs for the association between individual fatty acids and death during follow-up and linear regression to estimate the association between individual fatty acids and least-squared mean joint rank test score. We modeled fatty acids as continuous variables (per 1 SD increase) to maximize power if linear and as a categorical variable (quartiles) to explore possible nonlinear associations. For the categorical analyses, we categorized participants into batch-specific quartiles. To test for a linear trend across the quartiles, we assigned the median value to each quartile and modeled this as a continuous variable. To evaluate the influence of potential confounders, we first fitted models adjusted for age, sex, and baseline ALSFRS-R and then further adjusted for body mass index (BMI), race/ethnicity, symptom duration, site of onset, riluzole use, family history of ALS, predicted upright slow vital capacity, and treatment arm in the trial. We used separate models for each fatty acid.

Cumulative incidence curves were estimated with the Kaplan-Meier estimator to visualize the risk of death during follow-up according to ALA levels.¹⁷ The plot was curtailed when less than 10% of the participants were still in follow-up to avoid large variation and uncertainty in the estimation when the risk set is small.¹⁸

All analyses were conducted using R, version 4.0.3 (The R Foundation, Vienna, Austria).¹⁹ Tidyverse R packages were used for data import, manipulation, and visualization.²⁰ Survival analyses were conducted using the survival R package.²¹ Cumulative incidence curves were visualized using the survminer R package.²² All tests are 2-tailed. The α-level was set at 0.05.

Data Availability

The data sets analyzed in this study are not publicly available because of restricted access, but further information about the data sets is available from the corresponding author on reasonable request.

Results

Plasma samples were available from 450 participants who participated in the EMPOWER trial. One participant was excluded because of a questionable lipid profile, leaving 449 in the current analyses. Participant characteristics according to baseline plasma levels of ALA, the primary PUFA of interest in our study, are provided in Table 1. Overall, the characteristics were similarly distributed across the quartiles of ALA. There were a higher number of non-White participants in quartile 4 compared with the other quartiles. Furthermore, there were a lower number of participants with a family history of ALS in quartile 4 compared with the other quartiles. Compared with the first and fourth quartiles, the third quartile had a higher proportion of patients with bulbar onset, while the second quartile had a lower proportion. The baseline characteristics of the participants in our study closely resembled those of the complete study population in the EMPOWER trial.

Table 1.

Participant Characteristics by Quartiles of ALA Measured at Randomization

Time to Death With 18 Months of Follow-up

All participants with available plasma fatty acid levels (n = 449) were included in the survival analyses. Among these, 126 participants (28.1%) died during follow-up. A lower number of participants in the top quartile of ALA (n = 21, 18.9%) died during follow-up compared with quartile 1 (n = 37, 32.7%), quartile 2 (n = 31, 27.4%), and quartile 3 (n = 37, 33.0%), as illustrated in Figure 1. In the analysis adjusted for age, sex, and baseline ALSFRS-R score, the HR for death comparing the highest and the lowest quartile of ALA was 0.50 (95% CI 0.29–0.86, p-trend = 0.041; Table 2). The estimate remained similar in a multivariable analysis additionally adjusted for BMI, race/ethnicity, symptom duration, site of onset, riluzole use, family history of ALS, predicted upright slow vital capacity, and treatment arm in the trial (HR comparing highest vs lowest quartile 0.55, 95% CI 0.31–0.98, p-trend = 0.120). The risk estimates for ALA were similar in each treatment arm (p for interaction >0.05). Among the other n-3 and n-6 PUFAs, higher plasma levels of eicosapentaenoic acid (EPA; multivariable-adjusted HR comparing top vs bottom quartile 0.45, 95% CI 0.26–0.79, p-trend = 0.008) and linoleic acid (LA; multivariable-adjusted HR comparing top vs bottom quartile 0.54, 95% CI 0.31–0.93, p-trend = 0.048) were associated with a lower risk of death during follow-up. In exploratory analyses, higher levels of the monounsaturated fatty acid palmitic acid were associated with a higher risk of death in multivariable-adjusted analyses (Figure 2). None of the other fatty acids were nominally statistically significantly associated with the risk of death during follow-up (p > 0.05).

Figure 1. Cumulative Incidence of Death According to Quartiles of ALA Measured at the Time of Randomization.

ALA = alpha-linolenic acid.

Table 2.

Hazard Ratio for Death at 18 Months According to PUFA Levels Measured at Randomization

Figure 2. HR for Death From Amyotrophic Lateral Sclerosis According to Plasma Fatty Acid Levels.

Risk estimates obtained from multivariable Cox proportion hazards models adjusted for age, sex, baseline ALSFRS-R score, body mass index, race/ethnicity, symptom duration, site of onset, riluzole use, family history of ALS, predicted upright slow vital capacity, and treatment arm in the trial. The dotted line represents nominal significance (α = 0.05). ALA = alpha-linolenic acid; ALS = amyotrophic lateral sclerosis; ALSFRS-R = ALS Functional Rating Scale–Revised; EPA = eicosapentaenoic acid; HR = hazard ratio; LA = linoleic acid; PA = palmitic acid.

Joint Rank Test of Functional Decline and Death at 12 Months

A total of 423 participants with available plasma fatty acid levels were included in the joint rank test analyses, as 326 participants had available ALSFRS-R scores at month 12, while 97 participants died before month 12 in the trial. In a model adjusted for age, sex, and baseline ALSFRS-R score, higher ALA levels were associated with higher rank, which indicates slower functional decline, when ALA was modeled as a continuous variable (difference in joint rank test score according to 1 SD increase 10.7, 95% CI 0.2–21.1, p = 0.045; Table 3). The least-squared mean joint rank test score for participants in quartile 4 of ALA was 24.3 points (95% CI −5.0 to 53.5) higher than in quartile 1, but the difference was not significant. The estimates remained similar in multivariable-adjusted analyses additionally adjusted for BMI, race/ethnicity, symptom duration, site of onset, riluzole use, family history of ALS, predicted upright slow vital capacity, and treatment arm in the trial (Table 3). The risk estimates for ALA were similar in each treatment arm (p for interaction >0.05). None of the other n-3 or n-6 PUFAs were nominally statistically significantly associated with the joint rank test score at 12 months (Table 3).

Table 3.

Mean Joint Rank Score at 12 Months According to PUFA Levels Measured at Randomization

Discussion

In this study, we found that higher plasma levels of ALA measured at the time of randomization in the EMPOWER trial were associated with longer survival and slower functional decline during follow-up. Higher levels of the long-chain n-3 fatty acid EPA and the n-6 fatty acid LA were also associated with a lower risk of death during follow-up, and LA was also associated with slower functional decline. These results suggest that specific PUFAs, in particular ALA, may have a favorable effect on disease progression in patients with ALS.

Previous data on PUFAs and disease progression in ALS are limited to a cross-sectional study in which dietary factors were correlated with measures of disease progression a median of ∼1 year after symptom onset.¹³ In the main analyses, intake of a group of healthy micronutrients, including n-3 fatty acids, was associated with higher ALSFRS-R score and higher percentage of predicted FVC. On the other hand, strong data supports that a higher dietary intake and increased plasma levels of PUFAs, particularly the plant-derived n-3 ALA, are associated with a lower risk of ALS.^9,12 The combined evidence from analyses on dietary intake assessed by food frequency questionnaires, which provide a valid estimation of long-term dietary intake,²³ and biochemical indicators of dietary intake directly measured in plasma samples adds strengths to these results. Our findings that higher plasma levels of ALA are also associated with longer survival and slower functional decline in a large cohort of patients with ALS contribute to support its neuroprotective effect.

PUFAs can regulate and modulate pathways relevant to ALS. Most previous studies have focused on the effects of the long-chained n-3 fatty acid docosahexaenoic acid (DHA) and EPA, both of which have demonstrated neuroprotective and anti-inflammatory effects in animal models.³ EPA supplementation can attenuate activation of microglia cells,²⁴ which contribute to inflammation-mediated neurotoxicity,²⁵ while DHA can promote a switch of microglia to a more anti-inflammatory phenotype,²⁶ demonstrating the modulating effect on neuroinflammation, which plays an important role for neurodegenerative diseases,^27,28 including ALS.²⁹ Less is known about the direct effects of ALA. The association between ALA and ALS could be mediated by EPA because ALA can be converted to EPA.³ However, only a small fraction is metabolized,³ suggesting that ALA could itself have direct effects. ALA supplementation was associated with lower levels of inflammatory markers, such as interleukin (IL)-6, IL-1β, and tumor necrosis factor in some,^30,31 but not all,³² clinical trials. These cytokines play an important role in neuroinflammation and neurodegeneration³³ and have been reported to be elevated in patients with ALS compared with control participants.³⁴

Our study has several strengths. We included a large number of patients with ALS from a well-defined cohort with similar disease duration and progression. As the study population was part of a clinical trial, we had access to comprehensive data on demographic characteristics and clinical variables relevant for the diagnosis, prognosis, and progression of ALS. This allowed us to evaluate whether there were any systematic differences in participants with different PUFA levels and account for these differences in multivariable models. As patients were followed closely with repeated clinical visits during follow-up, we could conduct longitudinal analyses of disease progression over time, and with detailed data on time of death, study withdrawal, and discontinuation of participation for each patient with ALS, we could account for time-at-risk in detail. Our study has also some weaknesses. Fatty acid concentrations in plasma may not always reflect dietary intake because the levels could also be affected by genetic, behavioral, and metabolic factors.³⁵ Still, levels of fatty acids that are not synthesized endogenously, such as n-3 PUFAs, are more likely to reflect dietary intake than other fatty acids.³⁶ Furthermore, the consistency between studies finding an inverse association between PUFA intake, both assessed by dietary intake and plasma concentration, and ALS risk,^9,12 suggests that some PUFAs affect mechanisms relevant to ALS. We did not have access to data on the overall diet of the participants and could therefore not account for differences in other nutrients, supplements (e.g., n-3 fatty acids), or total caloric intake, which has been associated with survival time in ALS.³⁷ The low mortality rate over the follow-up period may have limited the statistical power of our analyses, especially for other fatty acids than ALA. Moreover, the highly selective nature of participants in clinical trials may restrict the generalizability of our findings to the broader ALS population. Additional limitations include the lack of data on genetic risk factors, such as repeat expansions in the C9orf72 gene, or cognitive and behavioral status, which may influence survival and could confound our analyses. Finally, as is inherent to any observational study, the results may be affected by residual or unmeasured confounding factors.

In summary, we observed that higher levels of ALA were associated with a longer survival and slower functional decline in a cohort of patients with ALS. Furthermore, EPA and LA were also associated with better disease outcomes. These results indicate that specific PUFAs, in particular ALA, may have a beneficial effect in patients with ALS.

Glossary

ALA

alpha-linolenic acid

ALS

amyotrophic lateral sclerosis

ALSFRS-R

ALS Functional Rating Scale–Revised

BMI

body mass index

DHA

docosahexaenoic acid

EPA

eicosapentaenoic acid

FVC

forced vital capacity

HR

hazard ratio

IL

interleukin

LA

linoleic acid

PUFA

polyunsaturated fatty acid

Appendix. Authors

Study Funding

Study funded by a grant from the ALS Association awarded to Alberto Ascherio.

Disclosure

K. Bjornevik reports no disclosures. M. Cortese has received a speaker honorarium from Roche. J.D. Furtado is currently an employee of and holds stock/stock options in Biogen. S. Paganoni reports research grants from Amylyx Pharmaceuticals, Revalesio Corporation, UCB, Biohaven, Clene, Seelos, and Prilenia and consulting fees from Orion, Cytokinetics, and Medscape. M.A. Schwarzschild received payment for consultative service from Denali, Lilly, Prevail, and the Parkinson Study Group (for advising Bial, Biogen, Chase Therapeutics, and Partner Therapeutics). M.E. Cudkowicz provided consultation to Avexis, Sunovion, Takeda, Biogen, Anelixis, Aclipse, Cytokinetics, Disarm, ALS Pharma, RRD, Immunity Pharma, Helixsmith, Denali, Wave, Orion, Transposon, QurAlis, Faze, Regeneron, AB Sciences, Lilly, MTPC, Locust Walk, and NeuroSense. She is on the board of directors for Praxis. A. Ascherio reports no disclosures. Go to Neurology.org/N for full disclosures.