Blood glycated hemoglobin level is not associated with disease progression in amyotrophic lateral sclerosis

Abstract

Objective:

A high glycemic index and high glycemic load diet has been associated with slower progression of amyotrophic lateral sclerosis (ALS), suggesting a benefit from high blood glucose levels. We examined the association between average blood glucose level and ALS progression in two independent cohorts.

Methods:

Sporadic ALS patients enrolled in the ALS Multicenter Cohort Study of Oxidative Stress (ALS COSMOS) who completed a 3-month follow-up visit and had available blood samples were included. Hemoglobin A1c (HbA1c) was measured from whole blood collected at the 3-month follow-up. From the Pooled Resource Open-Access ALS Clinical Trials (PRO-ACT) database, we included ALS patients with one or more HbA1c measurements at enrollment and available death information. Associations between HbA1c with revised ALS functional rating scale (ALSFRS-R)/ALSFRS total score change, and tracheostomy-free survival/survival were examined in these cohorts using linear regression, linear mixed-effects models, and Cox proportional hazard models, adjusted for covariates.

Results:

In the ALS COSMOS cohort (n=193), HbA1c level was not significantly associated with the change in the ALSFRS-R total score from baseline to the 3-month follow-up (p=0.8) nor baseline to the 6-month follow-up (p=0.4). No significant association was found between HbA1c level and tracheostomy-free survival (p=0.8). In the PRO-ACT cohort (n=928), no significant association was found between HbA1c level and the rate of ALSFRS decline in the first 200 days (p=0.81 for interaction) nor between HbA1c level and survival (p=0.45).

Interpretation:

We did not find convincing evidence that mean blood glucose level is associated with disease progression among ALS patients.

Introduction

Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease involving upper and lower motor neurons.¹ We have recently demonstrated that high dietary glycemic index (GI) and glycemic load (GL) from the food frequency questionnaire (FFQ) are associated with slower functional decline and longer survival among patients with ALS, suggesting a critical role of glucose in disease progression.² We hypothesized that higher dietary GI/GL might be associated with slower functional decline and longer survival through higher mean blood glucose levels. In this study, we used hemoglobin A1c (HbA1c), a marker of average blood glucose level in the past 3 months, to investigate if mean blood glucose level is associated with dietary GI and disease progression in the ALS Multicenter Cohort Study of Oxidative Stress (ALS COSMOS).² We further utilized the historical ALS cohorts from the Pooled Resource Open-Access ALS Clinical Trials (PRO-ACT) database³ to confirm the association between HbA1c with disease progression and survival.

Methods

ALS COSMOS cohort

Sporadic ALS patients less than 18 months after symptom onset were enrolled at 16 clinical centers in the United States, as described previously.² IRB approval and written or verbal consent with an independent witness were obtained at all sites. Participants who completed the baseline FFQ and a 3-month follow-up visit with available blood samples (n=193) were included in this study. Demographic, neurological, survival, and dietary data were extracted for analysis, as previously described.² HbA1c was measured from whole blood samples collected at the 3-month follow-up visit using a turbidimetric inhibition immunoassay on the automated analyzer Cobas Integra 400 plus (Roche Diagnostics, Indianapolis, IN).⁴

PRO-ACT historical combined cohorts

The latest version of PRO-ACT database from August 2022, combined data from 16 phase II and III ALS trials and one large observational study,³ was downloaded from the PRO-ACT website (www.alsdatabase.org). We included participants who had at least one HbA1c measurement within 90 days of enrollment and had survival information ascertained (n=928). Demographics, disease history, medication, functional measurements, and survival data were extracted. HbA1c measurement that was closest to the enrollment date was used for analysis.

Statistical Analyses

In the ALS COSMOS cohort, univariable linear regression models were used to examine associations between HbA1c and baseline dietary GI or GL. The association between HbA1c at the 3-month follow-up and change in the revised ALS functional rating scale (ALSFRS-R) total score from baseline to the 3-month or 6-month follow-up were examined with multivariable linear regression models, adjusted with covariates.

In the PRO-ACT cohort, an association between baseline HbA1c and the decline in ALSFRS over time was assessed using a linear mixed-effects model with intercept, time since baseline (in months), baseline HbA1c, and interaction between time since baseline and baseline HbA1c as fixed effects, adjusted for covariates.

The associations between baseline HbA1c and tracheostomy-free survival (COSMOS)/survival (PRO-ACT) were examined with the Cox proportional hazard model, adjusted for covariates. Kaplan-Meier survival curves and the log-rank test were used to compare survival between normal vs abnormal baseline HbA1c (cut-off 5.7) groups for ALS COSMOS (baseline to 24 months) and PRO-ACT cohorts (baseline to 18 months).

Covariates included age, sex, disease duration, bulbar onset, riluzole intake, and baseline measurements (body mass index [BMI], ALSFRS/ALSFRS-R total score).

R version 4.4.0 with the lmerTest package was used for statistical analyses.

Results

A total of 193 participants from the ALS COSMOS study and 928 trial participants from the PRO-ACT database were included.[Table 1]

Table 1.

Demographic and clinical characteristics at baseline.

	ALS COSMOS (n=193)	PRO-ACT (n=928)
Age, years (SD)	60.5 (10.1)	59.6 (10.3)
Sex, male (%)	121 (63%)	532 (57%)
Race: White, Black, other (%)*	175 (90.7), 9 (4.7), 9 (4.7)	904 (97%), 18 (1.9%), 6 (0.6%)
Ethnicity, Hispanic (%)*	18 (9.3%)	13 (1.4%), missing (835)
Disease duration, months (SD)*	12 (4.6)	20.4 (12.9), missing (6)
Bulbar onset (%)	52 (27%)	252 (27.1%)
Riluzole use (%)	100 (52%)	580 (62.5%), missing (1)
Body mass index (SD)*	27 (4.5)	24.8 (4.2), missing (10)
ALSFRS-R (SD)	36.6 (6.0)	37.4 (6.2), missing (843)
ALSFRS (SD)	-	28.9 (5.7), missing (62)
Follow-up duration (months), (SD)*	13.7 (11.5)	9.4 (3.8)
Hemoglobin A1C (SD)*	5.4 (0.5)	5.4 (0.8)
Glucose intolerance (HbA1c ≥5.7) (%)*	37 (19.2%)	310 (33.4%)
Comorbid diabetes mellitus	14 (7%)	10 (1.1%), missing (27)
Insulin use	5 (2.6%)	18 (1.9%), missing (2)
Metformin use*	10 (5%)	3 (0.3%), missing (2)
Other oral agents use	2 (1%) Janumet, Actos	acarbose 4 (0.4%), gliclazide 11 (1.2%), glimepiride 1 (0.1%), glimepizide 6 (0.6%), repaglinide 1 (0.1%), missing (2)
Glycemic index (SD)	49.4 (4.6)	-
Glycemic load (SD)	88.3 (41.1)	-

^*p-value<0.05

In the ALS COSMOS cohort, HbA1c measured at the 3-month follow-up was not associated with baseline dietary GI (β=0.009, 95% CI [−0.008, 0.03], p=0.3) or GL (β=−0.0007, 95% CI [−0.003, 0.001], p=0.5). HbA1c was not associated with ALSFRS-R change from baseline to the 3-month follow-up (β=1.3, 95% CI [−0.4, 2.1], p=0.8) nor from baseline to the 6-month follow-up (β=0.7, 95% CI [−1.0, 2.5], p=0.4), adjusted for covariates. No significant association was found between HbA1c level and tracheostomy-free survival (hazard ratio 0.94, 95% CI [0.6, 1.5], p=0.8), adjusted for covariates.

In the PRO-ACT cohort, no significant interaction was found between baseline HbA1c level and the trend of ALSFRS decline from baseline to 200 days (β= −0.008, 95% CI [−0.08, 0.06], p=0.81) in a linear mixed-effect model, adjusted for covariates. Baseline HbA1c level was not associated with survival (hazard ratio 0.91, 95% CI [0.71, 1.16], p=0.45), adjusted for covariates.

The Kaplan-Meier survival curve did not demonstrate a separation between normal and abnormal HbA1c groups in either of the cohorts [Figure 1].

Figure 1.

Survival comparison between HbA1C normal vs abnormal groups among the ALS COSMOS and PRO-ACT participants. No statistical differences were found between groups in either of the cohorts.

Discussion

We did not find an association between HbA1c with GI or disease progression in the ALS COSMOS cohort. Baseline HbA1c level was also not associated with functional decline or survival among ALS patients in the PRO-ACT database. These results argue against our initial hypothesis that higher dietary GI is associated with slower functional decline and longer survival among patients with ALS through elevating mean blood glucose levels.

Our study results are consistent with prior studies that did not find a significant association between comorbid type 2 diabetes mellitus (DM) and ALS disease progression.^{5, 6} DM is a disease of inadequate control of blood glucose levels due to insulin resistance (more prominent in type 2 DM), insufficient insulin production (more prominent in type 1 DM), or a combination of both. Current literature suggests that premorbid type 2 DM lowers the risk while type 1 DM increases the risk of ALS.^7–9 However, it is not clear whether this reflects the overall earlier age of onset of type 1 DM (comorbid type 2 DM is also associated with later ALS symptoms onset)⁶ or interactions between genetics, autoimmunity, dietary patterns, or other differences in mechanisms of glucose homeostasis/utilization between type 1 and 2 DM.

Hemoglobin A1c was not significantly associated with dietary GI in the COSMOS cohort. Previous studies have shown the association between dietary GI and glycemic parameters including HbA1c among people with DM.^{10, 11} However, the effect size of dietary GI/GL on HbA1c might be small in adults without DM or obesity.¹² Therefore, the effects of dietary carbohydrates on ALS may be due to aspects of glucose homeostasis or utilization that are not reflected in circulating glucose concentrations in individuals with ALS. More studies are needed to understand the relationship between dietary carbohydrate intake and the change in blood glycemic parameters among ALS patients.

We conclude that mean blood glucose level is not associated with disease progression and HbA1c is not a sensitive marker for dietary glycemic index among predominantly Caucasian ALS patients.