Sage Journals: Discover world-class research

Abstract

Guillain-Barré Syndrome (GBS) is a rare and potentially life-threatening autoimmune disorder affecting the peripheral nerves. We sought to identify demographic and clinical characteristics associated with GBS onset. In this retrospective case-control study, we used national 2005 to 2020 fee-for-service Medicare claims data to identify beneficiaries with incident GBS using a chart-validated algorithm (N = 16 280), matched 1:1 with non-GBS controls (N = 16 280) by age, sex, race/ethnicity, number of preventive care visits, and year of diagnosis. We then used 2 separate modified Poisson regressions to model GBS onset as a function of preceding (≤42 days) and preexisting (>42 days) clinical conditions, while further adjusting for age, sex, race/ethnicity, socioeconomic status, disability status, rurality of residence, and year of diagnosis. Preexisting conditions associated with GBS included disorders of lipid metabolism (aOR(approx.) = 1.04, P = .03) and intestinal infection (aOR(approx.) = 1.05, P = .03). Preceding conditions associated with GBS included disorders of lipid metabolism (aOR(approx.) = 1.05, P = .01), delirium, dementia, and amnestic and other cognitive disorders (aOR(approx.) = 1.14, P = .0004), and chronic ulcer of skin (aOR(approx.) = 1.07, P = .01). In both models, GBS is positively associated with Medicare-Medicaid dual eligibility and disability status. Our hypothesis-generating findings suggest areas for further study into the mechanisms underlying GBS, raise interesting questions around the role of socioeconomic and structural factors in GBS, and demonstrate the potential of using claims data to study rare conditions.

Keywords

Guillain-Barré syndrome Medicare neuropathy big data rare disease

Introduction

Guillain-Barré Syndrome (GBS) is a rare autoimmune disorder that may be triggered by an infection or other stimulus that causes the body’s immune system to mistakenly attack the peripheral nerves, resulting in muscle weakness, sensory loss, and autonomic dysfunction.¹ It presents with wide variation in clinical severity and outcomes and atypical cases can be challenging to diagnose.² In 30% of cases, respiratory muscle weakness leads to respiratory failure and the need for mechanical ventilation.^3,4 With existing treatments, including intravenous or subcutaneous immunoglobulin (IVIg/SCIg) and plasma exchange, most individuals experience significant recovery over the course of weeks, months, or years. However, up to 14% of GBS patients remain permanently disabled,³ and the 1-year mortality rate has been estimated from 3% to 20%.^5
-8

Prior research has described the incidence and prevalence of GBS and its subtypes,^3,9,10 and has identified suspected GBS risk factors,^5,10,11 but like other rare conditions, it is challenging to study.¹² Most of the literature is based on case reports and small samples, limiting generalizability. Even estimates of GBS incidence vary widely from 0.4 to 3.3 cases per 100 000 individuals annually, with the highest rates among older adults.¹³ Consequently, much remains unknown about GBS risk factors.

While technological advances have improved clinical data collection, facilitated international efforts to study GBS, and enabled the development of prognostic models using biospecimens and clinical data,^14,15 we still do not definitively know what causes GBS. Thus, there is considerable value in leveraging large-scale administrative data to further our understanding of GBS. Specifically, the purpose of this study was to use Medicare data to identify health conditions and demographic characteristics associated with GBS onset, which may lead to more timely diagnosis, improved treatment, and better patient outcomes.

Methods

Study Population

We conducted a retrospective case-control study using fee-for-service Medicare claims data from 2005 to 2020 accessed through the CMS Virtual Research Data Center. Following the results of a prior validation study with a positive predictive value of 79.5%,¹⁶ we identified all individuals with an ICD-9-CM or ICD-10-CM diagnosis code for GBS in the primary position of an inpatient claim as an incident case of GBS. Because we were interested in examining other health conditions in the claims prior to GBS onset, we excluded individuals without at least 1 year of prior continuous enrollment in Medicare Parts A and B. Given their significant medical comorbidities, we also excluded patients with end stage renal disease (ESRD). Since we were interested in rurality of residence, we also excluded individuals whose ZIP code was associated with a PO box, US territory, or not otherwise linkable to Rural-Urban Continuum Codes (RUCCs).¹⁷

Case-Control Matching

After excluding individuals with ESRD or whose ZIP code could not be linked to RUCCs from our group of potential controls, we used a cumulative sampling approach to match each GBS case to a control without a GBS diagnosis. We matched 1:1 on exact age, sex, racial and ethnic category (using the RTI race variable, which outperforms the standard Medicare race variable),¹⁸ the year of diagnosis, and the number of preventive health visits during the 1-year prior to diagnosis. The GBS diagnosis date for cases served as the index date for matched controls (who were also required to have at least 1 year of continuous enrollment in Medicare Parts A and B prior to the index date). We defined preventive health visits using a combination of CPT codes, provider specialty codes, and place of service codes as shown in Supplemental Appendix Table 1.

Measures

In addition to our matching variables of age, sex, race/ethnicity, year of diagnosis, and number of preventive health visits, we also included disability status, Medicaid dual eligibility (as a proxy for socioeconomic status), and rurality. Using the Master Beneficiary Summary File, we identified disability status using the current and original reason for entitlement codes, and flagged individuals as dually eligible for Medicaid if they had at least 1 month of full or partial dual eligibility in the year prior to diagnosis (or index date for controls).

To capture health conditions of interest, we identified all diagnosis codes from inpatient and outpatient claims for GBS cases and their matched controls during the 1 year prior to diagnosis (index date). We classified diagnoses occurring ≤ 42 days (6 weeks) prior to GBS onset as preceding conditions (ie, those that may directly precipitate GBS onset) and diagnoses occurring > 42 days prior to onset as preexisting conditions (ie, those that may predispose someone to developing GBS even if they do not precipitate it directly). This 42-day window is based on risk interval studies of vaccine-associated GBS.^19,20

To reduce data dimensionality, we classified diagnoses using the Healthcare Cost and Utilization Project Clinical Classifications Software (CCS). We followed the approach of Hong et al, using single-level CCS categories for ICD-9-CM diagnoses and single-level CCS categories from CCS 2019.1 (beta) for ICD-10-CM diagnoses since these categories are harmonized to the ICD-9-CM categories.²¹ To further reduce the number of CCS categories included in our regression models, we focused on those categories with an overall prevalence ≥ 0.5%. To ensure that we focused on conditions plausibly associated with an increased risk of developing GBS, we compared the prevalence between GBS cases and controls and included only those CCS categories that were significantly more prevalent (P < .05) among cases versus controls.

Data Analysis

First, we calculated descriptive statistics for baseline covariates and reported the most frequent CCS categories for preexisting and preceding conditions. We used chi square tests for CCS categories to make comparisons between GBS cases and matched controls. Next, we assessed preexisting and preceding conditions separately based on CCS category and the individual components of CCS categories. Finally, we used modified Poisson regression to model GBS onset as a function of patient demographics as well as preexisting and preceding conditions identified as more prevalent among those with GBS.²² While our model specification does not directly account for the matched case-control pairs in our study design, the outcomes can be interpreted as adjusted multiplicative associations which—given the extremely rare incidence of GBS— closely approximate odds ratios and relative risks. Moreover, except for preventive visits, we included matching variables as covariates to adjust for residual confounding that might result from our unconditional model specification. To assess the impact of the 42-day (6 weeks) cut-point differentiating preexisting and preceding conditions, we conducted a sensitivity analysis in which we redefined preceding conditions as those occurring within 3 weeks (21 days) and preexisting conditions as those occurring from 1 year prior to diagnosis up to 22 days before diagnosis before re-estimating our regression models. This study was deemed exempt human subjects research, and a waiver of informed consent was approved by the University of North Carolina Institutional Review Board in October 2021.

Results

We identified 16 280 patients with a GBS diagnosis and an equal number of matched patients without a GBS diagnosis giving us a total of 32 560 cases and controls. Demographic characteristics for our sample are shown in Table 1. More than 75% of our sample were older adults between the ages of 65 and 84, a majority were male, and more than 86% were non-Hispanic white. Because of our 1:1 exact matching, there were no differences in age, sex, race/ethnicity, or index year between GBS cases and controls. There were significant differences between GBS cases and controls for Medicaid status (16% vs 24%), disability status (24% vs 48%), and rurality (24% vs 22%). Chi square results for preexisting and preceding CCS categories are available in Supplemental Appendix Tables 2–4.

Table 1.

Descriptive Statistics for Matched Cases and Controls (N = 32 560).

	Patients with GBS (N = 16 280)		Matched controls (N = 16 280)
Variables	N	%	N	%
Age
21-44	427	2.62	427	2.62
45-54	714	4.39	714	4.39
55-64	1212	7.44	1212	7.44
65-74	7342	45.10	7342	45.10
75-84	5176	31.79	5189	31.79
85+	1409	8.65	1409	8.65
Sex
Male	9045	55.56	9045	55.56
Female	7235	44.44	7235	44.44
Race
Non-Hispanic White	14 059	86.36	14 059	86.36
Black	1022	6.28	1022	6.28
Hispanic	780	4.79	780	4.79
Other/unknown	419	2.57	419	2.57
Disability
Yes	3867	23.75	7805	47.94
No	12 413	76.25	8475	52.06
Any Medicaid
Yes	2524	15.50	3582	22.00
No	13 756	84.50	12 698	78.00
Rurality of residence
Metro county	12 339	75.79	12 760	78.38
Micro county	1975	12.13	1828	11.23
Small town	1147	7.05	967	5.94
Rural	819	5.03	725	4.45
GBS year
2006	1100	6.76	1100	6.76
2007	1111	6.82	1111	6.82
2008	1063	6.53	1063	6.53
2009	1034	6.35	1034	6.35
2010	1015	6.23	1015	6.23
2011	996	6.12	996	6.12
2012	949	5.83	949	5.83
2013	1001	6.15	1001	6.15
2014	1004	6.17	1004	6.17
2015	1102	6.77	1102	6.77
2016	1132	6.95	1132	6.95
2017	1183	7.27	1183	7.27
2018	1214	7.46	1214	7.46
2019	1291	7.93	1291	7.93
2020	1085	6.66	1085	6.66

Note. Other/unknown race includes unknown, other, Asian/Pacific Islander, and American Indian/Alaska Native.

Selected results from our regression model focused on preexisting conditions (demographics and CCS categories where P < .05) are shown in Table 2. The full regression model results are available in Supplemental Appendix Table 5. Only 2 CCS categories were positively associated with GBS onset: disorders of lipid metabolism (approximately 4% higher odds) and intestinal infections (approximately 5% higher odds). Individuals dually eligible for Medicaid had about 12% higher odds of developing GBS than individuals who were only enrolled in Medicare, and individuals with a disability had about 222% higher odds of developing GBS compared to individuals without a disability. We also found that the odds of developing GBS are lower among individuals who do not live in a metro area.

Table 2.

Selected Regression Results for Preexisting Characteristics.

Variable	Adjusted OR^a	95% confidence interval	P-value
Medicaid dual-eligibility	1.12	1.07, 1.18	<.0001
Disability	3.22	3.04, 3.4	<.0001
Location (metropolitan area ref.)
Micropolitan area	0.91	0.87, 0.96	.0002
Small town	0.87	0.82, 0.93	<.0001
Rural area	0.88	0.82, 0.95	.0004
Health conditions
Uterine and cervical cancer	0.83	0.71, 0.99	.026
Cancer of lymphatic and hematopoietic tissue	0.89	0.81, 0.97	.0104
Disorders of lipid metabolism	1.04	1, 1.08	.0308
Gout and other crystal arthropathies	0.93	0.87, 0.99	.0263
Fluid and electrolyte disorders	0.94	0.9, 0.99	.0139
Paralysis	0.87	0.8, 0.95	.0013
Other nervous system disorders	0.9	0.87, 0.94	<.0001
Hypertension	0.95	0.91, 0.99	.0188
Intestinal infection	1.05	1, 1.09	.0297
Symptoms; signs; and ill-defined conditions	0.96	0.92, 0.99	.0254

Results from modified Poisson regression (without conditioning on matched pair) closely approximate both adjusted odds ratios and relative risks because GBS is an exceedingly rare outcome. While all matching variables (except preventive visits) were included as covariates to minimize residual confounding, coefficients for these variables are not shown here and should not be interpreted as population effects given our study design.

Table 3 presents selected results from our regression model focused on preceding conditions (demographics and CCS categories where P < .05). The full regression model results are available in Supplemental Appendix Table 6. As with our model for preexisting conditions, only a few of the CCS categories were associated with increased odds of GBS: disorders of lipid metabolism (approximately 5% higher odds), chronic ulcer of skin (approximately 7% higher odds), and delirium, dementia, and amnestic and other cognitive disorders (approximately 14% higher odds). Like the model for preexisting conditions, individuals who were dually eligible for Medicaid had about 9% higher odds of developing GBS than individuals who were only enrolled in Medicare, and individuals with a disability had about 131% higher odds of developing GBS compared to individuals without a disability. However, we no longer observe a significant association between rurality of residence and the odds of developing GBS.

Table 3.

Regression Results for Preceding Characteristics (6-Week Window).

Variable	Adjusted OR^a	95% confidence interval	P-value
Medicaid dual-eligibility	1.09	1.04, 1.15	.0007
Disability	2.31	2.18, 2.45	<.0001
Location (metropolitan area ref.)
Micropolitan area	0.99	0.94, 1.04	.6354
Small town	0.97	0.91, 1.03	.3226
Rural area	0.94	0.88, 1.01	.1018
Health conditions
Disorders of lipid metabolism	1.05	1.01, 1.09	.0138
Fluid and electrolyte disorders	0.93	0.89, 0.97	.001
Delirium, dementia, and amnestic and other cognitive disorders	1.14	1.06, 1.23	.0004
Headache, incl. migraine	0.95	0.9, 1	.0491
Ear disorders	0.94	0.9, 0.99	.0111
Other nervous system disorders	0.59	0.56, 0.61	<.0001
Hypertension	0.95	0.91, 0.99	.0113
Diseases of the heart	0.92	0.89, 0.96	.0001
Respiratory infections	0.87	0.83, 0.91	<.0001
Other lower respiratory disease	0.91	0.87, 0.94	<.0001
Gastrointestinal hemorrhage	0.89	0.8, 0.98	.018
Diseases of the urinary system	0.94	0.91, 0.98	.0037
Chronic ulcer of skin	1.07	1.02, 1.13	.0104
Infective arthritis and osteomyelitis	0.89	0.79, 1	.0454
Nontraumatic joint disorders	0.86	0.83, 0.89	<.0001
Systemic lupus erythematosus and connective tissue disorders	0.89	0.85, 0.92	<.0001
Symptoms; signs; and ill-defined conditions	0.68	0.65, 0.71	<.0001
Residual codes; unclassified; all E codes	0.96	0.92, 0.99	.026

Supplemental Appendix Table 7 presents the full regression model from our sensitivity analysis, in which we redefined preceding conditions as those occurring within a 3-week window prior to the diagnosis/index date (rather than 6 weeks). Our findings were very similar with respect to Medicaid dual eligibility, disability status, and rurality. Interestingly, only 1 CCS category—developmental disorders—showed a significant increase of approximately 25% in the odds of developing GBS.

Discussion

In this retrospective case-control study, we sought to identify clinical and demographic factors associated with developing GBS, while using a 6-week window before diagnosis (based on clinical guidelines related to vaccination in GBS patients) to identify potential preceding conditions. While we found few CCS categories significantly associated with GBS onset, we did identify disorders of lipid metabolism as both a potential preexisting and preceding condition. Interestingly, given concerns about vaccine-induced GBS, we did not find any association between GBS and the CCS category for immunizations and screening for infectious disease. We also identified intestinal infections as a potential preexisting condition and both chronic ulcer of skin and delirium, dementia, and amnestic and other cognitive disorders as potential preceding conditions in our main analysis, with developmental disorders a potential preceding condition in a sensitivity analysis based on a 3-week window before diagnosis.

While other GBS studies in the U.S. and France have used hospital discharge data,^23
-25 we used event-level Medicare data, which can be linked across multiple files and years, allowing us to follow our GBS cohort before and after their diagnosis, and compare them to Medicare enrollees without GBS. To our knowledge, this study is one of the first to examine the other health conditions preceding a GBS diagnosis using Medicare claims. By adapting prior work among adolescents with GBS,²⁶ we have shown that adults with incident GBS can be accurately identified in Medicare claims, which makes it possible to conduct GBS studies using a much larger sample of cases than is typical in most studies.¹⁶

Disorders of lipid metabolism include hypercholesterolemia, hyperglyceridemia, mixed hyperlipidemia, hyperchylomicronemia, and other and unspecified hyperlipidemia. These are individuals with high cholesterol levels, particularly high LDL (ie, “bad”) cholesterol, high triglycerides, or both, as well as individuals with a rare condition marked by elevated levels of chylomicrons, which transport triglycerides from the intestines throughout the rest of the body. In short, all these conditions involve high blood lipid levels. This is a potentially interesting association with GBS, given the important role of lipids in maintaining the structure and function of the peripheral nervous system, and the myelin sheath in particular.^27,28 Indeed, emerging evidence points to the importance of lipid metabolism in inherited neuropathies,²⁹ diabetic neuropathy,³⁰ and even amyotrophic lateral sclerosis (ALS).³¹ Perhaps further exploration of the role of lipid metabolism in GBS—including the initiation of medications for hyperlipidemia—is warranted.

We also identified several demographic factors associated with increased odds of developing GBS. For example, we found that dual enrollment in Medicaid and having a disability were both associated with increased odds of developing GBS in our preexisting and preceding condition models. It is likely that Medicaid enrollment is a proxy for low socioeconomic status, and that other factors associated with low socioeconomic status (eg, psychosocial stress, food insecurity and poor diet, improper food handling and storage, chronic conditions and health status, etc.) may lead to GBS onset; these socioeconomic factors warrant further study. Interestingly, an individual’s disability status had the strongest association with developing GBS. This likely reflects some factor or set of factors associated with both having a disability and the likelihood of developing GBS. Whether the mechanism is one in which some disabling condition itself is associated with developing GBS, or one in which having a disability indirectly makes individuals more likely to develop GBS (eg, living in a long-term care facility where both viral respiratory and bacterial foodborne illnesses spread more readily), is a question for future research and suggests that investigating the role of non-clinical factors in GBS is warranted.

Finally, we found that individuals who live outside of metro areas have lower odds of developing GBS in our preexisting conditions model, but this association no longer held in our preceding conditions model. Given that rural residents are at increased risk of Campylobacter jejuni infection (a known GBS risk factor) through exposure to untreated well water and interactions with poultry,³² it is interesting that we did not find rurality to be positively associated with GBS. This suggests that other structural and environmental factors may be involved. For example, the lower odds of GBS outside of metro areas may reflect underdiagnosis stemming from reduced access to health care and the likelihood that physicians practicing in rural areas will not encounter a rare condition like GBS as often as physicians in urban areas with tertiary care centers and academic medical centers and therefore may be less likely to recognize it. Alternatively, living in a large metro area may be associated with developing GBS for reasons like those we speculated about regarding persons with disabilities living in long-term care facilities, namely that they are more densely populated, resulting in individuals being exposed to viral illnesses more often that may trigger GBS but seldom appear in the claims.

There are several study limitations. First, while we used a matched case-control design, we did not use a conditional logistic model for analyses. Thus, our model uses the Poisson likelihood rather than the Bernoulli likelihood underlying conditional logistic regression. However, we included most of our matching variables as covariates to minimize residual confounding associated with our model specification. Moreover, given that GBS is exceedingly rare, our outcomes closely approximate conditional odds ratios and relative risks. Nevertheless, while we expect any resulting bias to be minimal, some residual confounding is possible. Second, if individuals do not seek treatment for an illness preceding their GBS onset, it would not be identified as a potential preceding condition. It has been previously established that GBS cases exhibit seasonality that coincides with that of respiratory infections.¹ We did not find an association with respiratory infection diagnosis and incident GBS, but many of these infections do not trigger healthcare system visits and may have been missed in this analysis. Similarly, given the exploratory nature of our study, any associations we identify between particular diagnosis codes and the onset of GBS should be considered hypothesis generating rather than evidence of a causal relationship. Third, Medicare claims also lack specific clinical data, so we were unable to model predictors of different GBS subtypes. Fourth, because we lacked access to data beyond 2020, we could not properly examine the relationship between COVID-19 and GBS, although we urge researchers with access to more recent years of Medicare data to consider adapting our methods for such studies. Finally, our study used only Medicare data; expanding these methods to evaluate populations covered by other insurance types would be worth investigating and may yield different identified risk factors.

Conclusion

As a rare condition, GBS is difficult to study. By using Medicare claims and a validated algorithm for case identification, we were able to study a large cohort of Medicare beneficiaries diagnosed with GBS and compare them to similar individuals without GBS. We identified several demographic characteristics and a few preexisting and preceding conditions that are associated with developing GBS. This demonstrates both the strengths and the limitations of using large-scale administrative claims data to study rare diseases.

Supplemental Material

sj-docx-1-inq-10.1177_00469580251400659 – Supplemental material for Characterizing Guillain-Barré Risk Factors Among Traditional Medicare Beneficiaries

Supplemental material, sj-docx-1-inq-10.1177_00469580251400659 for Characterizing Guillain-Barré Risk Factors Among Traditional Medicare Beneficiaries by Brad Wright, Samantha R. Eiffert, Abagail Cirincione, James F. Howard, Joshua Nardin and Rebecca E. Traub in INQUIRY: The Journal of Health Care Organization, Provision, and Financing

Footnotes

ORCID iD

Brad Wright

Ethical Considerations

This study was deemed exempt human subjects research and was approved by the University of North Carolina Institutional Review Board in October 2021.

Informed Consent

A waiver of informed consent was approved by the University of North Carolina Institutional Review Board in October 2021.

Author Contributions

BW, JFH, and RET conceived of and obtained funding for the study. BW and SRE obtained the data. SRE led the analysis and drafted the methods section. BW led the drafting of the rest of the manuscript. All authors assisted with interpretation of the results and made substantive contributions to the text through their reviews and edits.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institute on Minority Health and Health Disparities (R21NS119867).

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

Under the terms of our data use agreement with the Centers for Medicare and Medicaid Services, we are prohibited from sharing these data with others.

Supplemental Material

Supplemental material for this article is available online.

References

Elendu

Osamuyi

Afolayan

, et al Clinical presentation and symptomatology of Guillain-Barré syndrome: a literature review. Medicine. 2024;103(30):e38890.

Walgaard

Lingsma

Ruts

van Doorn

Steyerberg

Jacobs

. Early recognition of poor prognosis in Guillain-Barre syndrome. Neurol. 2011;76(11):968-975.

Sundar

Abraham

Gharat

Yeolekar

Trivedi

Dwivedi

. Neuromuscular respiratory failure in Guillain-Barre syndrome: evaluation of clinical and electrodiagnostic predictors. J Assoc Physicians India. 2005;53:764-768.

Lawn

Fletcher

Henderson

Wolter

Wijdicks

. Anticipating mechanical ventilation in Guillain-Barré syndrome. Arch Neurol. 2001;58(6):893-898.

Rajabally

Uncini

. Outcome and its predictors in Guillain-Barre syndrome. J Neurol Neurosurg Psychiatry. 2012;83(7):711-718.

van Doorn

. Diagnosis, treatment and prognosis of Guillain-Barré syndrome (GBS). La Presse Médicale. 2013;42(6 Pt 2):e193-e201.

Wijdicks

EFM

Klein

. Guillain-Barré syndrome. Elsevier. 2017;92:467-479.

Islam

Papri

Ara

, et al Risk factors for respiratory failure in Guillain-Barré syndrome in Bangladesh: a prospective study. Ann Clin Transl Neurol. 2019;6(2):324-332.

Shui

Rett

Weintraub

, et al Guillain-Barré syndrome incidence in a large United States cohort (2000–2009). Neuroepidemiology. 2012;39(2):109-115.

10.

Sipilä

Soilu-Hänninen

. The incidence and triggers of adult-onset Guillain-Barré syndrome in southwestern Finland 2004-2013. Eur J Neurol. 2015;22(2):292-298.

11.

Yang

Lian

Liu

Duan

R-S

. A retrospective analysis of possible triggers of Guillain-Barre syndrome. J Neuroimmunol. 2016;293:17-21.

12.

Nguengang Wakap

Lambert

Olry

, et al Estimating cumulative point prevalence of rare diseases: analysis of the Orphanet database. Eur J Hum Genet. 2020;28(2):165-173.

13.

Goodfellow

Willison

. Guillain-Barré syndrome: a century of progress. Nat Rev Neurol. 2016;12(12):723-731.

14.

Jacobs

van den Berg

Verboon

, et al International Guillain-Barré syndrome outcome study: protocol of a prospective observational cohort study on clinical and biological predictors of disease course and outcome in Guillain-Barré syndrome. J Peripher Nerv Syst. 2017;22(2):68-76.

15.

GBS/CIDP Foundation International. GBS/CIDP Patient Registry. Accessed November 24, 2025. https://gbs-cidp.iamrare.org/

16.

Eiffert

Wright

Nardin

Howard

Traub

. Evaluating algorithms for identifying incident Guillain-Barré syndrome in Medicare fee-for-service claims. Glob Epidemiol. 2024;7:100145.

17.

U.S. Department of Agriculture Economic Research Service. Rural Urban Commuting Area Codes. Accessed September 28, 2020. https://www.ers.usda.gov/data-products/rural-urban-commuting-area-codes/

18.

Eicheldinger

Bonito

. More accurate racial and ethnic codes for Medicare administrative data. Health Care Financ Rev. 2008;29(3):27-42.

19.

Burwen

Ball

Bryan

, et al Evaluation of Guillain-Barré syndrome among recipients of influenza vaccine in 2000 and 2001. Am J Prev Med. 2010;39(4):296-304.

20.

Perez-Vilar

Wernecke

Arya

, et al Surveillance for Guillain-Barré syndrome after influenza vaccination among U.S. Medicare beneficiaries during the 2017-2018 season. Vaccine. 2019;37(29):3856-3865.

21.

Hong

Garcia-Albeniz

Friesen

Foskett

Beau-Lejdstrom

. Use of clinical classifications software to address ICD coding transition in large healthcare databases analyzed via high-dimensional propensity scores. Pharmacoepidemiol Drug Saf. 2024;33(1):e5702.

22.

Zou

. A modified Poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159(7):702-706.

23.

Dharaiya

Modi Naqvi

. Guillain Barre Syndrome: Predictors of Requirement of Mechanical Ventilation (P5. 068). AAN Enterprises; 2015.

24.

Roberts

Jin

Shin

Dhamoon

. Readmissions After Guillain-Barré Syndrome: Nationally Representative Data (P2. 4-027). AAN Enterprises; 2019.

25.

Delannoy

Rudant

Chaignot

Bolgert

Mikaeloff

Weill

. Guillain-Barré syndrome in France: a nationwide epidemiological analysis based on hospital discharge data (2008–2013). J Peripheral Nerv Syst 2017;22(1):51-58.

26.

Funch

Holick

Velentgas

, et al Algorithms for identification of Guillain–Barré syndrome among adolescents in claims databases. Vaccine. 2013;31(16):2075-2079.

27.

Hornemann

. Mini review: lipids in peripheral nerve disorders. Neurosci Lett. 2021;740:135455.

28.

Kister

. Overview of myelin, major myelin lipids, and myelin-associated proteins. Front Chem. 2023;10:1041961. doi: 10.3389/fchem.2022.1041961. PMID: 36896314; PMCID: PMC9989179.

29.

López

. A new look at lipids in nerves. Nat Rev Neurosci. 2003;4(12):936-936.

30.

Afshinnia

Reynolds

Rajendiran

, et al Serum lipidomic determinants of human diabetic neuropathy in type 2 diabetes. Ann Clin Transl Neurol. 2022;9(9):1392-1404.

31.

Tracey

Kirk

Steyn

Ngo

. The role of lipids in the central nervous system and their pathological implications in amyotrophic lateral sclerosis. Semin Cell Dev Biol. 2021;112:69-81.

32.

Kuhn

Nygård

Löfdahl

, et al Campylobacteriosis in the Nordic countries from 2000 to 2015: trends in time and space. Scand J Public Health. 2020;48(8):862-869.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.09 MB