Journal of the Neurological Sciences 381 (2017) 272–277

Contents lists available at ScienceDirect

Journal of the Neurological Sciences

journal homepage: www.elsevier.com/locate/jns

Zika virus disease-associated Guillain-Barré syndrome—Barranquilla, T

Colombia 2015–2016
Jorge L. Salinasa,⁎, Diana M. Walterosb, Ashley Styczynskia, Flavio Garzónb, Hernán Quijadab,
Elsa Bravoc, Pablo Chaparrob, Javier Maderob, Jorge Acosta-Reyesd, Jeremy Ledermanna,
Zuleima Artetac, Erin Borlanda, Paul Burnsa, Maritza Gonzalezb, Ann M. Powersa,
Marcela Mercadob, Alma Solanoc, James J. Sejvara, Martha L. Ospinab
a
Centers for Disease Control and Prevention, United States
b
Instituto Nacional de Salud, Colombia
c
Secretaria de Salud de Barranquilla, Colombia
d
Universidad del Norte, Barranquilla, Colombia

A R T I C L E I N F O A B S T R A C T

Keywords: Background: An outbreak of Guillain-Barré syndrome (GBS), a disorder characterized by acute, symmetric limb
Guillain-Barré syndrome weakness with decreased or absent deep-tendon reflexes, was reported in Barranquilla, Colombia, after the
Zika virus introduction of Zika virus in 2015. We reviewed clinical data for GBS cases in Barranquilla and performed a case-
Case-control studies control investigation to assess the association of suspect and probable Zika virus disease with GBS.
Colombia
Methods: We used the Brighton Collaboration Criteria to confirm reported GBS patients in Barranquilla during
October 2015–April 2016. In April 2016, two neighborhood and age range-matched controls were selected for
each confirmed GBS case-patient. We obtained demographics and antecedent symptoms in the 2-month period
before GBS onset for case-patients and the same period for controls. Sera were collected for Zika virus antibody
testing. Suspected Zika virus disease was defined as a history of rash and ≥ 2 other Zika-related symptoms (fever,
arthralgia, myalgia, or conjunctivitis). Probable Zika virus disease was defined as suspected Zika virus disease
with laboratory evidence of a recent Zika virus or flavivirus infection. Conditional logistic regression adjusted for
sex and race/ethnicity was used to calculate odds ratios (ORs) and 95% confidence intervals (CIs).
Results: We confirmed 47 GBS cases. Incidence increased with age (10-fold higher in those ≥60 years versus
those < 20 years). We interviewed 40 case-patients and 79 controls. There was no significant difference in
laboratory evidence of recent Zika virus or flavivirus infection between case-patients and controls (OR: 2.2; 95%
CI: 0.9–5.1). GBS was associated with having suspected (OR: 3.0, 95% CI: 1.1–8.6) or probable Zika virus disease
(OR: 4.6, CI: 1.1–19.0).
Conclusions: Older individuals and those with suspected and probable Zika virus disease had higher odds of
developing GBS.
Key points: We confirmed a Guillain-Barré syndrome (GBS) outbreak in Barranquilla, Colombia, during October
2015–April 2016. A case-control investigation using neighborhood controls showed an association of suspected
and probable Zika virus disease with GBS.

1. Background development of GBS include Campylobacter jejuni, Mycoplasma pneu-

Guillain-Barré syndrome (GBS) is a disease of the peripheral ner-
vous system typically characterized by acute, symmetric limb weakness
with decreased or absent deep-tendon reflexes [1]. Nearly 70% of pa-
tients with GBS report having had symptoms of an infectious illness in
the days or weeks prior to onset of neurologic illness [2]. Although no
causal agent is known, infectious agents previously linked with the
moniae, cytomegalovirus, and Epstein-Barr virus [1,2]. Globally, the
annual incidence of GBS is estimated as 1.1–1.8 cases per 100,000
population [3], but these estimates may vary depending on the regional
prevalence of infectious triggers.
Zika virus infection is usually asymptomatic. Symptomatic Zika
virus infection is known as Zika virus disease and typically manifests as
fever, rash, conjunctivitis, and myalgias/arthralgias [4]. Additionally,

⁎
Corresponding author at: Centers for Disease Control and Prevention, 1600 Clifton Road, NE, Mailstop E-10, United States.
E-mail address: jsalinas@cdc.gov (J.L. Salinas).

http://dx.doi.org/10.1016/j.jns.2017.09.001
Received 14 June 2017; Received in revised form 31 August 2017; Accepted 1 September 2017
Available online 04 September 2017
0022-510X/ Published by Elsevier B.V.
J.L. Salinas et al. Journal of the Neurological Sciences 381 (2017) 272–277

Fig. 1. Flow diagram of case and control selection for

84 (100%) Guillain-Barré syndrome clinical description —
Cases reported as GBS during Barranquilla, Colombia, Oct, 1 2015–April, 2 2016.

Oct,1st 2015–April, 2nd 2016

Excluded 10 (12%)
Recorded home address outside of
Barranquilla
74 (88%)
Cases with a home address in Barranquilla

Excluded 2 (3%)
Insufficient clinical data to determine
72 (97%) Brighton Collaboration level
Individual cases with
ICD codes for GBS
Excluded 25(35%)
Brighton Collaboration level 4, 7 (10%)
47 (65%) Brighton Collaboration level 5, 18 (25%)
Confirmed GBS case-patients

Brighton Collaboration level 1, 2 (4%)

Brighton Collaboration level 2, 11 (24%)
Brighton Collaboration level 3, 34 (72%)
Excluded 7(15%)
Moved out of Barranquilla, 4
Died, 2
40 (85%) Not found, 1
GBS case-patients included in
case-control investigation

79
neighborhood controls

Zika virus acquisition during pregnancy can lead to adverse birth out-
comes [5]. Furthermore, increased incidence of GBS has been reported
after the introduction of Zika virus in French Polynesia [6] and the
Americas [7–9]. After the introduction of Zika virus in Colombia in
2015 [10], the municipality of Barranquilla reported a large number of
Zika virus disease cases and a concomitant increase in GBS cases [11].
Because there are few comparative analytic studies linking GBS and
Zika virus beyond spatiotemporal associations or case series, we aimed
to: 1) confirm GBS status in cases given GBS diagnosis and describe
their clinical characteristics, 2) estimate the incidence of GBS during
the Zika virus outbreak, and 3) conduct a case-control investigation to
assess whether GBS case-patients had increased odds of having Zika
virus disease.
2. Methods
2.1. Ethics statement
Institutional review at CDC determined the investigation to be
public health practice, and as such did not require Institutional Review
Board approval. The Instituto Nacional de Salud de Colombia relied on
CDC's determination.
2.2. Confirmation of GBS status and clinical description
To identify patients reported with GBS (cases), we used surveillance
data from two sources. The first source was the Colombia National
Surveillance System (Sivigila), a longstanding, country-wide disease
surveillance system housed at the Instituto Nacional de Salud de
Colombia. Sivigila began recording suspected Zika virus disease in
October 2015 and suspected Zika virus-associated GBS cases starting in
December 2015. The second source was a local surveillance system
housed at the Secretaria de Salud de Barranquilla that captures ICD
code data for all clinical encounters in Barranquilla. We identified GBS
cases by their ICD-10 code (G61.0). We then obtained medical records
for all patients who were reported to Sivigila or the Secretaria de Salud
de Barranquilla as having GBS during October 1, 2015–April 2, 2016,
the period encompassing the majority of reported Zika virus illnesses.
We used a standardized medical record abstraction form to collect de-
mographic characteristics, clinical features, and antecedent illnesses in
the 2 months before onset of neurologic signs or symptoms, and any
rendered treatment for reported GBS cases. We used the Brighton
Collaboration criteria to assess whether reported cases met the defini-
tion for being a confirmed GBS case (Brighton levels 1–3) [12]. Any
patient who resided in Barranquilla and met levels 1, 2, or 3 of the
Brighton Criteria was considered to be a GBS case. Patients lacking
documentation to fulfill minimal clinical criteria were assigned a
Brighton level 4, and those with an alternative diagnosis were assigned
a level 5.
Antecedent non-neurological symptom(s)/illness were defined as
any symptom recorded to have occurred within 2 months before onset
of neurologic manifestations. We recorded any infectious disease
testing for suspected GBS infectious triggers performed during the
hospitalization for GBS which were available for review, and we re-
corded any immunizations given in the 2 months prior to neurologic
illness onset. GBS treatments recorded were intravenous im-
munoglobulin (IVIG) and plasma exchange. Hospital stay outcomes
were transcribed from discharge summaries.

273
J.L. Salinas et al.
2.3. Incidence of GBS in Barranquilla and temporal relationship with Zika
virus
To estimate the GBS incidence during the Zika virus outbreak period
in Barranquilla, we divided the number of confirmed GBS cases (i.e.,
Brighton levels 1–3) identified during October 1, 2015–April 2, 2016
(the GBS outbreak period), by the midyear 2015 population estimate
for the municipality of Barranquilla (1,218,475) [13]. To assess the
temporal relationship of confirmed GBS and Zika virus disease cases, we
juxtaposed the incidence of confirmed GBS cases over time to the
number of Zika virus disease cases reported to Sivigila during October
2015– April 2016, which was the Zika virus disease outbreak period.
We also plotted cases of dengue and Chikungunya reported to Sivigila
for comparison (Fig. 1).
2.4. Case-control investigation
In April 2016, we searched for confirmed GBS case-patients who
were Barranquilla residents according to phone numbers and to their
residency as listed in the medical record. We approached potential case-
patients for in-person interviews and obtained written informed con-
sent (for adults) or assent and parental consent for children < 18 years
of age, for inclusion into the investigation. Those who moved out of
Barranquilla, died, or were not found, were excluded. We used a
standardized questionnaire to interview case-patients and controls to
obtain demographic information, and antecedent symptoms within a 2-
month period before the onset of neurological signs/symptoms. We
visited neighboring houses to enroll two controls for each case-patient.
Controls with a neurologic illness during the evaluated period were
excluded. We flipped a coin to determine the direction of the search
(left or right) and used a random number generator to determine how
many properties [1–20] to skip to choose the first house. If the house
had any consenting person from the same age group (0–19, 20–39,
40–59, ≥ 60 years) as the case-patient, we asked for antecedent
symptoms and exposures in the 2-months period before the date of
neurologic onset of their frequency matched case-patient. Following the
same direction, a new random number was generated to identify a
second control. Serum samples were collected from case-patients and
controls for laboratory testing.
2.5. Laboratory testing and Zika virus disease case definitions
We tested the serum samples by enzyme-linked immunosorbent
assay (ELISA) for IgM antibodies against Zika and dengue viruses [14].
We determined neutralizing antibody titers against Zika and dengue
viruses using a 90% cutoff value for plaque-reduction neutralization
tests (PRNT90) [15]. We defined recent Zika virus infection as a positive
or equivocal result for either Zika or dengue virus IgM, along with a
Zika virus PRNT ≥ 10 and a dengue virus PRNT < 10. Recent flavi-
virus infection was defined as a positive or equivocal result for either
Zika or dengue viruses IgM ELISA and Zika and dengue viruses
PRNT ≥ 10 [16]. Using information from the case-control ques-
tionnaire for cases and controls, we defined suspected Zika virus disease
as a history of rash and ≥ 2 other Zika-related symptoms (fever, ar-
thralgia, myalgia, or conjunctivitis). Probable Zika virus disease was
defined as suspected Zika virus disease with additional laboratory evi-
dence of recent infection by Zika virus or a flavivirus.
2.6. Statistical analyses
To determine a possible association between GBS and a preceding
Zika virus disease, we estimated that a total of 37 case-patients and 74
controls would provide a power of 80% to detect a difference of 30% in
ZIKV prevalence between the two groups, with an alpha level of 5%. We
present categorical variables as counts and proportions, and continuous
variables as medians and ranges. We used bivariate conditional logistic
274
Journal of the Neurological Sciences 381 (2017) 272–277
regression to calculate odds ratios (OR) with 95% confidence intervals
(CI) for the association of GBS with demographics, antecedent symp-
toms and suspected or probable Zika virus disease. We adjusted ORs by
controlling for sex and race/ethnicity and report adjusted odds ratios
(AORs). Analyses were performed using Stata version 14 (College
Station, Texas).
3. Results
3.1. Confirmation of GBS status and clinical description
We identified a total of 84 patients reported to have a GBS diagnosis
reported during October 1, 2015–April 2, 2016. We excluded 10 (12%)
patients who resided outside of Barranquilla. Of the remaining 74 pa-
tients, 2 (3%) had insufficient data to assign a Brighton level. Among
the 72 patients with available data, 25 (35%) did not have a confirmed
GBS diagnosis (i.e., Brighton Collaboration criteria levels 4 [n = 7,
10%] and 5 [n = 18, 25%]). In total, 47 (65%) of the 72 Barranquilla
residents with sufficient data for classification were confirmed to meet
Brighton Collaboration criteria levels 1–3 (Fig. 1).
Among confirmed GBS patients (n = 47), most (78%) had neuro-
logic illness onset during December 2015–January 2016 (Fig. 2), the
median age was 49 years (range, 10–83), and 25 (53%) were female.
Thirty-six of the confirmed GBS case-patients (77%) had a documented
antecedent illness in the 2 months prior to neurologic symptoms onset.
The most frequently documented antecedent symptoms were fever
(n = 23; 69%), rash (n = 16; 44%), polyarthralgia (n = 12; 33%), and
myalgia (n = 8; 22%). Median time from onset of antecedent illness to
onset of neurologic symptoms was 6 days (range, 0–55). No doc-
umentation of any vaccinations in the 2 months prior to GBS onset were
found in the medical records. Laboratory testing for infectious diseases
was infrequent. Patients were tested for HIV (n = 9; 19%), cytomega-
lovirus (n = 3; 6%), syphilis (n = 3; 6%), Campylobacter jejuni (n = 1;
2%), hepatitis C virus (n = 1; 2%), leptospirosis (n = 1; 2%) and
dengue (n = 1; 2%). Only one patient (2%) had a positive result (de-
tection of dengue IgM).
Median time from onset of neurologic symptoms to hospitalization
was 9 days (range, 0–56), and median duration of hospitalization was
22 days (range, 3–102). The most frequently reported neurologic
manifestations during hospitalization were hyporeflexia (100%), lower
or upper extremity weakness (100%; 98% and 57%, respectively), and
facial weakness (38%) (Table 1). All seven patients (100%) with
available cerebrospinal fluid (CSF) data had cytoalbuminologic dis-
sociation. In the 13 (28%) patients with electrodiagnostic testing,
median time between neurologic symptoms onset and electrodiagnostic
studies was 16 days (range = 3–67). Ten patients (77%) had electro-
diagnostic studies interpreted as consistent with acute inflammatory
demyelinating polyneuropathy (AIDP), one (8%) was interpreted as
acute motor axonal neuropathy (AMAN), and the remaining two (15%)
were interpreted as indeterminate.
Twenty-seven patients received a specific treatment for GBS (57%).
Of them, 20 received intravenous immune globulin (IVIG) only (74%),
one patient received plasma exchange only (4%) and six received IVIG
followed by plasma exchange (22%). A total of 32 (68%) patients re-
quired intensive care, and 11 (23%) required mechanical ventilation. Of
36 (77%) patients with a recorded hospital outcome, 32 (88%) were
discharged home, one (3%) was discharged to a rehabilitation facility,
one (3%) remained hospitalized at the time of medical record review,
and two (6%) died in the hospital.
3.2. Incidence of GBS in Barranquilla and temporal association with Zika
virus
Using the total number of confirmed GBS cases (n = 47) among
those reported from the Municipality of Barranquilla during October 1,
2015–April 2, 2016, we estimated a GBS incidence of 3.9 cases per
J.L. Salinas et al. Journal of the Neurological Sciences 381 (2017) 272–277

Fig. 2. Epidemiologic curve of incident confirmed Guillain-Barré syndrome (GBS) cases juxtaposed with suspected and laboratory confirmed dengue, chikungunya disease (CHIK), and
Zika virus disease reported to Sivigila in Barranquilla, Colombia, 2015–2016.

Table 1 Laboratory evidence of recent Zika virus or flavivirus infection was

Clinical characteristics documented in medical records of confirmed Guillain-Barré syn- found in 21 (53%) case-patients and 30 (38%) controls (AOR: 2.2; 95%
drome patients (N = 47) — Barranquilla, Colombia, Oct, 1st 2015–April, 2nd 2016.
CI: 0.9–5.1). Case-patients had higher odds of having had suspected
Clinical characteristic No. % Zika virus disease (AOR: 3.0, 95% CI: 1.1–8.6), or probable Zika virus
disease (AOR: 4.6, CI: 1.1–19.0).
Hyporeflexia/areflexia 47 100.0%
Extremity weakness 47 100.0%
Lower extremity weakness or paresthesia 46 97.9%
4. Discussion
Upper extremity weakness or paresthesia 27 57.4%
Facial weakness 18 38.3%
Dyspnea 12 25.5% We present a clinical and epidemiological description of an out-
Dysarthria 12 25.5% break of confirmed GBS in the setting of a Zika virus outbreak in
Dysphagia 10 21.3% Barranquilla, Colombia, in 2015–2016. The estimated incidence of
Ataxia 10 21.3%
Facial paresthesia 8 17.0%
nearly 4 GBS cases/100,000 population (7.8 cases/100,000/year) ex-
Diplopia/ophtalmoplegia 6 12.8% ceeded the expected GBS incidence based on global background in-
Dysautonomia 5 10.6% cidence estimates. Our study showed that GBS cases had significantly
Total 47 100.0% higher odds of having suspected and probable Zika virus disease than
controls.
Although the baseline rate of GBS in Colombia is currently unclear,
100,000 population during the GBS outbreak period (7.8 cases/
the observed GBS incidence during this outbreak was nearly 8 cases/
100,000/year). Most cases occurred during December 2015–January
100,000/year which is higher than expected using baseline GBS in-
2016 coinciding with the outbreak of Zika virus (Fig. 2). Age-specific
cidence data from other countries (1–2 cases/100,000/year). This
incidence varied considerably, and increased with age; incidence was
finding was similar to reports from other regions in the Americas af-
1.5/100,000/year in the 0–20 years age group, 6.1 for the 20–39 years
fected by Zika virus where GBS incidence was considered to have in-
age group, 13.4 for the 40–59 years age group, and 16.2 for persons 60
creased 2–10 times over the expected baseline incidence [9]. Prior as-
and older (~ 10-fold higher in those ≥ 60 years versus those < 20
sessments of GBS have demonstrated a clear increase in incidence with
years).
age [3]; normally, persons ≥ 60 years have an approximately 2–3-fold
higher incidence compared to persons < 20 years of age. However, the
3.3. Case-control investigation higher incidence in our investigation (≥60 years versus < 20 years)
was nearly 10-fold. A similar unusual increase in incidence among older
Of the 47 confirmed GBS case-patients, four moved out of individuals was observed in investigations of Zika-associated GBS in
Barranquilla, two had died, and one was not found (Fig. 1). The re- French Polynesia and Brazil (A. Styczynski, personal communication).
maining 40 case-patients and 79 controls were included. Basic demo- Similar large increases have not been reported in cases of GBS asso-
graphic variables were similar in cases and controls. The odds of having ciated with other antecedent antigenic stimuli [17]. Zika virus infection
several Zika virus disease symptoms in the 2 months preceding GBS or the host response to Zika virus may result in a greater susceptibility
onset were higher for cases versus controls: rash (AOR: 4.4; 95% CI: to developing GBS among older individuals.
1.7–11.8), fever (AOR: 3.2; 95% CI: 1.4–7.5), myalgia (AOR: 4.1; 95% Similar to previous reports from other regions, we found a clear
CI 1.4–11.7), and arthralgia (AOR: 2.8; 95% CI 1.1–7.4) (Table 2). temporal association between peak incidence of GBS and reported Zika

275
J.L. Salinas et al. Journal of the Neurological Sciences 381 (2017) 272–277

Table 2
Associations of antecedent symptoms and Zika virus disease case definitions with Guillain-Barré syndrome, 40 case-patients and 79 controls — Barranquilla, Colombia, April 2016.

Casesa Controls ORb 95% CI AORc 95% CI

N = 40 % N = 79 %

Demographics
Age in years (median) 47 49 – – –
Male sex 21 52.2% 51 64.6% 0.6 [0.2–1.4] – –
African descent 38 95.0% 79 100.0% – – – –

Risk factors
Vaccines 0 0.0% 1 0.0% – – – –
Surgery 5 12.5% 6 7.6% 2.7 [0.5–16.0] 3.0 [0.5–18.0]

Previous illness
Viral-like syndrome
Rash 16 40.0% 11 13.9 4.3 [1.7–11.2] 4.4 [1.7–11.8]
Fever 20 50.0% 19 24.1% 3.3 [1.4–7.8] 3.2 [1.4–7.5]
Chills 12 30.0% 11 13.9% 3.1 [1.1–8.5] 3.4 [1.2–9.6]
Myalgia 16 40.0% 15 19.0% 4.1 [1.4–11.6] 4.1 [1.4–11.7]
Arthralgia 13 32.5% 13 16.5% 2.8 [1.1–7.2] 2.8 [1.1–7.4]
Conjunctivitis 10 25.0% 9 11.4% 2.6 [1.0–6.8] 2.6 [1.0–7.0]
Headache 9 22.5% 11 13.9% 1.9 [0.7–5.0] 1.8 [0.7–4.9]
Eye pain 3 7.5% 11 13.9% 0.5 [0.1–2.0] 0.5 [0.1–1.8]
Gastrointestinal syndrome
Nausea/vomiting 2 5.0% 4 5.1% 1.0 [0.2–5.5] 1.2 [0.2–6.5]
Diarrhea 7 17.5% 8 10.1% 2.2 [0.7–7.1] 2.1 [0.6–6.8]
Abdominal pain 2 5.0% 5 6.3% 0.8 [0.1–5.4] 0.7 [0.1–5.1]
Upper respiratory syndrome
Cough 5 12.5% 8 10.1% 1.3 [0.4–4.9] 1.3 [0.4–4.8]
Nasal congestion 5 12.5% 8 10.1% 1.4 [0.4–4.9] 1.4 [0.4–4.9]

Laboratory findings
Zika IgM 19 47.5% 30 38.0% 1.6 [0.7–3.6] 1.7 [0.7–3.8]
Recent Zika or flavivirus infection 21 53.0% 30 38% 2.0 [0.9–4.7] 2.2 [0.9–5.1]

Zika virus disease case definitions

Suspected 10 25.0% 8 10.1% 3.0 [1.1–8.3] 3.0 [1.1–8.6]
Probable 6 15.0% 3 3.8% 4.2 [1.1–16.8] 4.6 [1.1–19.0]

Abbreviations: OR, odds ratio; CI, confidence interval; AOR, adjusted odds ratio.
a
Column percentages reported for those with available data.
b
ORs and CIs are rounded to the nearest decimal. Highlighted ORs were statistically significant.
c
Adjusted for sex and race/ethnicity.

virus disease in Barranquilla (December 2015–January 2016) [6,9,18].
Contrary to pre-Zika GBS reports in Colombia, where respiratory or
gastrointestinal symptoms preceded the onset of GBS [19], we found
that most GBS patients during this outbreak reported an antecedent
illness consistent with Zika virus disease (rash, fever, polyarthralgia,
myalgia, and conjunctivitis), providing further support to the associa-
tion between Zika virus and GBS. GBS clinical manifestations were si-
milar to those previously reported from other regions in Colombia, with
a high proportion of facial palsy noted [8,20]. Nearly 40% of cases
lacked upper extremity weakness (Table 1) which may be due to in-
complete physical exam documentation although GBS variants with
only lower extremity weakness have been previously described
[21–23].
The pathophysiology of Zika virus-associated GBS is poorly under-
stood. Although GBS is considered a post-infectious autoimmune con-
dition, several reports have demonstrated presence of Zika virus in
serum or CSF at the onset of neurologic symptoms, postulating a
parainfectious rather than a post-infectious association between Zika
virus and GBS [8,18]. Because our evaluation occurred nearly three
months after GBS onset, we did not assess specimens for the presence of
Zika virus using molecular methods. Several studies have suggested that
Zika virus-associated GBS presents predominantly as the AMAN subtype
of GBS [6,20]. In our investigation, electrophysiologic findings were
heterogeneous, with some cases having the AIDP subtype rather than
the AMAN subtype that reportedly predominated among Zika virus-
associated GBS cases from French Polynesia and Cúcuta, Colombia
[6,20]. Some authors have proposed that the technical challenges in-
volved with electrophysiologic assessment of GBS cases in resource-
limited settings may raise questions about the association of the AMAN
and Zika virus-associated GBS cases [24].
A previous case-control study in French Polynesia [6] found a sig-
nificant difference in prevalence of Zika virus infection among case-
patients and controls based on PCR testing and Zika virus-specific an-
tibodies. In contrast, cases and controls in our study had a similar
prevalence of Zika virus infection. This discrepancy might be explained
by differences in control selection methods: while hospital-based con-
trols, who may not have had similar Zika virus exposure risk as case-
patients, were used in French Polynesia, we used age group-matched
neighborhood controls, likely leveling the exposure risk to Zika virus
among case-patients and controls in our investigation. Our findings
suggest that, although case-patients and controls had similar prevalence
of Zika virus infection, GBS case-patients had a higher prevalence of
Zika virus disease than controls. These findings are similar to results
from another case-control investigation performed in Salvador, Brazil
(A. Styczynski, personal communication). Because only a small pro-
portion of persons with Zika virus infection develop GBS [6], the higher
prevalence of Zika virus disease among GBS case-patients may have
implications in the understanding of the underlying pathophysiologic
process related to Zika virus infection and GBS [25].
Our investigation had limitations, the clinical description of GBS is
limited to the data available for review and may not be complete, the
use of Zika virus IgM serology to confirm recent Zika virus infection.
Because titers of anti-Zika virus IgM antibodies, necessary for de-
termining acute Zika virus infection, presumably decrease over time,
prevalence of Zika infection in case-patients and controls may have
been higher than the results presented. We performed a case-control

276
J.L. Salinas et al.
investigation nearly 3 months after the peak of GBS incidence which
could have introduced recall bias regarding exposure status. This time
gap also limited our ability to perform testing for other antecedent
events (i.e. Campylobacter infection) that may result in GBS. However,
we were able to confirm through both medical record review and case-
control interviews that symptoms consistent with Zika virus disease
were commonly reported among case-patients on acute presentation.
In conclusion, our investigation confirmed a strong geo-temporal
association between the Zika virus outbreak in Barranquilla and a GBS
outbreak. GBS cases had higher odds of suspected and probable Zika
virus disease compared with controls. These findings suggest an asso-
ciation between GBS and Zika virus disease.
Acknowledgments
We are grateful to the patients, family members, and neighborhood
controls that participated in this investigation. We also thank Eloina
Goenaga, Maribel Perez, Eberto Messino, Gabriel Rivera, Paula Matos,
Nineth Pabón, Sandra Silvera, Jenner Suárez, Edilma Loayza, Rosa
Aljure, Elina Ruiz, Yesenia Rambal, Angélica Carrillo, Jemima
Camargo, Margarita Villanueva, Sandra Terán, Sebastián Vergara, Ediie
Jiménez, Remberto Quintero, Yamile Sánchez, Eva Imitola, Joveth
Llanos, María González, Elena Bula, Jazmin Leon, Ingrid Cepeda, Mabel
de La Hoz, Alejandra Ortega, Arturo Coba, Dellys Guzman, Fedra
Contreras, Kupertino Bravo, Manuel Toro, Orieta Donado, Ruth Gueto,
Yulisca Reález, and Zuleima Cera from the Equipo de Vigilancia
Epidemiológica de la Secretaría de Salud de Barraquilla for collabora-
tion in patient tracing, phlebotomy support, and patient interviews. We
thank Ibrahim Zaganjor, Christina Renquist, Diana Valencia, and Carol
Rao from CDC's Zika incident management system for logistic and sci-
entific support. We would like to thank Dr. Larry Schonberger and Dr.
Ermias Belay from CDC's Division of High-Consequence Pathogens and
Pathology for valuable input and review of data analysis. We thank
Mariluz Gunturiz for laboratory assistance. We thank senior staff at
Instituto Nacional de Salud de Colombia and CDC for support for this
investigation. This study was funded by the United States Centers for
Disease Control and Prevention, the Colombian Instituto Nacional de
Salud, and the Secretaria de Salud de Barranquilla. The findings and
conclusions in this report are those of the authors and do not necessarily
represent the official position of the Centers for Disease Control and
Prevention.
References
[1] H.J. Willison, B.C. Jacobs, P.A. van Doorn, Guillain-Barre syndrome, Lancet
(London, England) (2016).
[2] A. McGrogan, G.C. Madle, H.E. Seaman, C.S. de Vries, The epidemiology of
Guillain-Barre syndrome worldwide. A systematic literature review,
Neuroepidemiology 32 (2) (2009) 150–163.
[3] J.J. Sejvar, A.L. Baughman, M. Wise, O.W. Morgan, Population incidence of
Guillain-Barre syndrome: a systematic review and meta-analysis,
Neuroepidemiology 36 (2) (2011) 123–133.
277
Journal of the Neurological Sciences 381 (2017) 272–277
[4] L.R. Petersen, D.J. Jamieson, A.M. Powers, M.A. Honein, Zika virus, N. Engl. J. Med.
374 (16) (2016) 1552–1563.
[5] L.R. Petersen, D.J. Jamieson, M.A. Honein, Zika virus, N. Engl. J. Med. 375 (3)
(2016) 294–295.
[6] V.M. Cao-Lormeau, A. Blake, S. Mons, S. Lastere, C. Roche, J. Vanhomwegen, et al.,
Guillain-Barre syndrome outbreak associated with Zika virus infection in French
Polynesia: a case-control study, Lancet (London, England) 387 (10027) (2016)
1531–1539.
[7] L.M. Araujo, M.L. Ferreira, O.J. Nascimento, Guillain-Barre syndrome associated
with the Zika virus outbreak in Brazil, Arq. Neuropsiquiatr. 74 (3) (2016) 253–255.
[8] B. Parra, J. Lizarazo, J.A. Jimenez-Arango, A.F. Zea-Vera, G. Gonzalez-Manrique,
J. Vargas, et al., Guillain-Barre syndrome associated with Zika virus infection in
Colombia, N. Engl. J. Med. (2016).
[9] T. Dos Santos, A. Rodriguez, M. Almiron, A. Sanhueza, P. Ramon, W.K. de Oliveira,
et al., Zika virus and the Guillain-Barre syndrome-case series from seven countries,
N. Engl. J. Med. (2016).
[10] O. Pacheco, M. Beltran, C.A. Nelson, D. Valencia, N. Tolosa, S.L. Farr, et al., Zika
virus disease in Colombia-preliminary report, N. Engl. J. Med. (2016).
[11] INS S, Vigilancia rutinaria Colombia, Available from: http://portalsivigila.ins.gov.
co/sivigila/reportes/pages/map_colombia.php, (2017) (1/24/2017).
[12] J.J. Sejvar, K.S. Kohl, J. Gidudu, A. Amato, N. Bakshi, R. Baxter, et al., Guillain-
Barre syndrome and Fisher syndrome: case definitions and guidelines for collection,
analysis, and presentation of immunization safety data, Vaccine 29 (3) (2011)
599–612.
[13] DANE, Proyecciones de Poblacion Nacionales y Departamentales 2006-2020
Colombia2017 [01/24/2017], Available from: http://www.dane.gov.co/index.
php/estadisticas-por-tema/demografia-y-poblacion/censo-general-2005-1.
[14] D.A. Martin, D.A. Muth, T. Brown, A.J. Johnson, N. Karabatsos, J.T. Roehrig,
Standardization of immunoglobulin M capture enzyme-linked immunosorbent as-
says for routine diagnosis of arboviral infections, J. Clin. Microbiol. 38 (5) (2000)
1823–1826.
[15] C.H. Calisher, N. Karabatsos, J.M. Dalrymple, R.E. Shope, J.S. Porterfield,
E.G. Westaway, et al., Antigenic relationships between flaviviruses as determined
by cross-neutralization tests with polyclonal antisera, J. Gen. Virol. 70 (1) (1989)
37–43.
[16] I.B. Rabe, J.E. Staples, J. Villanueva, K.B. Hummel, J.A. Johnson, L. Rose, et al.,
Interim guidance for interpretation of Zika virus antibody test results, MMWR
Morb. Mortal. Wkly Rep. 65 (21) (2016) 543–546.
[17] L.B. Schonberger, D.J. Bregman, J.Z. Sullivan-Bolyai, R.A. Keenlyside, D.W. Ziegler,
H.F. Retailliau, et al., Guillain-Barre syndrome following vaccination in the
National Influenza Immunization Program, United States, 1976–1977, Am. J.
Epidemiol. 110 (2) (1979) 105–123.
[18] E. Dirlikov, C.G. Major, M. Mayshack, N. Medina, D. Matos, K.R. Ryff, et al.,
Guillain-Barre syndrome during ongoing Zika virus transmission-Puerto Rico,
January 1–July 31, 2016, MMWR Morb. Mortal. Wkly Rep. 65 (34) (2016)
910–914.
[19] M.P. Mahecha, E. Ojeda, D.A. Vega, J.C. Sarmiento-Monroy, J.M. Anaya, Guillain-
Barre syndrome in Colombia: where do we stand now? Immunol. Res. (2016).
[20] A. Arias, L. Torres-Tobar, G. Hernandez, D. Paipilla, E. Palacios, Y. Torres, et al.,
Guillain-Barre syndrome in patients with a recent history of Zika in Cucuta,
Colombia: a descriptive case series of 19 patients from December 2015 to March
2016, J. Crit. Care 37 (2016) 19–23.
[21] T. Kimachi, N. Yuki, N. Kokubun, S. Yamaguchi, B.R. Wakerley, Paraparetic
Guillain-Barre syndrome: nondemyelinating reversible conduction failure restricted
to the lower limbs, Muscle Nerve 55 (2) (2017) 281–285.
[22] A.H. Ropper, Unusual clinical variants and signs in Guillain-Barre syndrome, Arch.
Neurol. 43 (11) (1986) 1150–1152.
[23] B. van den Berg, C. Fokke, J. Drenthen, P.A. van Doorn, B.C. Jacobs, Paraparetic
Guillain-Barre syndrome, Neurology 82 (22) (2014) 1984–1989.
[24] A. Uncini, N. Shahrizaila, S. Kuwabara, Zika virus infection and Guillain-Barre
syndrome: a review focused on clinical and electrophysiological subtypes, J.
Neurol. Neurosurg. Psychiatry (2016).
[25] W. Dejnirattisai, P. Supasa, W. Wongwiwat, A. Rouvinski, G. Barba-Spaeth,
T. Duangchinda, et al., Dengue virus sero-cross-reactivity drives antibody-depen-
dent enhancement of infection with zika virus, Nat. Immunol. 17 (9) (2016)
1102–1108.