Vol.3, No.4, 208-213 (2013) Journal of Diabetes Mellitus
http://dx.doi.org/10.4236/jdm.2013.34032
S troke in type 2 diabetes mellitus p atient s admitted to
emergency unit in Central African country (Congo):
Preliminary findings*
Bertrand Fikahem Ellenga-Mbolla1,2#, Henri Germain Monabeka2, Paul Macaire Ossou-Nguiet2,
Gilbert Fabrice Otiobanda3, Kryste Chancel Mahoungou Guimbi3, Thierry Raoul Gombet2,4,
Suzy-Gisèle Kimbally-Kaky1,2, Benjamin Longo Mbenza5
1Department of Cardiology, University Hospital of Brazzaville, Brazzaville, Congo;
#Corresponding Author: ellenga_bertrand@hotmail.com
2Department of Medicine, Faculty of Health Sciences, Marien Ngouabi University, Brazzaville, Congo
3Intensive Care Unit, University Hospital of Brazzaville, Brazzaville, Congo
4Emergengy Unit, University Hospital of Brazzaville, Brazzaville, Congo
5Faculty of Health Sciences, Walter Sisulu University, Mthatha, South Africa
Received 5 October 2013; revised 2 November 2013; accepted 9 November 2013
Copyright © 2013 Bertrand Fikahem Ellenga-Mbolla et al. This is an open access article distributed under the Creative Commons
Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is
properly cited.
ABSTRACT
Background: The cardiovascular risk factors
including type 2 diabetes mellitus (T2DM) are a
public health problem in sub-Saharan Africa.
The aim of this study is to determine the preva-
lence and factors associated with stroke in
T2DM p atients admitted to the emergency Unit in
Brazzaville. Patients and Methods: This is the
preliminary findings of a cross-sectional study
including patients with T2DM, admitted in Emer-
gency Unit of University Hospital of Brazzaville
from January to April 2011. One hundred and
seven patients were included. Sex ratio was 1.5.
Results: The mean age was 60.3 ± 10.2 years
(range 40 - 80 years). S troke story was noted in 5
cases (4.6%). The main pathologies were meta-
bolic complications (n = 51; 47.6%) and cardio-
vascular diseases (n = 36; 33.6%), dominated by
stroke (17 cases). Epidemiological factors asso-
ciated with stroke were high standard living (OR
= 3, 95% CI: 1.02 - 8.9, p = 0.03), polypharmacy
(OR = 3.7, 95% CI: 1.27 - 10.8, p = 0.01), previous
hospitalization (OR = 3.1, 95% CI: 1.07 - 8.9, p =
0.03), and the absence of antiplatelet therapy
(OR = 4.2, 95% CI 1.2 to 15, p = 0.03). Clinical
associated factors were coma (OR = 3.3, 95% CI
1.14 to 9.6, p = 0.02) and the presence of severe
hypertension (OR = 4, 95% CI: 1.2 - 12, p = 0.02).
Finally, prognostic factors were the transfer in
intensive care unit (O R = 9.8, 95% CI: 2.7 - 34, p <
0.001). Conclusion: The f irst card iovascular co m-
plication in p atient s with T2DM admitted in emer-
gency at University Hospital of Brazzaville is
stroke. Primary prevention in high-risk patients
is still inadequate.
Keywords: Type 2 Diabetes Mellitus; Stroke;
Emergency; Brazzaville
1. INTRODUCTION
The type 2 diabetes mellitus (T2DM) is a public health
problem in sub-Saharan Africa (SSA) where its preva-
lence constantly increases and the complications of this
disease become concerned [1]. The stroke represents a
serious and frequent complication of T2DM because of
its lethality, disability and relapse [2]. T2DM patient of-
ten has multiple risk factors, which lead to frequent oc-
currence of cardiovascular complications [1]. Thus, a
comprehensive and efficient management of T2DM pa-
tients, includes the management of all risk factors. Sev-
eral predictors of stroke have been described in black
patient with T2DM. Among them, there is an increase in
blood pressure, presence of infection, poor medical care,
the addictions and occurrence of microvascular compli-
cations [3]. In SSA, the management of T2DM is diffi-
cult because of poverty, poor access to medical care, lack
of equipment and qualified healthcare professionals [4,
*Competing interests: None.
Copyright © 2013 SciRes. OPEN ACCESS
B. F. Ellenga-Mbolla et al. / Journal of Diabetes Mellitus 3 (2013) 208-213 209
5].
The aim of this study is to determine the prevalence
and factors associated with stroke in T2DM patients ad-
mitted to the emergency Unit in Brazzaville.
2. PATIENTS AND METHODS
It is a cross-sectional study, conducted from January to
April 2011 (4 months) in the emergency Unit of Univer-
sity Hospital of Brazzaville.
We included all patients admitted with a diagnosis of
T2DM, with records including at least of standard boil-
ogy, an electrocardiogram and chest X-ray. A CT scan
performed within 48 hours was required for the diagnosis
of stroke.
2.1. Variables Anal yzed
They were epidemiological (age, sex, anthropometry,
diabetes follow up, duration of diabetes, clinical history,
outpatient treatment, adherence and reference), clinical
(signs and diagnosis) and prognostic (transfer in inten-
sive care unit and death).
2.2. Definitions
Obesity was defined as a body mass index greater than
30 kg/m2. Polypharmacy was seen when taking more
than three medications in outpatient treatment. Poor
compliance was indicated by irregular intake of treat-
ment. Reference represented patients addressed by a sec-
ondary health center. The high standard of living was
based on occupation and monthly salary. Major hyper-
glycemia was defined when glycemia 5 g/L. Severe
hypertension was defined when blood pressure 180/110
mm Hg.
2.3. Statistical Analysis
Data were treated with EPI INFO 3.3.2 software (CDC
Atlanta, USA) and SPSS 10.0 for Windows (Chocago, IL,
USA). Quantitative variables were expressed as mean
and standard deviation, and qualitative staffing percent-
age. The chi-square test was used to compare qualitative
variable and ANOVA for quantitative variable. The uni-
variate risk of stroke was assessed in calculating Odds
ratio (OR) with 95% confidence intervals (95% CI).
Multivariates analyses such as logistic regression models
were used to assess the independent determinants of
stroke. The receiving operating characteristic (ROC) curve
to predict stroke was plotted. The significance level was
0.05.
3. RESULTS
One hundred and seven patients were included. They
were 65 women (60.7%), the sex ratio was 1.5. The mean
age was 60.3 ± 10.2 years (range 40 and 80 years). The
mean age by sex was 60.4 ± 9.8 years (range: 40 - 78
years) for women and 60.3 ± 11 years (range: 40 - 80
years) for men (p = 0.98). In ROC curve (Figure 1), age
>62.5 ans to predict stroke (sensitivity 64.7%, specificity
57.8%) was no significant.
3.1. Epidemiological Status
High standard living was noted in 45 cases (42%), in-
cluding 22 women (33.8%) and 23 men (54.8%). Patients
were referred from a secondary health center in 26 cases
(24.3%). The mean duration of T2DM was 7.6 ± 5.4
years (range 0 - 22 years), respectively 5 ± 7.7 years
(range 0 - 2 years) for women and 7.6 ± 6 years (range 0
- 22 years) for men (p = 0.6). Duration of diabetes >10.5
years in ROC curve (Figure 1) was predicted stroke with
sensibility 35.3% and specificity 74.4% (p = 0.048). The
history of diabetes was known in 96 cases (90%), respec-
tively 59 cases (90.8%) for women and 37 (88%) for
men. Previous hospitalization was noted in 33 cases
(30.8%). The distribution of patients by sex, duration of
diabetes, poor adherence, polypharmacy and the number
of medications is reported in Table 1.
Diabetes treatment included an oral antidiabetic alone
in 63 cases (58.9%), insulin alone in 16 cases (15%) and
both associated in 10 cases (9.3%). There wasn’t diabetes
treatment in 18 cases (16.8%).
A history of stroke was noted in 5 cases (4.6%). The
antiplatelet therapy was ongoing in 13 patients (12.1%).
3.2. Clinical Features
The clinical signs presented by patients on admission
are reported in Table 2.
Seventeen cases of cerebral ischemia divided in 16
constituted strokes and one transient ischemic attack
Figure 1. ROC curve of age and duration of diabetes to predict
stroke. For age, area under curve (AUC) = 0.622 (95% CI:
0.497 - 0.747; p = 0.112). For duration of diabetes AUC =
0.652 (95% CI: 0.538 - 0.774).
Copyright © 2013 SciRes. OPEN ACCESS
B. F. Ellenga-Mbolla et al. / Journal of Diabetes Mellitus 3 (2013) 208-213
210
Tabl e 1. Sex distribution, monitoring of diabetes, poor adher-
ence, polypharmacy, and the number of drugs being.
Females (n = 65) Males (n = 42) All (n = 107)
Diabetes monitoring
Although monitoring 26 (40) 12 (28.6) 38 (35.5)
Poorly monitored 33 (50.8) 24 (57.1) 57 (53.3)
Not followed 6 (9.2) 6 (14.3) 12 (11.2)
Poor compliance 26 (40) 15 (35.7) 41 (38.3)
Drug
None 4 (6.2) 7 (16.7) 11 (10.3)
1 24 (36.9) 14 (33.3) 38 (35.5)
2 15 (23.1) 9 (21.4) 24 (22.4)
Polypharmacy 23 (35.4) 12 (28.6) 35 (32.7)
Table 2. Clinical signs listed in the admission of patients.
n %
Major hyperglycemia
Coma
Polyuri polydipsia signs
Fever
Dyspnea
Severe hypertension
Dehydration
Hemiplegia
Cough
Psychomotor agitation
State of shock
Convulsions
Hypoglycemia
Vomiting
Abdominal pain
Brief loss of consciousness
Heart rhythm disorders
Aphasia
45
37
32
26
23
17
17
15
13
12
10
10
10
7
7
6
5
3
42.1
34.6
30
24.3
21.5
15.8
15.8
14
12.1
11.2
9.3
9.3
9.3
6.5
6.5
5.6
4.6
2.8
were registered. It represents the first condition beyond
those attributable to glycemic control (major hypergly-
cemia and hypoglycemia). All patients with cerebral
ischemia were hypertensive (p < 0.0001) and were know
diabetics (p = 0.06). The mean age of patients with stroke
was 64.2 ± 7.8 years vs 59.6 ± 10.5 for other patients (p
= 0.09). The mean duration of diabetes in patients with
stroke was 9.7 ± 4.8 years vs 7.2 ± 5.4 years for other (p
= 0.04). Diabetes follow up was regular in patients with
stroke in 12 cases (70.6%) vs. 26 cases (28.9%) for oth-
ers (p < 0.01).
The distribution of the main diseases is reported in
Table 3.
Tab le 4 shows univariates Odds of stroke. Any vari-
Table 3. Synopsis of pathologies.
n %
Metabolic Complications
Decompensation of diabetes
Diabetic coma
Hypoglycemia
Cardiovascular disease
Stroke and transcient ischemic attack
Hypertensive encephalopathy
Heart failure
Arteriopathy
Infectious diseases
Bronchopulmonary infection
Diabetic foot
Gastrointestinal infection
Urinary tractus infection
Septicemia
Myositis
Erysipelas
Other
Uremic syndrome
51
28
16
7
36
17
11
6
1
26
8
7
5
2
2
1
1
1
1
47.6
26.2
15
6.5
33.6
15.8
10.3
5.6
0.9
28
7.5
6.5
4.7
1.9
1.9
0.9
0.9
0.9
0.9
able was independent determinant of stroke in logistic
regression (Ta bl e 5 ). In ROC curve, only polypharmacy
(Figure 2) was significant to predict stroke: sensitivity
58.8%, specificity 72.2% (p = 0.043).
The average duration in emergency unit was 32.5 ± 20
hours (range 12 to 96 hours). This duration was 47.4 ±
23.5 hours for patient with stroke vs 29.6 ± 18.5 hours
for others (p < 0.001). A treatment in Intensive care Unit
of University Hospital of Brazzaville was required in 13
cases (12.1%). Three cases of deaths were recorded
(2.8%), including 2 cases of stroke (OR 11.8, 95% CI
1.01 to 139, p = 0.03).
4. DISCUSSION
4.1. Study Limitations
The study we conducted in an emergency unit, we
were able to raise the profile of T2DM and other risk
factors. It’s a preliminary finding. The best would be to
conduct a cohort study to determine the long-term prog-
nosis. Indeed, glucose and blood pressure levels, recur-
rent stroke, readmissions and mortality are important to
assess the real risk of morbidity and mortality. In addi-
tion, we had included the outcome of patients after ori-
entation in intensive care unit or department of diabetol-
ogy.
4.2. Epidemiological Aspects
The profile of our patients reflects the socio-demo-
graphic and epidemiological realities in SSA. Indeed, the
low standard of living (58%), poor reference (24.3%),
Copyright © 2013 SciRes. OPEN ACCESS
B. F. Ellenga-Mbolla et al. / Journal of Diabetes Mellitus 3 (2013) 208-213
Copyright © 2013 SciRes. OPEN ACCESS
211
Table 4. Univariates factors associated with stroke.
Stroke (n = 17)Without stroke (n = 90)OR 95% IC p-value
Epidemiological aspects
Female
Age >75 years
High standard of living
Polypharmacy
History of stroke
Anterior hospitalization
Without antiplatelet
Poor compliance
12 (70.6)
3 (17.6)
11 (64.7)
10 (58.8)
2 (11.8)
9 (52.9)
5 (29.4)
5 (29.4)
53 (58.9)
9 (10)
34 (37.8)
25 (27.8)
3 (3.3)
24 (26.7)
8 (8.9)
36 (40)
1.67
1.92
3.01
3.7
3.86
3.09
4.2
0.6
0.54 - 5.1
0.46 - 8
1.02 - 8.9
1.27 - 10.8
0.59 - 25
1.07 - 8.9
1.2 - 15
0.2 - 1.9
0.26
0.19
0.03
0.01
0.1
0.03
0.03
0.29
Clinical aspects
Obesity
Severe hypertension
Coma
Major hyperglycemia
16 (94.1)
6 (35.3)
10 (58.8)
3 (17.6)
70 (77.8)
11 (12.2)
27 (30)
42 (46.7)
4.5
4
3.3
0.24
0.5 - 36
1.2 - 12
1.14 - 9.6
0.06 - 0.91
0.06
0.02
0.02
0.01
Prognosis
Transfer in intensive care unit 7 (41.2) 6 (6.7) 9.8 2.7 - 34 <0.001
Table 5. Independents determinants of stroke by logistic regression.
Independant variables β coefficient Standard errorWald Χ2 OR (95% CI) p-value
Age (years)
Age >75 years (yes vs no)
Duration of diabetes
Anterior hospitalization (yes/no)
Coma (yes vs no)
Major glycemia (yes vs no)
Severe hypertension (yes vs no)
HSL (yes vs no)
Obesity (yes vs no)
Polypharmacy (yes vs no)
Transfert intensive care (yes/no)
Sex (male vs female)
0.026
–0.274
0.042
1.138
2.069
–14.59
1.001
1.247
1.27
0.734
14.23
–1.464
0.047
1.418
0.081
0.92
1.159
253.8
0.842
0.777
1.229
0.855
253.8
0.875
0.559
–0.382
0.519
1.237
1.784
–0.05
1.188
1.605
1.033
0.858
0.056
–1.672
1.02 (0.93 - 1.12)
0.58 (0.03 - 9)
1.04 (0.88 - 1.22)
3.12 (0.51 - 19)
7.92 (0.81 - 76)
<0.01 (0 - >105)
2.72 (0.52 - 14)
3.48 (0.75 - 15)
3.56 (0.32 - 39)
2.08 (0.38 - 11)
>105 (<105 - >105)
0.23 (0.04 - 1.28)
0.575
0.701
0.603
0.216
0.074
0.954
0.234
0.108
0.301
0.39
0.955
0.094
HSL: High standard of living.
poor adherence (64.5%), demonstrate the realities of ma-
nagement of cardiovascular disease in our environment
[6].
Among metabolic complications, glycemic imbalance
was found in 51.4% of cases and 42.1% of major hyper-
glycemia. It is recognized that the major hyperglycemia
is pejorative in the acute stage of cardiovascular compli-
cations of T2DM [7]. It plays also a role in the early de-
velopment of complications [8]. The specificity of this
major hyperglycemia in the black people is the presence
of ketosis, thus defining the ketosis prone T2DM [1,9].
The others complications is represented by hypoglyce-
mia (9.3%). It is often secondary to poor adherence to
treatment, but also the low status and illiteracy [6].
Associated pathologies, we identified cardiovascular
disease with stroke in the foreground and infectious dis-
eases. There were mainly represented by the broncho-
pulmonary infections and diabetic foot. The diabetic foot
is often mixed in our context with a tricky treatment,
sometimes leading to amputation in 20% of cases ac-
cording to Longo-Mbenza et al. [10] and 43% reported
by Monabeka et al. [11].
4.3. Stroke
Cardiovascular complications were mainly represented
by stroke (15.8%) and hypertensive encephalopathy
(10.3%). Any studied factor was independent determi-
nant of stroke in our study. Many factors were commonly
independently associated with ischemic stroke in T2DM
patients [12]. Arboix et al., identified hypertension, atrial
B. F. Ellenga-Mbolla et al. / Journal of Diabetes Mellitus 3 (2013) 208-213
212
Figure 2. ROC curve to predict stroke for polypharmacy (AUC
= 0.655; 95% CI: 0.508 - 0.803) Coma (AUC = 0.644; 95% CI:
0.497 - 0.791; p = 0.06) severe hypertension (AUC = 0.615;
95% CI: 0.457 - 0.774; p = 0.133) anterior hospitalization
(AUC = 0.631; 95% CI: 0.481 - 0.782; p = 0.07).
fibrillation, congestive heart failure, and valvumopathy
was causes of stroke [12]. The female sex was signifi-
cantly associated with stroke and mortality [12]. This
aspect was not significant in our study. According to
Mbanya et al., the prevalence of cardiovascular compli-
cations varies from 4 to 28 in diabetes, and 15% of pa-
tients with stroke have diabetes, and 5% of diabetes de-
velop stroke [1]. In our series, the number of stroke ap-
pears to be limited, because we included only patients
who had realized the CT brain scan. However, in Braz-
zaville, 50% of hypertensive emergencies are represented
by stroke [13]. The diagnosis of stroke remains difficult
in SSA. Indeed, the lack of equipment and qualified
healthcare professionals limit the diagnosis and man-
agement [5]. It is certain that the achievement of CT scan
improves the initial emergency treatment and prognosis
[14]. The poor glycemic control in patients contributes to
the early onset of complications including stroke [8], and
alters the prognosis [7] what motivated the transfer re-
suscitation in a large number in our series. In addition,
the social level of patients limit access to care [6]. In our
series, stroke is univariate factor of mortality in T2DM
patients. In addition, stroke is the leading cause of car-
diovascular emergencies at University Hospital of Braz-
zaville [15]. Given these aspects, primary prevention
using the lifestyle measures involving diet and physical
activity reduces the risk of occurrence of complications
especially stroke [16,17]. In primary prevention, the oc-
currence of diabetes is significantly lower in subjects
with regular physical activity in addition to appropriate
diet [16].
5. CONCLUSION
The risk of stroke is higher in T2DM. Given the diffi-
culties of management of cardiovascular risk factors in
sub-Saharan Africa, an active primary prevention would
not only lessen the cost of treatment, but also reduce the
occurrence of complications.
6. ACKNOWLEDGEMENTS
We thank Dr Charley Elenga Bongo for his contribution in the
manuscript translation.
REFERENCES
[1] Mbanya, J.C., Motala, A.A., Sobngwi, E., Assah, F.K. and
Enoru, S.T. (2010) Diabetes in sub-Saharan Africa. Lan-
cet, 375, 2254-2266.
http://dx.doi.org/10.1016/S0140-6736(10)60550-8
[2] Sagui, E. (2007) Stroke in sub Saharan Africa. Médecine
Tropicale, 67, 596-600.
[3] Longo-Mbenza, B., Mombo Ngimbi, R., Vangu Ngoma,
D., Mbungu Fuele, S. and Buassa-bu-Tsumbu, B. (2008)
Risk factors of stroke in blacks congolese patients with
type 2 diabetes mellitus. Annales de Cardiologie et dAn-
géologie, 57, 37-43.
[4] Gning, S.B., Thiam, M., Fall, F., Ba-Fall, K., Mbaye, P.S.
and Fourcade L. (2007) Diabetes mellitus in sub Saharan
Africa: Epidemiological aspects and management issues.
Médecine Tropicale, 67, 607-611.
[5] Chin, J.H. (2012) Stroke in Sub-Saharan Africa: An ur-
gent call for prevention. Neurology, 78, 1007.
http://dx.doi.org/10.1212/WNL.0b013e318248df95
[6] Gombet, T.R., Ellenga Mbolla, B.F., Ikama, M.S., Ekoba,
J. and Kimbally-Kaky G. (2009) Cost of emergency car-
diovascular care at the university hospital center in Braz-
zaville, Congo. Médecine Tropicale, 69, 45-47.
[7] Longo-Mbenza, B., Lelo Tshinkwela, M. and Mbuilu Pu-
kuta, J. (2008) Rates and predictors of stroke-associated
case fatality in Black Central African Patients. Cardio-
vascular journal of Africa, 19, 72-76.
[8] Tuei, V.C., Maiyoh, G.K. and Ha, C.E. (2010) Type 2
diabetes mellitus and obesity in sub-Saharan Africa. Dia-
betes and Metabolic Research and Review, 26, 433-445.
http://dx.doi.org/10.1002/dmrr.1106
[9] Djrolo, Fr., Amoussou-Guenou, D., Wanvoegbe, A. and
Glitho, S. (2011) Diabète sucré à tendance cétosique à
Cotonou. Louvain Médical, 130, 97-98.
[10] Longo Mbenza, B., Ndungo Kidinda, F., Buassa-Bu-
Tsumbu, B., Mbungu Fuele, S. and Vunga Ngoma, D.
(2009) Epidemiological and clinical profil of diabetic foot
in General Hospital of Kinshasa. Médecine dAfrique
Noire, 56, 401-409.
[11] Monabeka, H.G. and Nsakala Kibangou, N. (2001) Epi-
demiological and clinical aspects of diabetic foot (CHU
Brazzaville). Bulletin de la Société de Pathologie Exoti-
que, 94, 246-248.
[12] Arboix, A., Milian, M., Oliveres, M., Garcia-Eroles, L.
and Massons, J. (2006) Impact of female gender on prog-
nosis in type 2 diabetic patients with ischemic stroke.
European Neurology, 56, 6-12.
http://dx.doi.org/10.1159/000094249
[13] Ellenga Mbolla, B.F., Gombet, T.R., Mahoungou Guimbi,
Copyright © 2013 SciRes. OPEN ACCESS
B. F. Ellenga-Mbolla et al. / Journal of Diabetes Mellitus 3 (2013) 208-213
Copyright © 2013 SciRes. OPEN ACCESS
213
K.C., Otiobanda, G.F., Ikama, M.S., Kimbally-Kaky, G.
and Etitiele, F. (2011) Hypertensives emergencies in Uni-
versity hospital of Brazzaville (Congo). Médecine Tropi-
cale, 71, 97-98.
[14] Ossou-Nguiet, P.M., Otiobanda, G.F., Obondzo Aloba, K.,
Ellenga Mbolla, B.F., Bandzouzi-Damba, B. and Makos-
so, E. (2013) Contribution of CT scan on epidemiology
and management of stroke in a Central African country.
International Journal of Stroke, 8, E27.
http://dx.doi.org/10.1111/ijs.12118
[15] Gombet, Th., Ellenga Mbolla, B.F., Ikama, M.S., Okiemy,
G. and Etitiele, F. (2007) Cardiovascular emergencies at
Teaching Hospital of Brazzaville. Médecine dAfrique
Noire, 54, 505-511.
[16] Lloyd, C.E. and Barnett, A.H. (2008) Physical activity
and risk of diabetes. Lancet, 371, 5-6.
http://dx.doi.org/10.1016/S0140-6736(08)60044-6
[17] Feigin, V.L. and Krishnamurthi, R. (2011) Stroke preven-
tion in the developing world. Stroke, 42, 3655-3658.
http://dx.doi.org/10.1161/STROKEAHA.110.596858