Abstract
Background:
Acute infections have been posited as potential precipitants or triggers of the occurrence of stroke among adults with traditional vascular risk factors. We evaluated associations between stroke occurrence and reported febrile illness within 4 weeks (potential antecedent infections) among West Africans.
Methods:
The Stroke Investigative Research and Educational Network (SIREN) is a multicenter, case-control study involving 15 sites in Ghana and Nigeria. Cases include adults aged ≥18 years with radiologically confirmed strokes. Controls were stroke-free adults matched with cased by age, gender and ethnicity. Detailed evaluations for vascular, lifestyle and psychosocial factors were performed. Participants were asked for evidence of any febrile illness within the past 4 weeks. We used conditional logistic regression to estimate adjusted odds ratios (aOR) with 95% Confidence Interval.
Results:
Among 3588 stroke cases recruited in Ghana and Nigeria between August 2014 and July 2018, 363 cases (10.1%) reported having a febrile illness within the 4 weeks prior to stroke occurrence. Having an antecedent infection was associated with stroke occurrence with an unadjusted OR of 1.19 (1.00–1.51) but aOR of 0.83 (0.59–1.17) upon adjusting for traditional vascular risk factors. Stress, aOR of 4.69 (2.59–8.50) and consumption of green vegetables 2.27 (1.35–2.85) were associated with antecedent febrile illness.
Conclusion:
1 in 10 stroke cases reported antecedent history of febrile illness prior to occurrence of stroke but no independent association was observed in this study. Infectious exposures may be important triggers of cardiovascular events requiring further exploratory studies to better understand the role of this emerging risk factor.
Keywords: Infection, Stroke trigger, Febrile illness, Stroke risk, Africa
1. Introduction
Although stroke is the second leading cause of mortality globally, 90% of all strokes are due to cardio-metabolic and behavioral risk factors which are amenable to modification [1,2]. While a great majority of population harbor these modifiable risk factors, the exact precipitants of acute stroke events in individuals at risk for cardiovascular events remain to be elucidated. An emerging area of stroke epidemiology investigation entails the determination of stroke triggers or precipitants [3].
One such stroke trigger is recent infection. Evidence from prospective studies suggest that recent hospitalizations with acute infection is a potential trigger for stroke [4–6]. These studies have shown that the association between infection and stroke occurrence shows a time-dependent relation with increasing stroke risk as the time interval after hospitalization decreased. Furthermore, while all infection types are associated more strongly with subsequent acute ischemic stroke than hemorrhagic stroke, the greatest trigger is urinary tract infection [7]. The association between recent infection and stroke occurrence has hitherto not been studied in sub-Saharan Africa, a region at the epicenter of a double burden of infectious diseases and non-communicable cardiovascular diseases exemplified by rapidly rising indices of stroke in recent decades [8–11]. Our objectives for this paper are to assess the frequency and factors associated with reported febrile illness in the preceding 4 weeks before stroke onset among stroke cases and assess associations between febrile illness and stroke type and by etiologic subtypes in West Africa.
2. Methods
2.1. Study design
The Stroke Investigative Research and Educational Network study is a large multicenter case-control study involving 15 sites in northern and southern belts of Nigeria and Ghana. The study protocol has been previously published [12]. In brief, stroke cases were consecutive consenting (in unconscious or aphasic patients, consent was obtained from next of kin) adults aged ≥18 years with clinical stroke presenting within 8 days of current symptom onset or ‘last seen without deficit’. We confirmed all stroke diagnosis using either CT or MRI scan typically within 10 days of symptom onset. Controls were consenting stroke-free adults, mostly from the communities in the catchment areas of the SIREN hospitals where cases were recruited. Stroke-free status was confirmed with the 8-item questionnaire for verifying stroke-free status (QVSFS) which has 98% negative predictive value [13].
Controls were matched to cases by age (+/− 5 years), sex and ethnicity in a ratio of 1:1 to minimize the potential confounding effect of these variables on the relationship between stroke and the main environmental risk factors. Hospital-based controls may be attendants or relatives of another (non-stroke) patient, or patients admitted to the hospital or visiting the hospital for conditions or procedures not related to stroke or TIA. Ethical approval was obtained from all study sites and informed consent was obtained from all subjects [12].
2.2. Stroke phenotyping
Stroke diagnosis and phenotyping were based on clinical evaluation and brain neuroimaging (CT or MRI), ECG, transthoracic echocardiography, and carotid Doppler ultrasound performed according to standardized protocols (SOP) at each site. Presumed etiological sub-types of ischemic stroke were defined etiologically using the A-S-C-O-D classification into A: Atherosclerosis, S: Small-vessel occlusion, C: Cardiac pathology, O: Other causes and D: dissection [14] and intracerebral hemorrhage was classified etiologically into Structural, Medication-related, Amyloid angiopathy, Systemic/other disease, Hypertension and Undetermined causes (SMASH-U) [15].
2.3. Definition of risk factors
We collected basic demographic and lifestyle data including, socioeconomic status, cardiovascular risk profile, dietary patterns, routine physical activity, stress using a validated INTERSTROKE instrument, depression, cigarette smoking, and alcohol use [16].
Antecedent febrile illness: Antecedent febrile illness was defined as any acute febrile illness within the past 4 weeks prior to the onset of stroke (for stroke cases) or within the past 4 weeks (for stroke free controls). Acute febrile illness was defined as a patient with fever of 38 °C or higher or history of fever that persisted for 2–7 days. We did not ascertain the cause of febrile illness.
Hypertension: Blood pressure was recorded at baseline and daily for 7 days. Hypertension was defined as a sustained elevation of blood pressure ≥ 140/90 mmHg > 72 h after stroke, a premorbid history of hypertension, use of antihypertensive drugs before stroke or > 72 h after stroke onset. Adjustments to systolic BP (SBP) based on reported associations between pre-morbid BP and acute post-stroke BP in the Oxford Vascular Study (OXVASC) were also applied in sensitivity analyses [17]. Definition of hypertension in controls was self-reported history of hypertension or use of antihypertensive drugs or average of 3 recorded BP at first clinical encounter ≥140/90 mmHg [16].
Diabetes mellitus was defined based on history of diabetes mellitus, use of medications for DM, an HBA1c > 6.5% or a fasting blood glucose (FBG) levels > 7.0 mmol/l at first encounter in controls or measured after the post-acute phase in cases due to the known acute transient elevation of glucose as a stress response after stroke [18].
Dyslipidemia was defined as fasting total cholesterol ≥5.2 mmol/l, HDL cholesterol ≤1.03 mmol/l, triglyceride ≥1.7 mmol/l or LDL cholesterol ≥3.4 mmol/l according to NCEP guidelines [19] or use of statin prior to stroke onset. Based on distribution of the LDL/HDL ratio in the present study, the LDL/HDL ratio was dichotomized using the lowest two tertiles (≤1.97 and 1.98–2.95) as normal versus highest tertile (≥2.96) as high.
Cardiac disease was defined after evaluation by study cardiologists based on history or current diagnosis of atrial fibrillation, cardiomyopathy, heart failure, ischemic heart disease, rheumatic heart disease or valvular heart diseases.
For obesity, we assessed both waist-to-hip ratio (WHR) and body-mass index. Subjects were classified individually either using the WHO guidelines cutoffs of 0.90 (men) and 0.85 (women) for WHR or 30 kg/m2 for BMI (Obesity) [20].
Individuals were classified as physically active if they were regularly involved in moderate exercise (walking, cycling, or gardening) or strenuous exercise (jogging, football, and vigorous swimming) for 4 h or more per week [16].
Dietary history included regularity of intake of food items such as meat, fish, green leafy vegetables, addition of salt at table, nuts, sugar and other local staple food items. Regular intake was defined as intake on daily, weekly or at least once monthly versus none in a month.
Alcohol use was categorized into current users (users of any form of alcoholic drinks) or never/former drinker while alcohol intake was categorized as low drinkers (1–2 drinks per day for female and 1–3 drinks per day for male) and high drinker (> 2 drinks per day for female and > 3 drinks per day for male. 1 drink or 1 unit of alcohol = 8 g of alcohol) [16].
Smoking status was defined as current smoker (individuals who smoked any tobacco in the past 12 months) or never/former smoker [16].
We adapted the measures of psychosocial stress and depression in the INTERSTROKE study for assessment of psychosocial risk factors [13]. Psychosocial stress combined measures of stress at home/work (e.g. irritability, anxiety or sleeping difficulties) and life events, experienced in the 2 weeks preceding the stroke. Depression combined depressed mood and a checklist of other depression symptoms experienced in the 4 weeks preceding the stroke.
Family history of cardiovascular risk/diseases was defined based on self-reported history of any of hypertension, diabetes, dyslipidemia, stroke, cardiac disease or obesity in participants’ father, mother, sibling or second degree relative.
2.4. Statistical analysis
We compared demographic and vascular risk factor data among stroke cases who reported a prior history of febrile illness versus those without any prior febrile using Student’s t-test for parametrically distributed continuous data and Chi-squared tests for categorical data. Next, we assessed factors associated with reported antecedent history of febrile illness among stroke cases and by stroke types (ischemic and hemorrhagic strokes). We then compared information on stroke types, subtypes, and stroke severity according to antecedent history of febrile illness or not. Finally, we performed analyses to determine the adjusted associations between reported prior infection and risk of stroke occurrence for the total sample and stratified by stroke types. We first performed sequential adjusted estimates of association between prior febrile illness and stroke occurrence followed by a fully adjusted model used conditional logistic regression to estimate the adjusted odds ratio (OR) and 95% confidence intervals. The adjusted models included selected covariates depending on considerations from the literature on stroke risk factors, and or a p-value < 0.10 in unadjusted analysis. Additionally, the final adjusted models were assessed for collinearity using variance inflation factor (VIF) and goodness of fit using residual analysis. We calculated the adjusted Population Attributable Ratios (PARs) with their respective 95%CI for each exposure variable included in the best-fitted adjusted models. The PARs were estimated as the proportion of the risk of the stroke in the population that is attributable to the individual risk factors including prior infection (i.e. the proportion of cases that would not occur in the population if the factor were eliminated) [21]. The 95% CI for the PAR were obtained using the AF R-package [22] where the variance is estimated via the delta method. Composite PARs for the dominant risk factors for stroke, and stroke subtypes were calculated using the ATTRIBRISK R package with its 95% CI computed via the bootstrap method. All statistical tests of hypotheses were two-sided. Statistical analyses and graphics were performed with SAS 9.4 and R statistical program (version 3.4.2).
3. Results
3.1. Comparison of demographic and vascular risk factors among those with antecedent history of febrile illness and those without
Among 3588 stroke cases recruited in Ghana and Nigeria between August 2014 to July 2018, 363 cases (10.1%) reported having antecedent febrile illness within the 4 weeks prior to stroke occurrence. Overall, 83% of controls were recruited from communities where stroke cases resided, 17% were hospital-based controls.
Stroke patients with antecedent history of febrile illness were not significantly different in terms of mean age, gender, monthly income and educational attainment compared with those without history of febrile illness. However, those resident in rural settings had significantly higher frequency of reported febrile illness prior to stroke 24.0% versus 8.9% than among those without antecedent febrile illness, p < 0.001. The vascular risk factor profile between the two groups is shown in Table 1. Frequency of hypertension, dyslipidemia, diabetes, cardiac disease and obesity were not significantly different between the two groups. But, total cholesterol, LDL-cholesterol and HDL-cholesterol concentrations were significantly lower among those with antecedent febrile illness than those without. At the time of presentation with stroke, those with antecedent febrile illness were significantly more likely to report having psychosocial stress 41.9% versus 15.7%, p < 0.001, depression 19.3% versus 5.8% and with a higher frequency of family history of cardiovascular disease 47.7% versus 36.7%, p < 0.001. Conversely, green leafy vegetable, legume and fruit consumption was higher among those with antecedent febrile illness than those without. (Table 1).
Table 1.
Comparison of demographic and clinical characteristics of stroke cases with antecedent febrile illness versus those with no prior febrile illness.
| Variable | No prior febrile illness | Prior febrile illness | P-value |
|---|---|---|---|
| N = 3225 | N = 363 | ||
| Country, Ghana, n (%) | 1052 (32.62) | 103 (28.37) | 0.100 |
| Gender, Male, n (%) | 1794 (55.63) | 212 (58.40) | 0.331 |
| Age, mean ± SD | 59.97 ± 14.38 | 59.09 ± 14.85 | 0.272 |
| < 30 | 50 (1.55) | 12 (3.31) | 0.015 |
| 30–49 | 666 (20.65) | 75 (20.66) | 0.996 |
| 50–69 | 1604 (49.74) | 174 (47.93) | 0.515 |
| ≥70 | 885 (27.44) | 102 (28.10) | 0.790 |
| Domicile | |||
| Rural, n (%) | 288 (8.93) | 87 (23.97) | < 0.001 |
| Semi-urban, n (%) | 911 (28.25) | 126 (34.71) | < 0.001 |
| Urban, n (%) | 2010 (62.33) | 186 (51.24) | < 0.001 |
| Monthly Income > $100, n (%) | 1733 (53.74) | 187 (51.52) | 0.334 |
| Education, (some) n (%) | 2632 (81.61) | 288 (79.34) | 0.266 |
| Hypertension, n (%) | 3093 (95,91) | 343 (94.49) | 0.204 |
| Dyslipidemia, n (%) | 2664 (82.60) | 313 (86.23) | 0.09 |
| Diabetes, n (%) | 1218 (37.77) | 155 (42.70) | 0.067 |
| Cardiac Disease, n (%) | 375 (11.63) | 44 (12.12) | 0.783 |
| HDL-Cholesterol, mg/dl, mean ± SD | 48.18 ± 19.38 | 44.71 ± 19.61 | 0.003 |
| HDL-Cholesterol ≤18.54 mg/dl, n (%) | 1008 (31.26) | 141 (38.84) | 0.001 |
| LDL-Cholesterol, mg/dl, mean ± SD | 122.37 ± 51.33 | 111.61 ± 49.49 | 0.001 |
| LDL-Cholesterol ≥61.2 mg/dl, n (%) | 1072 (33.24) | 95 (26.17) | 0.007 |
| LDL/HDL ratio, mean ± SD | |||
| LDL/HDL ratio > 2.96, n (%) | 1005 (31.16) | 115 (31.68) | 0.616 |
| LDL/HDL ratio by thirds: | |||
| ≤ 2.00, n (%) | 854 (26.48) | 92 (25.34) | 0.641 |
| 2.01–2.96, n (%) | 803 (24.90) | 86 (23.69) | 0.613 |
| ≥ 2.97, n (%) | 1005 (31.16) | 115 (31.68) | 0.840 |
| Total Cholesterol, mmol/l, mean ± SD | 192.79 ± 57.65 | 179.15 ± 56.90 | < 0.001 |
| Total Cholesterol ≥93.6 mg/dl, n (%) | 1152 (35.72) | 107 (29.48) | 0.021 |
| Triglyceride, mg/dl, mean ± SD | 127.97 ± 85.05 | 126.81 ± 87.29 | 0.825 |
| Triglyceride ≥30.6 mg/dl, n (%) | 683 (21.18) | 73 (20.11) | 0.747 |
| Waist-to-hip Ratio, mean ± SD | |||
| Waist-to-hip Ratio raised, n (%) | 2463 (76.37) | 288 (79.32) | 0.092 |
| Waist-to-hip Ratio by thirds: | |||
| ≤ 0.90, n (%) | 799 (24.78) | 76 (20.94) | 0.106 |
| 0.91–0.96, n (%) | 1077 (33.40) | 122 (33.61) | 0.935 |
| ≥ 0.97+, n (%) | 1118 (34.67) | 137 (37.74) | 0.256 |
| WHR**, Lowest vs highest thirds, n (%) | 1119 (34.70) | 137 (37.74) | 0.091 |
| WHR**, 1st vs 2nd + 3rd thirds, n (%) | 2195 (68.06) | 259 (71.35) | 0.115 |
| BMI*** (kg/m2), mean ± SD | 26.76 ± 5.29 | 27.00 ± 5.78 | 0.471 |
| BMI*** > 30 kg/m2, n (%) | 541 (16.78) | 70 (19.28) | 0.280 |
| Physical Activity (some activity), n (%) | 2992 (92.78) | 335 (92.29) | 0.410 |
| Tobacco use in past 12 months, n (%) | 105 (3.26) | 15 (4.13) | 0.375 |
| Tobacco (any use), n (%) | 293 (9.09) | 41 (11.29) | 0.165 |
| Alcohol (current user), n (%) | 543 (16.84) | 63 (17.36) | 0.795 |
| Alcohol (any use), n (%) | 1025 (31.78) | 132 (36.36) | 0.074 |
| Alcohol use categories: | |||
| Never Use, n (%) | 2177 (67.50) | 228 (62.81) | 0.071 |
| Ever Low Use, n (%) | 581 (18.02) | 56 (15.43) | 0.221 |
| Ever High Use, n (%) | 75 (2.33) | 21 (5.79) | < 0.001 |
| Stress, n (%) | 507 (15.72) | 152 (41.87) | < 0.001 |
| Cancer, n (%) | 21 (0.65) | 3 (0.83) | 0.929 |
| Depression, n (%) | 186 (5.77) | 70 (19.28) | < 0.001 |
| Family history of CVD, n (%) | 1184 (36.71) | 173 (47.66) | < 0.001 |
| Adding salt at table, n (%) | 220 (6.82) | 30 (8.26) | 0.315 |
| Adding salt at table categories: | |||
| Never/rarely, n (%) | 2878 (89.24) | 320 (88.15) | 0.529 |
| Occasionally, n (%) | 2878 (89.24) | 320 (88.15) | 0.529 |
| Very often, n (%) | 220 (6.82) | 30 (8.26) | |
| Green vegetable consumption, n (%) | 2118 (65.67) | 292 (80.44) | < 0.001 |
| Whole grains consumption, n (%) | 2536 (78.64) | 276 (76.03) | 0.161 |
| Legumes consumption, n (%) | 1967 (60.99) | 259 (71.35) | < 0.001 |
| Fruit consumption, n (%) | 2505 (77.67) | 304 (83.75) | 0.002 |
| Sugar consumption or otherwise, n (%) | 854 (26.48) | 100 (27.55) | 0.615 |
| Meat consumption or otherwise, n (%) | 2574 (79.81) | 293 (80.72) | 0.360 |
| Fish consumption or otherwise, % | 2972 (92.16) | 332 (91.46) | 0.067 |
3.2. Factors associated with antecedent febrile illness prior to stroke onset
Two factors were independently associated with antecedent febrile illness prior to stroke, adjusted OR (95% CI): stress 4.69 (2.59–8.50) and regular consumption of green leafy vegetables 2.27 (1.35–2.85). By stroke types, one factor, stress was associated with antecedent febrile illness for ischemic stroke, adjusted OR of 5.65 (2.72–11.72). However, for hemorrhagic stroke, there were 4 independent factors: dyslipidemia 18.81 (2.44–145.16), diabetes mellitus 8.18 (1.76–38.07), green leafy vegetable 4.76 (1.19–20.00) and stress 9.00 (1.45–55.81). Unadjusted estimates of covariates are shown in Table 2.
Table 2.
Multivariable logistic regression analysis for factors associated with antecedent febrile illness prior to stroke onset.
| Risk factor | All Stroke strokes | Ischemic stroke | Hemorrhagic stroke | |||
|---|---|---|---|---|---|---|
| Unadjusted OR (95% CI) | Adjusted OR (95% CI) | Unadjusted OR (95% CI) | Adjusted OR (95% CI) | Unadjusted OR (95% CI) | Adjusted OR (95% CI) | |
| Domicile | 0.63 (0.38–1.05) | 0.56 (0.30–1.04) | 0.60 (0.22–1.65) | |||
| HDL Cholesterol | 1.61 (1.11–2.35) | 1.19 (0.73–1.94) | 1.58 (1.01–2.48) | 1.47 (0.82–2.66) | 2.43 (1.01–5.86) | 1.31 (0.33–5.30) |
| LDL Cholesterol | 0.85 (0.61–1.19) | 0.98 (0.64–1.49) | 0.54 (0.28–1.06) | |||
| Dyslipidemia | 1.83 (1.22–2.75) | 1.65 (0.92–2.97) | 1.58 (0.96–2.58) | 3.20 (1.17–8.73) | 18.81 (2.44–145.16) | |
| Diabetes | 1.63 (1.11–2.41) | 1.38 (0.87–2.21) | 1.48 (0.94–2.34) | 3.40 (1.25–9.22) | 8.18 (1.76–38.07) | |
| Regular consumption of green leafy vegetables | 2.86 (1.82–4.55) | 2.27 (1.35–3.85) | 2.50 (1.45–4.35) | 1.79 (0.92–3.44) | 3.23 (1.18–9.09) | 4.76 (1.19–20.00) |
| Regular consumption of legumes | 1.24 (0.81–1.90) | 1.25 (0.76–2.05) | 1.43 (0.54–3.75) | |||
| Regular consumption of fruit | 1.44 (0.88–2.36) | 0.59 (0.32–1.07) | 0.89 (0.34–2.30) | |||
| Alcohol | 1.51 (0.99–2.32) | 1.16 (0.69–1.93) | 1.63 (0.98–2.70) | 1.34 (0.72–2.51) | 1.22 (0.51–2.95) | |
| Stress | 6.47 (3.75–11.14) | 4.69 (2.59–8.50) | 6.80 (3.50–13.21) | 5.65 (2.72–11.72) | 5.50 (1.90–15.96) | 9.00 (1.45–55.81) |
| Depression | 3.64 (1.87–7.09) | 1.72 (0.77–3.82) | 3.83 (1.56–9.41) | 1.89 (0.64–5.58) | 5.00 (1.45–17.72) | 2.25 (0.38–13.30) |
3.3. Comparison of stroke types, subtypes, severity and outcomes in relation to antecedent history of febrile illness
There were no differences in frequency of the primary stroke types in relation to antecedent febrile illness. However, among ischemic stroke etiologic subtypes, 61.4% of those with small vessel occlusive disease reported having an antecedent febrile illness while 53.8% with this ischemic stroke subtype reported not having febrile illness, p = 0.06. There were no significant differences by etiologic subtypes for hemorrhagic stroke and stroke severity using mean scores on the National Institutes of Health Stroke scale (NIHSS). (Table 3).
Table 3.
Comparison of stroke characteristics among cases with reported febrile illness versus those without prior febrile illness.
| Variable | No prior febrile illness | Prior febrile illness | P-value |
|---|---|---|---|
| N = 3225 | N = 363 | ||
| Stroke type | 0.13 | ||
| Ischemic | 2013 (70.4) | 239 (74.5) | |
| Hemorrhagic | 845 (29.6) | 82 (25.5) | |
| Ischemic stroke subtypes | |||
| OCSP | |||
| TACI | 260 (14.5) | 34 (15.7) | 0.64 |
| PACI | 608 (34.0) | 60 (27.8) | 0.07 |
| LACI | 185 (10.3) | 26 (12.0) | 0.44 |
| POCI | 736 (41.1) | 96 (44.4) | 0.35 |
| ASCO | |||
| Large artery atherosclerosis | 345 (26.2) | 43 (25.1) | 0.76 |
| Small vessel occlusion | 708 (53.8) | 105 (61.4) | 0.06 |
| Cardio-embolic | 222 (16.9) | 21 (12.3) | 0.13 |
| Others | 41 (3.1) | 2 (1.2) | 0.15 |
| ICH subtypes | |||
| Structural | 33 (4.4) | 5 (7.5) | 0.25 |
| Medication-related | 4 (0.5) | 0 (0.00) | 0.55 |
| Amyloid angiopathy | 10 (1.3) | 2 (3.0) | 0.28 |
| Systemic diseases | 4 (0.5) | 0 (0.00) | 0.55 |
| Hypertension | 682 (90.7) | 57 (85.1) | 0.14 |
| Undetermined | 19 (2.5) | 3 (4.5) | 0.34 |
| Stroke severity | |||
| NIHSS, mean ± SD | 13.0 ± 8.7 | 13.3 ± 8.6 | 0.65 |
| NIHSS | 0.28 | ||
| Mild | 555 (26.9) | 51 (22.1) | 0.12 |
| Moderate | 1056 (51.1) | 128 (55.4) | 0.22 |
| Severe | 455 (22.0) | 52 (22.5) | 0.87 |
3.4. Association between antecedent febrile illness and stroke occurrence
Table 4 shows a comparative analysis between stroke cases and stroke-free controls. While 363 (10.1%) of stroke cases reported an antecedent febrile illness, 280 (7.8%) of stroke-free controls had had a febrile illness in the previous four weeks, p = 0.0006. There were significant differences in several demographic and vascular risk factors between cases and controls as shown in Table 4. Having an antecedent febrile illness was associated with stroke occurrence with an unadjusted OR of 1.19 (1.00–1.51). The unadjusted odds ratio for hemorrhagic stroke was 1.32 (0.81–2.15), ischemic stroke was 1.11 (0.83–1.47) and for small vessel occlusive ischemic strokes was 1.26 (0.83–1.93). As shown in Table 5, sequential adjustments for key demographic and vascular risk factors attenuated the observed associations between antecedent febrile and stroke occurrence. The fully adjusted models for stroke overall, ischemic stroke and hemorrhagic strokes are shown in Table 6. Having an antecedent febrile illness was not independently associated with stroke occurrence, adjusted OR of 0.83 (95% CI: 0.59–1.17).
Table 4.
Comparison of demographic and clinical characteristics of stroke cases versus stroke-free controls.
| Variable | Stroke status | ||
|---|---|---|---|
| Control | Case | ||
| (n = 3588) | (n = 3588) | p-value | |
| Gender, Male, % [1] | 1942 (54.1) | 1939 (54.0) | 0.92 |
| Age, mean ± SD | 58.71 ± 13.95 | 59.17 ± 13.94 | 0.17 |
| Age > 65 years, % | 1123 (31.3) | 1130 (31.5) | 0.84 |
| Monthly Income > $100, % | 1340 (37.3) | 1910 (53.2) | < 0.001 |
| Education, (some) % | 2751 (76.7) | 2879 (80.2) | < 0.001 |
| Antecedent febrile illness, % | 280 (7.8) | 363 (10.1) | 0.0006 |
| Hypertension, % | 2093 (58.3) | 3361 (93.7) | < 0.001 |
| Dyslipidemia, % | 1931 (53.8) | 2912 (81.2) | < 0.001 |
| Diabetes, % | 403 (11.2) | 1322 (36.8) | < 0.001 |
| Cardiac Disease, % | 154 (4.3) | 414 (11.5) | < 0.001 |
| HDL-Cholesterol, mmol/l, mean ± SD | 52.62 ± 17.59 | 47.89 ± 19.35 | < 0.001 |
| HDL-Cholesterol ≤ 1.03 mmol/l, % | 729 (20.3) | 1113 (31.0) | < 0.001 |
| LDL-Cholesterol, mmol/l, mean ± SD | 120.52 ± 45.06 | 121.17 ± 50.63 | 0.60 |
| LDL-Cholesterol ≥ 3.4 mmol/l, % | 1090 (30.4) | 1138 (31.7) | 0.157 |
| LDL/HDL ratio, mean ± SD | 2.55 ± 1.35 | 2.92 ± 1.88 | < 0.001 |
| LDL/HDL ratio > 2.96, % | 890 (24.8) | 1081 (30.1) | < 0.001 |
| Total Cholesterol, mmol/l, mean ± SD | 194.26 ± 49.28 | 191.72 ± 57.11 | 0.07 |
| Total Cholesterol ≥ 5.2 mmol/l, % | 1202 (33.5) | 1246 (34.7) | 0.329 |
| Triglyceride, mmol/l, mean ± SD | 105.61 ± 50.86 | 128.11 ± 85.31 | < 0.001 |
| Triglyceride ≥ 1.7 mmol/l, % | 440 (12.3) | 754 (21.0) | < 0.001 |
| Waist-to-hip Ratio raised, % | 2432 (67.8) | 2687 (74.9) | < 0.001 |
| BMI*** (kg/m2), mean ± SD | 26.18 ± 5.78 | 26.76 ± 5.30 | < 0.001 |
| BMI*** > 30 kg/m2, % | 730 (20.3) | 593 (16.5) | 0.984 |
| Physical Activity (some activity), % | 3374 (94.0) | 3255 (90.7) | < 0.001 |
| Tobacco use in past 12 months, % | 46 (1.3) | 121 (3.4) | < 0.001 |
| Tobacco (any use), % | 254 (7.1) | 334 (9.3) | < 0.001 |
| Alcohol (current user), % | 493 (13.7) | 575 (16.0) | 0.003 |
| Alcohol (any use), % | 997 (27.8) | 1108 (30.9) | 0.001 |
| Stress, % | 467 (13.0) | 652 (18.2) | < 0.001 |
| Cancer, % | 3 (0.08) | 22 (0.61) | < 0.001 |
| Depression, % | 197 (5.5) | 253 (7.1) | < 0.001 |
| Family history of CVD, % | 913 (25.4) | 1330 (37.1) | < 0.001 |
| Adding salt at table, % | 282 (7.9) | 232 (6.5) | 0.06 |
| Green vegetable consumption, % | 2827 (78.8) | 2356 (65.7) | < 0.001 |
| Whole grains consumption, % | 2924 (81.5) | 2741 (76.4) | 0.705 |
| Legumes consumption, % | 2165 (60.3) | 2180 (60.8) | 0.001 |
| Fruit consumption, % | 2973 (82.9) | 2742 (76.4) | 0.04 |
| Sugar consumption or otherwise, % | 1158 (32.3) | 942 (26.3) | < 0.001 |
| Meat consumption or otherwise | 2827 (78.8) | 2792 (77.8) | < 0.001 |
| Fish consumption or otherwise, % | 2352 (65.6) | 2431 (67.8) | < 0.001 |
Table 5.
Associations between antecedent febrile illness and stroke occurrence.
| Stroke overall | Ischemic stroke | Small vessel occlusive stroke | Hemorrhagic stroke | |
|---|---|---|---|---|
| Adjusted OR (95% CI) | Adjusted OR (95% CI) | Adjusted OR (95% CI) | Adjusted OR (95% CI) | |
| Model 1 | 1.19 (1.00–1.51) | 1.11 (0.83–1.47) | 1.26 (0.83–1.93) | 1.32 (0.81–2.15) |
| Model 2 | 1.20 (0.95–1.51) | 1.12 (0.84–1.48) | 1.29 (0.85–1.99) | 1.33 (0.813–2.18) |
| Model 3 | 1.22 (0.96–1.56) | 1.09 (0.82–1.48) | 1.33 (0.84–2.11) | 1.55 (0.93–2.61) |
| Model 4 | 1.11 (0.83–1.48) | 1.02 (0.73–1.43) | 1.05 (0.64–2.08) | 1.35 (0.64–2.85) |
| Model 5 | 0.89 (0.65–1.22) | 0.81 (0.55–1.19) | 1.09 (0.60–1.98) | 0.93 (0.43–2.04) |
| Model 6 | 0.83 (0.59–1.17) | 0.70 (0.47–1.06) | 1.06 (0.56–2.01) | 0.84 (0.34–2.05) |
Model 1: unadjusted; Model 2: aged- and sex-adjusted; Model 3: adjusted for age, sex, location of residency, income, educational attainment; Model 4: adjusted for hypertension only; Model 5: adjusted for hypertension, dyslipidemia and diabetes; Model 6: Full adjustment (see Table 5).
Table 6.
Factors Associated with occurrence of stroke by its types (Multivariate).
| Risk factor | All Stroke strokes | Ischemic stroke | Hemorrhagic stroke | |||
|---|---|---|---|---|---|---|
| Adjusted OR (95% CI) | PAR (95% CI) | Adjusted OR (95% CI) | PAR (95% CI) | Adjusted OR (95% CI) | PAR (95% CI) | |
| Antecedent febrile illness | 0.83 (0.59–1.17) | −1.96 (−5.73–1.82) | 0.70 (0.47–1.06) | −4.17 (−9.49–1.15) | 0.84 (0.34–2.05) | −1.71 (−12.21–8.79) |
| Age | 0.55 (0.35–0.86) | −26.01 (−52.97–0.95) | 0.57 (0.33–0.98) | −28.98 (−68.91–10.95) | 0.46 (0.15–1.48) | −19.55 (−61.31–22.31) |
| Gender | 0.83 (0.21–3.27) | −11.11 (−95.32–73.09) | 0.59 (0.10–3.54) | −36.50 (−89.85–16.84) | 9.15 (0.11–744.06) | 57.32 (46.81–67.83) |
| Hypertension | 17.96 (12.54–25.72) | 90.359 (88.12–92.57) | 10.54 (6.99–15.87) | 85.51 (81.38–89.63) | 95.24 (28.81–314.80) | 97.34 (95.70–98.98) |
| Dyslipidemia | 3.68 (2.90–4.67) | 61.49 (55.92–67.06) | 5.18 (3.76–7.12) | 70.82 (65.15–76.48) | 2.16 (1.31–3.56) | 43.16 (24.82–61.50) |
| Diabetes mellitus | 3.25 (2.57–4.10) | 26.06 (22.94–29.19) | 3.71 (2.78–4.97) | 30.78 (26.95–34.60) | 2.34 (1.40–3.91) | 15.40 (8.71–22.09) |
| Obesity | 0.74 (0.59–0.93) | −7.67 (−14.23 – −1.11) | 0.84 (0.63–1.12) | −4.36 (−12.03–3.30) | 0.50 (0.31–0.81) | −18.70 (−36.23 – −1.17) |
| Cigarette smoking | 1.96 (1.00–3.98) | 1.54 (0.23–2.85) | 1.52 (0.62–3.72) | 0.93 (−0.79–2.65) | 2.43 (0.58–10.17) | 2.64 (−0.60–5.87) |
| Stress | 1.69 (1.28–2.22) | 8.92 (5.19–12.65) | 1.79 (1.26–2.54) | 10.09 (5.38–14.80) | 1.54 (0.89–2.67) | 7.45 (0.25–14.65) |
| Cardiac disease | 1.71 (1.16–2.51) | 3.98 (1.45–6.51) | 2.02 (1.28–3.18) | 6.08 (2.88–9.28) | 1.01 (0.39–2.61) | 0.07 (−6.17–6.30) |
| Alcohol | 1.07 (0.84–1.35) | 2.09 (−5.72–9.92) | 0.87 (0.64–1.18) | −4.93 (−16.79–6.92) | 1.77 (1.08–2.91) | 17.74 (6.14–29.35) |
| Meat consumption | 2.12 (1.71–2.64) | 39.67 (31.75–47.59) | 2.01 (1.53–2.65) | 36.08 (26.11–46.04) | 1.99 (1.26–3.16) | 38.69 (20.50–56.87) |
| Low Vegetable consumption | 2.27 (1.81–2.86) | 18.95 (15.22–22.69) | 2.21 (1.64–2.97) | 17.50 (12.88–22.13) | 2.10 (1.31–3.38) | 19.44 (10.15–28.74) |
| Depression | 0.89 (0.59–1.36) | −0.89 (−4.23–2.45) | 1.06 (0.62–1.82) | 0.51 (−3.40–4.41) | 0.67 (0.28–1.64) | −3.25 (−12.37–5.87) |
| Physical inactivity | 0.78 (0.43–1.42) | −0.76 (−2.96–1.46) | 0.74 (0.35–1.54) | −1.00 (−4.20–2.20) | 0.72 (0.17–3.03) | −1.09 (−6.30–4.13) |
| Salt intake | 1.51 (1.01–2.21) | 2.66 (0.74–4.58) | 1.31 (0.80–2.16) | 1.60 (−0.82–4.01) | 2.97 (1.25–7.07) | 7.57 (4.37–10.76) |
| Composite PAR | 90.55 (87.93–93.17) | 87.45 (83.09–91.81) | 96.21 (91.99–99.810) | |||
4. Discussion
Among nearly 3600 stroke cases in Ghana and Nigeria, 10% reported an antecedent history of a febrile illness within the preceding 4 weeks before stroke onset. We assume that a significant majority of our study population with a history of antecedent febrile illness might have had an infection. Our local experience has taught us that many adults with fever may self-medicate and do not seek healthcare at medical facilities. Thus we relied on self-reports by stroke cases or their valid proxies for aphasic or unconscious stroke cases for a history of febrile illness as a possible proxy for antecedent infection. Previous studies have reported higher rates of pre-stroke infections. In the United States, 35% (n = 37) of subjects with incident ischemic strokes reported had an antecedent infection within 1 week of stroke onset [23]. Furthermore, 38% (n = 50) of subjects with cerebral infarction were found to have evidence of infections after detailed history and serological testing for several bacteria and viruses in Finland [24] while 57% of 70 subjects reported having a respiratory or urinary tract infection within 2 weeks of ischemic stroke onset in India [25]. The lower frequency of pre-stroke febrile illness identified in the present study may thus be due to unavailability of medical records for verification of in-patient hospitalizations, out-patient visits or visits to pharmacy shops for treatment for potential causes of febrile illness including infections. Hospital data are notoriously difficult to access in our settings due to a pervasive lack of electronic health records. Given the above limitations and caveats, we discuss our findings using the terminology ‘febrile illness’ instead of ‘infections’.
Pre-stroke febrile illness was significantly associated with reported psychosocial stress and regular consumption of green leafy vegetables in our study. Our previous reports from the SIREN study have demonstrated associations between stress and green leafy vegetable consumption and stroke occurrence overall [26,27], by stroke types [28], by gender [29] and by age sub-groups [30] in West Africa. Presently, we show that antecedent febrile illness is associated with psychosocial stress among ischemic stroke cases. Interestingly, among hemorrhagic stroke cases four independent factors were associated with antecedent febrile illness namely stress, dyslipidemia, diabetes and regular green leafy vegetable consumption. The consistency of the observed associations between pre-stroke febrile illness and prior psychosocial stress for both ischemic and hemorrhagic strokes is intriguing because both factors are considered triggers or precipitants of stroke and may potentiate risk of incident stroke among individuals with established cardiovascular risk factors [4–7].
The mechanistic pathways by which infections could trigger stroke include increased platelet activation, elevation of circulating leucocytes contributing to atherogenesis and thrombogenesis via elaboration of cytokine, perturbation of clotting factor function and concentrations and disordered endothelial function [4,31,32]. Indeed, in previous candidate gene studies of the SIREN cohort, we had demonstrated an association between the interleukin – 6 gene locus (rs1800796) and ischaemic stroke [33]. Promoter polymorphisms of the IL-6 gene regulate the circulating plasma level of interleukin – 6 a pleiotropic cytokine which plays critical roles in the acute inflammatory response and could trigger endothelial dysfunction and activation of the coagulation – fibrinolysis system [34]. It is noteworthy that hemorrhagic stroke cases with prior infection were more likely to have dyslipidemia with an adjusted odds ratio of 18.81 (2.44–145.16), diabetes mellitus 8.18 (1.76–38.07), and stress 9.00 (1.45–55.81). The wide confidence intervals observed is undoubtedly due to the limited sample size of 927 subjects with intracerebral hemorrhage out of which 82 (8.8%) reported an antecedent febrile illness. The implication of this finding-given that nearly 95% of all intracerebral hemorrhages (ICH) in West Africa are causally attributable to hypertension [28] – is that a vigorous systemic inflammatory cascade is perhaps orchestrated possibly by a febrile illness prior to a catastrophic ICH event. These observations require further studies to unravel the exact pathways that are triggered by infection prior to stroke. Overall, reporting an antecedent febrile illness was associated with a modest risk for stroke occurrence with odds of 19% (0% to 51%) in unadjusted models compared with stroke-free controls. However, adjustment for demographic and traditional vascular risk factors, no independent associations remained between antecedent febrile illness and stroke risk as well as subtypes (Tables 5 and 6). Furthermore, pre-stroke infectious exposures did not demonstrate a clear predilection for one stroke type or etiologic subtypes.
A limitation of the present study is that data on the causes and time frame within the 4- week period of antecedent febrile illness were not collected. This is important because some studies on this subject have shown that urinary tract infections and respiratory tract infections are common antecedent infections for strokes [23–25]. Furthermore, a more recent case-over analytic study (each stroke case served as its own control) has demonstrated all infection types namely infections of the urinary tract, respiratory tract, skin, abdomen and sepsis to be associated with stroke [7]. There is potential recall bias and reporting bias which might have led to an underestimation of the burden of antecedent febrile illness. This is because among stroke cases who were either aphasic or unconscious, reliable proxies responded to the questionnaires while stroke free controls provided responses themselves. We cannot draw causal inferences between antecedent infections and stroke occurrence in the present study due the case-control design. Additionally, we did not account for multiple comparisons in our analysis. Approximately, 17% of stroke-free controls were recruited from hospitals either as relatives of another patient, or as patients visiting the hospitals for other conditions not related to stroke. It is possible that some of these controls might have come to hospital on account of antecedent febrile illness with the potential to attenuate the associations between stroke and antecedent febrile illness. The aforementioned limitations will be addressed in subsequent studies. This however is the first study out of Africa to address potential associations between febrile illness and stroke. In Africa, the burden of stroke is escalating [35–47] but no prior studies have assessed the associations between infectious exposures and risk of stroke.[48–52]
In conclusion 1 in 10 stroke cases reported antecedent history of febrile illness prior to occurrence of stroke but no independent association was observed in this study. Infectious exposures may be important triggers of cardiovascular events requiring further exploratory studies to better understand the role of this emerging risk factor.
Funding
The SIREN (Stroke Investigative Research and Education Networks) and SIBS Genomics studies were funded by the National Institutes of Health grant U54 HG007479 and R01NS107900 under the H3Africa initiative.
Footnotes
Disclosures
All authors have no conflicts to declare.
References
- [1].Feigin VL, Forouzanfar MH, Krishnamurthi R, Global Burden of Diseases, Injuries, and Risk Factors Study 2010 (GBD 2010), the GBD Stroke Experts Group, et al. , Global and regional burden of stroke during 1990–2010: findings from the Global Burden of Disease Study 2010, Lancet 383 (2014) 245–254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Feigin VL, Lawes CM, Bennett DA, Barker-Collo SL, Parag V, Worldwide stroke incidence and early case fatality reported in 56 population-based studies: a systematic review, Lancet Neurol. 8 (2009) 355–369. [DOI] [PubMed] [Google Scholar]
- [3].Elkind MS, Why now? Moving from stroke risk factors to stroke triggers, Curr. Opin. Neurol 20 (2007) 51–57. [DOI] [PubMed] [Google Scholar]
- [4].Elkind MS, Carty CL, O’Meara ES, Lumley T, Lefkowitz D, Kronmal RA, et al. , Hospitalization for infection and risk of acute ischemic stroke: the cardiovascular health study, Stroke. 42 (2011) 1851–1856. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Boehme AK, Ranawat P, Luna J, Kamel H, Elkind MS, Risk of acute stroke after hospitalization for sepsis: a case-crossover study, Stroke. 48 (2017) 574–580. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Yawn BP, Wollan PC, Nagel MA, Gilden D, Risk of stroke and myocardial infarction after herpes zoster in older adults in a US community population, Mayo Clin. Proc 91 (2016) 33–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Sebastian S, Stein LK, Dhamoon MS, Infection as a stroke trigger, Stroke. 50 (8) (2019) 2216–2218. [DOI] [PubMed] [Google Scholar]
- [8].Walker R, Whiting D, Unwin N, Mugusi F, Swai M, Stroke incidence in rural and urban Tanzania: a prospective, community-based study, Lancet Neurol. 9 (2010) 786–792. [DOI] [PubMed] [Google Scholar]
- [9].Ezejimofor MC, Uthman OA, Maduka O, Ezeabasili AC, Onwuchekwa AC, Ezejimofor BC, et al. , Stroke survivors in Nigeria: a door-to-door prevalence survey from the Niger Delta region, J. Neurol. Sci 372 (2017. January 15) 262–269. [DOI] [PubMed] [Google Scholar]
- [10].Sarfo FS, Awuah DO, Nkyi C, Akassi J, Opare-Sem OK, Ovbiagele B, Recent patterns and predictors of neurological mortality among hospitalized patients in Central Ghana, J. Neurol. Sci 363 (2016) 217–224. [DOI] [PubMed] [Google Scholar]
- [11].Sarfo FS, Akassi J, Awuah D, Adamu S, Nkyi C, Owolabi M, et al. , Trends in stroke admission and mortality rates from 1983 to 2013 in central Ghana, J. Neurol. Sci 357 (1–2) (2015. October 15) 240–245. [DOI] [PubMed] [Google Scholar]
- [12].Akpalu A, Sarfo FS, Ovbiagele B, Akinyemi R, Gebregziabher M, Obiako R, et al. , Phenotyping stroke in sub-Saharan Africa: stroke investigative research and education network (SIREN) phenomics protocol, Neuroepidemiology 45 (2015) 73–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Sarfo F, Gebregziabher M, Ovbiagele B, Akinyemi R, Owolabi L, Obiako R, et al. , Multilingual validation of the questionnaire for verifying stroke-free status in West Africa, Stroke 47 (2016) 167–172. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Amarenco P, Bogousslavsky J, Caplan LR, Donnan GA, Hennerici MG, New approach to stroke subtyping: the A-S-C-O (phenotypic) classification of stroke, Cerebrovasc. Dis 27 (2009) 502–508. [DOI] [PubMed] [Google Scholar]
- [15].Meretoja A, Strbian D, Putaala J, Curtze S, Haapaniemi E, Mustanoja S, et al. , SMASH-U: a proposal for etiologic classification of intracerebral hemorrhage, Stroke 43 (2012) 2592–2597. [DOI] [PubMed] [Google Scholar]
- [16].O’Donnell M, Xavier D, Diener C, Sacco R, Lisheng L, Zhang H, et al. , Rationale and design of INTERSTROKE: a global case-control study of risk factors for stroke, Neuroepidemiology 35 (2010) 36–44. [DOI] [PubMed] [Google Scholar]
- [17].Fischer U, Conney MT, Bull LM, Silver LE, Chalmers J, Anderson CS, et al. , Acute post-stroke blood pressure relative to premorbid levels in intracerebral hemorrhage versus major ischemic stroke: a population-based study, Lancet Neurol. 13 (2014) 374–384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Alberti KG, Zimmet PZ, Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation, Diabet. Med 15 (1998) 539–553. [DOI] [PubMed] [Google Scholar]
- [19].National, Cholesterol education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (adult treatment panel). Third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high cholesterol in adults (adult treatment panel III), Circulation 106 (2002) 3143–21. [PubMed] [Google Scholar]
- [20].World Health Organization, Waist Circumference and Waist-hip Ratio. Report of a WHO Expert Consultation, http://www.who.int/nutrition/publications/obesity/WHO_report_waistcircumference_and_waisthip_ratio/en/, (2008) (Assessed May 27, 2017).
- [21].Llorca J, Delgado- Rodriguez M, A comparison of several procedures to estimate the confidence interval for attributable risk in case-control studies, Stat. Med 19 (2000) 1089–1099. [DOI] [PubMed] [Google Scholar]
- [22].Dahlqwist E, Sjölander A, AF Package- Model Based Estimation of Confounder-adjusted Attributable Fractions, https://cran.r-project.org/web/packages/AF/index.html.
- [23].Macko RF, Ameriso SF, Barndt R, et al. , Precipitants of brain infarction, Stroke. 27 (1996) 1999–2004. [DOI] [PubMed] [Google Scholar]
- [24].Syrjanan J, Vattonen VV, Livanainen M, et al. , Preceding infection as an important risk factor for ischemic brain infarction in youn and middle aged patients, Br. Med. J 296 (1988) 1156–1160. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Das S, Ghosh KC, Pulai S, Pulai D, Bhar D, Gangopadhyay PK, Systemic infection and inflammation as trigger factors for ischemic stroke, Ann. Neurosci 18 (1) (2011) 17–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Owolabi M, Sarfo F, Howard VJ, Irvin MR, Gebregziabher M, Akinyemi R, et al. , Stroke in indigenous Africans, African Americans, and European Americans: interplay of racial and geographic factors, Stroke. 48 (5) (2017) 1169–1175. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Owolabi MO, Sarfo F, Akinyemi R, Gebregziabher M, Akpa O, Akpalu A, et al. , Dominant modifiable risk factors for stroke in Ghana and Nigeria (SIREN): a case-control study, Lancet Global Health. 6 (4) (2018) e436–e446. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Sarfo FS, Ovbiagele B, Gebregziabher M, Akpa O, Akpalu A, Wahab K, et al. , Unraveling the risk factors for spontaneous intracerebral hemorrhage among West Africans, Neurology. 94 (10) (2020. March 10) e998–e1012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Akpalu A, Gebregziabher M, Ovbiagele B, Sarfo F, Iheonye H, Akinyemi R, et al. , Differential impact of risk factors on stroke occurrence among men versus women in West Africa, Stroke. 50 (4) (2019. April) 820–827. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Sarfo FS, Ovbiagele B, Gebregziabher M, Wahab K, Akinyemi R, Akpalu A, et al. , Stroke among young West Africans: evidence from the SIREN (stroke investigative research and educational network) large multisite case-control study, Stroke. 49 (5) (2018. May) 1116–1122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [31].Grau AJ, Boddy AW, Dukovic DA, Buggle F, Lichy C, Brandt T, CAPRIE Investigators, et al. , Leucocyte count as an independent predictor of recurrent ischemic events, Stroke. 35 (2004) 1147–1152. [DOI] [PubMed] [Google Scholar]
- [32].Mattila KJ, Viral and bacterial infections in patients with acute myocardial infarction, J. Intern. Med 225 (1989) 293–296. [DOI] [PubMed] [Google Scholar]
- [33].Akinyemi R, Arnett DK, Tiwari HK, et al. , Interleukin-6 (IL-6) rs1800796 and cyclin dependent kinase inhibitor (CDKN2A/CDKN2B) rs2383207 are associated with ischemic stroke in indigenous West African Men, J. Neurol. Sci 379 (2017) 229–235. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [34].Cole JW, Brown DW, Giles WH, et al. , Ischemic stroke risk, smoking, and the genetics of inflammation in a biracial population: the stroke prevention in young women study, Thromb. J 6 (2008) 11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].Sarfo FS, Mobula LM, Sarfo-Kantanka O, Adamu S, Plange-Rhule J, Ansong D, et al. , Estimated glomerular filtration rate predicts incident stroke among Ghanaians with diabetes and hypertension, J. Neurol. Sci 396 (2019) 140–147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [36].Walker RW, Jusabani A, Aris E, et al. , Post-stroke case fatality within an incident population in rural Tanzania, J. Neurol. Neurosurg. Psychiatry 82 (9) (2011) 1001–1005. [DOI] [PubMed] [Google Scholar]
- [37].Danesi MA, Okubadejo NU, Ojini FI, Ojo OO, Incidence and 30-day case fatality rate of first-ever stroke in urban Nigeria: the prospective community based Epidemiology of Stroke in Lagos (EPISIL) phase II results, J. Neurol. Sci 331 (2013) 43–47. [DOI] [PubMed] [Google Scholar]
- [38].Gomes J, Damasceno A, Carrilho C, et al. , Determinants of early case-fatality among stroke patients in Maputo, Mozambique and impact of in-hospital complications, Int. J. Stroke 8 (2013) 69–75 Suppl. 100. [DOI] [PubMed] [Google Scholar]
- [39].Sarfo FS, Akassi J, Kyem G, Adamu S, Awuah D, Kantanks OS, et al. , Long-term outcomes of stroke in a Ghanaian outpatient clinic, J. Stroke Cerebrovasc. Dis 27 (3) (2018) 1090–1099. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [40].Sarfo FS, Jenkins C, Singh A, Owolabi M, Ojagbemi A, Adusei N, Saulson R, Ovbiagele B, Post-stroke depression in Ghana: characteristics and correlates, J. Neurol. Sci 379 (2017) 261–265. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [41].Sarfo FS, Akassi J, Adamu S, Obese V, Ovbiagele B, Burden and predictors of vascular cognitive impairment among long-term Ghanaian stroke survivors, J Stroke Cerebrovas Dis 26 (11) (2017) 2553–2562. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [42].Sarfo FS, Nichols M, Qanungo S, Teklehaimanot A, Singh A, Mensah N, et al. , Stroke-related stigma among West Africans: patterns and predictors, J. Neurol. Sci 375 (2017) 270–274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [43].Sarfo FS, Berchie P, Singh A, Nichols M, Agyei-Frimpong M, Jenkins C, et al. , Prevalence, trajectory and predictors of poststroke fatigue among Ghanaians, J. Stroke Cerebrovasc. Dis 28 (5) (2019) 1353–1361. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [44].Sarfo F, Treiber F, Gebregziabher M, Adamu S, Patel S, Nichols M, et al. , PINGS (phone-based intervention under nurse guidance after stroke): interim results of a pilot randomized controlled trial, Stroke. 49 (1) (2018) 236–239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [45].Sarfo FS, Treiber F, Gebregziabher M, Adamu S, Nichols M, Singh A, et al. , Phone-based intervention for blood pressure control among Ghanaian stroke survivors: a pilot randomized controlled trial, Int. J. Stroke 14 (6) (2019) 630–638. [DOI] [PubMed] [Google Scholar]
- [46].Sarfo FS, Opare-Sem O, Agyei M, Akassi J, Owusu D, Owolabi M, et al. , Risk factors for stroke occurrence in a low HIV endemic West African country: a case-control study, J. Neurol. Sci 395 (2018. December 15) 8–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [47].Sarfo FS, Nichols M, Agyei B, Singh A, Ennin E, Nyantakyi AD, et al. , Burden of subclinical carotid atherosclerosis and vascular risk factors among people living with HIV in Ghana, J. Neurol. Sci 397 (2019. February 15) 103–111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [48].Sarfo FS, Nichols M, Singh A, Hardy Y, Norman B, Mensah G, et al. , Characteristics of hypertension among people living with HIV in Ghana: impact of new hypertension guideline, J Clin Hypertens (Greenwich). 21 (6) (2019. June) 838–850. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [49].Sarfo FS, Kasim A, Phillips R, Geretti AM, Chadwick DR, Long-term responses to first-line antiretroviral therapy in HIV and hepatitis B co-infection in Ghana, J. Inf. Secur 69 (5) (2014. November) 481–489. [DOI] [PubMed] [Google Scholar]
- [50].Sarfo FS, Mensah NO, Opoku FA, Adusei-Mensah N, Ampofo M, Ovbiagele B, COVID-19 and stroke: experience in a Ghanaian healthcare system, J. Neurol. Sci 416 (2020. July 16) 117044. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [51].Sarfo FS, Akassi J, Adamu S, Obese V, Agbenorku M, Ovbiagele B, Frequency and factors linked with refractory hypertension among stroke survivors in Ghana, J. Neurol. Sci 415 (2020. August 15) 116976. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [52].Sarfo FS, Gyamfi RA, Adamu S, Sarfo-Kantanka O, Owolabi M, Ovbiagele B, Administration of a pictorial questionnaire to screen for stroke among patients with hypertension or diabetes in rural Ghana, J. Neurol. Sci 373 (2017. February 15) 289–294. [DOI] [PMC free article] [PubMed] [Google Scholar]