p u b l i c h e a l t h x x x ( 2 0 1 5 ) 1 e5
Available online at www.sciencedirect.com
Public Health journal homepage: www.elsevier.com/puhe
Original Research
Stroke and periodontal disease in Senegal: caseecontrol study M. Diouf a,*, A. Basse b, M. Ndiaye b, D. Cisse c, C.M. Lo c, D. Faye c a
UCAD, Dakar, Senegal Neurology Department, Dakar, Senegal c Medicine Pharmacy and Dentistry Faculty, Dakar, Senegal b
article info
abstract
Article history:
Objective: To determine the periodontal factors associated with stroke in melanodermic
Received 23 August 2014
subjects in Senegal.
Received in revised form
Study design: Caseecontrol study.
16 December 2014
Methods: One hundred and twenty cases and 120 controls were included in this study. Cases
Accepted 24 February 2015
had been diagnosed with stroke by a neurologist, with the diagnosis confirmed by scanner.
Available online xxx
Controls had never had any type of stroke. Data were collected regarding sociodemographic characteristics, lifestyle behaviours, general history, type of stroke (ischaemic or
Keywords:
haemorrhagic) and periodontal parameters [plaque index, papillary bleeding index, pocket
Stroke
depth, clinical attachment loss, Community Periodontal Index of Treatment Needs and
Periodontal disease
periodontitis (defined by clinical attachment loss >2 mm and pocket depth >3 mm)]. Lo-
Melanodermic subject
gistic regression analysis was performed using R software to isolate a final model after
Senegal
adjustment for the 5% threshold. Results: All periodontal characteristics were more common among cases than among controls. Periodontitis (odds ratio 1.58, 95% confidence interval 1.1e3.022) and periodontal parameters were significantly associated with stroke, adjusted for hypertension, sedentary lifestyle, and the interaction between periodontitis and age. Conclusions: Periodontal disease is associated with stroke in the Senegalese population. Prospective longitudinal studies should be undertaken to improve understanding. © 2015 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
Introduction In its World Health Statistics 2008 report, the World Health Organization stated that stroke was the second most important cause of death around the world, accounting for 9.7% of
all recorded deaths. It is projected that this will increase to 12% by 2030.1 Stroke is a cosmopolitan condition with annual incidence of 145 new cases per 100,000 population in France, and prevalence of 250 cases per 100,000 population in Vietnam.2 The condition also constitutes a major public health challenge in developing countries, accounting for 45% of all
* Corresponding author. E-mail address: [email protected] (M. Diouf). http://dx.doi.org/10.1016/j.puhe.2015.02.033 0033-3506/© 2015 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Diouf M, et al., Stroke and periodontal disease in Senegal: caseecontrol study, Public Health (2015), http://dx.doi.org/10.1016/j.puhe.2015.02.033
2
p u b l i c h e a l t h x x x ( 2 0 1 5 ) 1 e5
deaths at the Department of Neurology of Cocody University Hospital, Abidjan in Ivory Coast, and is the principal condition and cause of death at the Neurology Clinic of Fann University Hospital, Dakar, Senegal.3 The extent and severity of stroke have resulted in much interest in its risk factors. A study involving 6000 subjects found that approximately 90% of cases of stroke can be attributed to 10 risk factors, including high blood pressure, physical inactivity and alcohol.4 Periodontal disease is polymicrobial, primarily inflammatory in nature and involves frequent formation of periodontal pockets. The prevalence of both stroke and periodontal disease is high in Africa,5 and the relationship between these two diseases has been reported previously.6 In Korea, Kim et al. found that periodontitis was an independent risk factor for stroke in men, non-diabetics and obese subjects.7 In Europe, particularly in Germany, Grau et al. found that patients with severe periodontitis (clinical attachment level >6 mm) had a 4.3-fold higher risk of ischaemic stroke compared with patients without periodontitis or with mild periodontitis (clinical attachment level 3 mm).8 A meta-analysis including eight cohort studies concluded that the risk of stroke was 19% higher in patients with periodontal disease.9 In Africa, studies focussing on the relationship between periodontal disease and stroke are very rare despite the magnitude and severity of this problem. For some diseases, being of Black ethnicity is a protective factor or ‘masks’ the association in melanodermic subjects. In Senegal, this relationship has not yet been documented. As such, it was hypothesized that there is an association between periodontal disease and stroke, and a caseecontrol study was conducted to test this hypothesis. The objective was to determine the periodontal factors associated with stroke in melanodermic subjects.
Methods This caseecontrol study took place in the neurological clinic of Fann National University Hospital. This is the only neurological clinic in Senegal. The clinic also receives patients from other countries.
Population and inclusion criteria The study population consisted of outpatients and inpatients attending the neurological clinic of Fann National University Hospital.
Case definition Cases were defined as patients who had been diagnosed with stroke by a neurologist, with the diagnosis confirmed by scanner.
Control definition Controls were defined as patients in the clinic who had not had any type of stroke.
Exclusion criteria Patients whose general condition did not allow a periodontal clinical examination (comatose patients), patients who were taking immunosuppressants or had received periodontal treatment within the last 6 months, and patients who had lost more than 75% of their teeth were excluded from the study.
Sample size and matching The sample size was calculated using EpiInfo 2000 (Centers for Disease Control and Prevention, Atlanta, GA, USA). With risk of 5% and power of 80%, and with reference to Diallo et al.,10 theoretical exposure to periodontal diseases of 32% was assumed among the controls. The risk of stroke for patients with periodontitis was fixed at 2. Thus, the sample size was 120 cases and 120 controls. The groups were matched in terms of age and sex in order to minimize the influence of these factors.
Collection procedure and variables Data were collected using a questionnaire that had been validated and used previously for the same purpose. This questionnaire was pretested on 10 patients attending the neurosurgery clinic in the hospital, to enable the authors to observe respondents' reactions to the survey and to estimate the time spent on various sections of the questionnaire. Data collection took place every day, and included all patients who came into the clinic who met the selection criteria. Data collection took place from April to July 2011. The information collected by questionnaire included sociodemographic characteristics (age, occupation, income, weight, size, body mass index) and lifestyle behaviours [smoking (active: 5 cigarettes/day; passive: 3 cups/day), drinking alcohol (>3 glasses/day), chewing cola seeds and activity (sedentary or physical)]. Medical history covered diabetes, depression, infection and hypertension. Information on periodontal factors included plaque, papillary bleeding index, pocket depth, attachment loss and Community Periodontal Index of Treatment Needs (CPITN). One week after the stroke, the patient or their relative/ friend was interviewed, and data collection was completed by a periodontal clinical examination of the patient.
Clinical observation and evaluation of periodontal data Evaluation of the level of hygiene of the gingival condition, extent of bleeding, clinical attachment loss and pocket depth was performed according to the recommendations of Martin and Bercy.11 Average values were calculated, first for individuals, by calculating the average value per tooth in the individual, adding the numbers together and dividing by the number of teeth scored (individual score). The average for the sample was calculated as the average value of all individual scores for the members of the sample.
Please cite this article in press as: Diouf M, et al., Stroke and periodontal disease in Senegal: caseecontrol study, Public Health (2015), http://dx.doi.org/10.1016/j.puhe.2015.02.033
3
p u b l i c h e a l t h x x x ( 2 0 1 5 ) 1 e5
Exposure measurement Exposure to risk factors was assessed from the independent variables (i.e. sociodemographic variables, medical history and periodontitis). A complete examination instrument tray, including a graduated periodontal World Health Organization probe, was used for periodontal measurements. Clinical examinations were undertaken by a graduate dental student, previously trained in these types of measurement. Periodontitis was diagnosed in patients with clinical attachment loss >2 mm at least two sites and pocket depth of >3 mm.12
Outcome measurement The outcome measure of interest was occurrence of stroke, and this was diagnosed using a cerebral scanner, manipulated by a neurologist.
Ethical considerations For ethical reasons, prior to the administration of the questionnaire, the objectives and the importance of the investigation were explained to the patient and their relative/friend to obtain their consent to participate. At the end of the clinical examination, individuals benefited from a session of motivation, including oral hygiene instruction and calculus removal.
were employed, compared with 28.3% of controls. Monthly income was 100,000 CFA (152 V) for 13.3% of cases vs 18.3% of controls. Among the cases, 8.3% were smokers and 27.5% were passive smokers. More than 60% of cases had sedentary occupations and drank coffee, 8.3% drank alcohol and 41.7% chewed cola seeds, a local postprandial habit. In terms of medical history, 18.3% of cases reported diabetes and 76.7% reported hypertension. Diabetes, history of infection and hypertension were reported by 10.8%, 0.8% and 22.5% of controls, respectively. Table 1 shows that indicators of periodontal disease were more prevalent in cases than controls. Periodontitis was recorded in 73.3% of cases compared with 40.8% of controls. Clinical attachment loss was 2.0 (SD 1.657) mm in cases and 1.0 (0.913) mm in controls, and pocket depth was 2.7 (SD 0.673) mm in cases and 2.4 (SD 0.510) mm in controls. The interactions between periodontitis and age, periodontitis and smoking, and periodontitis and alcohol were tested. Only the interaction between periodontitis and age was statistically significant. Table 2 presents the final model used. Periodontitis was found to be significantly associated with stroke (P ¼ 0.021), attachment loss (P ¼ 0.005), the interaction between periodontitis and age (P ¼ 0.008), physical inactivity (P ¼ 0.01), plaque index [OR 10.732 (95% CI 3.779e30.475)], pocket depth [OR 1.776 (95% CI 1.082e4.438)] and hypertension [OR 7.711 (95% CI 3.558e16.711)].
Data analysis strategy Data were entered using EpiInfo 2000 Version 5.3.1. Univariate descriptive analysis was undertaken with the same software. This provided proportions and means or medians of the studied variables. Some variables were dichotomized. The associations between categorical variables were studied using Chi-squared test, and their importance was determined by calculating odds ratios (ORs) with confidence intervals (CIs). Modelling of multiple logistic regression with a backward selection procedure was conducted for variables with P < 0.25 on univariate analysis. At each stage, the P-value associated with the variable was eliminated from the model at the 5% threshold. The procedure was stopped when all of the variables in the model had a P-value less than the threshold. In an attempt to control confounding, the analysis checked the variation of the main independent variable ‘periodontitis’ for each output variable by calculating the relative change. Interactions with periodontitis were tested using R Version 2.9.0 (R Core Development Team, Vienna, Austria). This analytical strategy helped to identify the independent variables that were significantly associated with the dependent variable (stroke). P ¼ 0.05 was taken to indicate marginal significance, and P < 0.05 was taken to indicate significance.
Results Mean weight was 65.3 [standard deviation (SD) 13.6] kg for cases and 64.3 (SD 15.1) kg for controls. The average body mass index of cases and controls was approximately equal [23.0 (SD 3.8) and 23.2 (SD 9.3), respectively]. Of the 120 cases, 17.5%
Discussion Periodontal disease and various periodontal parameters were significantly associated with stroke, adjusted for hypertension, physical inactivity, and the interaction between periodontitis and age.
Limitations and methodological considerations The quality of this study could have been increased if ischaemic and haemorrhagic strokes had been analysed separately, or the study population had been restricted to a single type of stroke. This may have increased the precision of the results, but the adjustments made in this study should have neutralized this confounding factor. In addition, in the
Table 1 e Periodontal indicators: mean values in cases and controls. Variable
Case (SD) Control (SD) P-value
Periodontal characteristics Plaque index Bleeding index Clinical attachment loss (mm) Pocket depth (mm) CPITN Periodontitis present
1.9 (0.404) 0.5 (0.374) 2.0 (1.657) 2.7 (0.673) 2.3 (0.744) 73.3%
1.4 (0.371) 0.3 (0.287) 1.0 (0.913) 2.4 (0.510) 1.7 (0.498) 40.8%
Available online at www.sciencedirect.com
Public Health journal homepage: www.elsevier.com/puhe
Original Research
Stroke and periodontal disease in Senegal: caseecontrol study M. Diouf a,*, A. Basse b, M. Ndiaye b, D. Cisse c, C.M. Lo c, D. Faye c a
UCAD, Dakar, Senegal Neurology Department, Dakar, Senegal c Medicine Pharmacy and Dentistry Faculty, Dakar, Senegal b
article info
abstract
Article history:
Objective: To determine the periodontal factors associated with stroke in melanodermic
Received 23 August 2014
subjects in Senegal.
Received in revised form
Study design: Caseecontrol study.
16 December 2014
Methods: One hundred and twenty cases and 120 controls were included in this study. Cases
Accepted 24 February 2015
had been diagnosed with stroke by a neurologist, with the diagnosis confirmed by scanner.
Available online xxx
Controls had never had any type of stroke. Data were collected regarding sociodemographic characteristics, lifestyle behaviours, general history, type of stroke (ischaemic or
Keywords:
haemorrhagic) and periodontal parameters [plaque index, papillary bleeding index, pocket
Stroke
depth, clinical attachment loss, Community Periodontal Index of Treatment Needs and
Periodontal disease
periodontitis (defined by clinical attachment loss >2 mm and pocket depth >3 mm)]. Lo-
Melanodermic subject
gistic regression analysis was performed using R software to isolate a final model after
Senegal
adjustment for the 5% threshold. Results: All periodontal characteristics were more common among cases than among controls. Periodontitis (odds ratio 1.58, 95% confidence interval 1.1e3.022) and periodontal parameters were significantly associated with stroke, adjusted for hypertension, sedentary lifestyle, and the interaction between periodontitis and age. Conclusions: Periodontal disease is associated with stroke in the Senegalese population. Prospective longitudinal studies should be undertaken to improve understanding. © 2015 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
Introduction In its World Health Statistics 2008 report, the World Health Organization stated that stroke was the second most important cause of death around the world, accounting for 9.7% of
all recorded deaths. It is projected that this will increase to 12% by 2030.1 Stroke is a cosmopolitan condition with annual incidence of 145 new cases per 100,000 population in France, and prevalence of 250 cases per 100,000 population in Vietnam.2 The condition also constitutes a major public health challenge in developing countries, accounting for 45% of all
* Corresponding author. E-mail address: [email protected] (M. Diouf). http://dx.doi.org/10.1016/j.puhe.2015.02.033 0033-3506/© 2015 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Diouf M, et al., Stroke and periodontal disease in Senegal: caseecontrol study, Public Health (2015), http://dx.doi.org/10.1016/j.puhe.2015.02.033
2
p u b l i c h e a l t h x x x ( 2 0 1 5 ) 1 e5
deaths at the Department of Neurology of Cocody University Hospital, Abidjan in Ivory Coast, and is the principal condition and cause of death at the Neurology Clinic of Fann University Hospital, Dakar, Senegal.3 The extent and severity of stroke have resulted in much interest in its risk factors. A study involving 6000 subjects found that approximately 90% of cases of stroke can be attributed to 10 risk factors, including high blood pressure, physical inactivity and alcohol.4 Periodontal disease is polymicrobial, primarily inflammatory in nature and involves frequent formation of periodontal pockets. The prevalence of both stroke and periodontal disease is high in Africa,5 and the relationship between these two diseases has been reported previously.6 In Korea, Kim et al. found that periodontitis was an independent risk factor for stroke in men, non-diabetics and obese subjects.7 In Europe, particularly in Germany, Grau et al. found that patients with severe periodontitis (clinical attachment level >6 mm) had a 4.3-fold higher risk of ischaemic stroke compared with patients without periodontitis or with mild periodontitis (clinical attachment level 3 mm).8 A meta-analysis including eight cohort studies concluded that the risk of stroke was 19% higher in patients with periodontal disease.9 In Africa, studies focussing on the relationship between periodontal disease and stroke are very rare despite the magnitude and severity of this problem. For some diseases, being of Black ethnicity is a protective factor or ‘masks’ the association in melanodermic subjects. In Senegal, this relationship has not yet been documented. As such, it was hypothesized that there is an association between periodontal disease and stroke, and a caseecontrol study was conducted to test this hypothesis. The objective was to determine the periodontal factors associated with stroke in melanodermic subjects.
Methods This caseecontrol study took place in the neurological clinic of Fann National University Hospital. This is the only neurological clinic in Senegal. The clinic also receives patients from other countries.
Population and inclusion criteria The study population consisted of outpatients and inpatients attending the neurological clinic of Fann National University Hospital.
Case definition Cases were defined as patients who had been diagnosed with stroke by a neurologist, with the diagnosis confirmed by scanner.
Control definition Controls were defined as patients in the clinic who had not had any type of stroke.
Exclusion criteria Patients whose general condition did not allow a periodontal clinical examination (comatose patients), patients who were taking immunosuppressants or had received periodontal treatment within the last 6 months, and patients who had lost more than 75% of their teeth were excluded from the study.
Sample size and matching The sample size was calculated using EpiInfo 2000 (Centers for Disease Control and Prevention, Atlanta, GA, USA). With risk of 5% and power of 80%, and with reference to Diallo et al.,10 theoretical exposure to periodontal diseases of 32% was assumed among the controls. The risk of stroke for patients with periodontitis was fixed at 2. Thus, the sample size was 120 cases and 120 controls. The groups were matched in terms of age and sex in order to minimize the influence of these factors.
Collection procedure and variables Data were collected using a questionnaire that had been validated and used previously for the same purpose. This questionnaire was pretested on 10 patients attending the neurosurgery clinic in the hospital, to enable the authors to observe respondents' reactions to the survey and to estimate the time spent on various sections of the questionnaire. Data collection took place every day, and included all patients who came into the clinic who met the selection criteria. Data collection took place from April to July 2011. The information collected by questionnaire included sociodemographic characteristics (age, occupation, income, weight, size, body mass index) and lifestyle behaviours [smoking (active: 5 cigarettes/day; passive: 3 cups/day), drinking alcohol (>3 glasses/day), chewing cola seeds and activity (sedentary or physical)]. Medical history covered diabetes, depression, infection and hypertension. Information on periodontal factors included plaque, papillary bleeding index, pocket depth, attachment loss and Community Periodontal Index of Treatment Needs (CPITN). One week after the stroke, the patient or their relative/ friend was interviewed, and data collection was completed by a periodontal clinical examination of the patient.
Clinical observation and evaluation of periodontal data Evaluation of the level of hygiene of the gingival condition, extent of bleeding, clinical attachment loss and pocket depth was performed according to the recommendations of Martin and Bercy.11 Average values were calculated, first for individuals, by calculating the average value per tooth in the individual, adding the numbers together and dividing by the number of teeth scored (individual score). The average for the sample was calculated as the average value of all individual scores for the members of the sample.
Please cite this article in press as: Diouf M, et al., Stroke and periodontal disease in Senegal: caseecontrol study, Public Health (2015), http://dx.doi.org/10.1016/j.puhe.2015.02.033
3
p u b l i c h e a l t h x x x ( 2 0 1 5 ) 1 e5
Exposure measurement Exposure to risk factors was assessed from the independent variables (i.e. sociodemographic variables, medical history and periodontitis). A complete examination instrument tray, including a graduated periodontal World Health Organization probe, was used for periodontal measurements. Clinical examinations were undertaken by a graduate dental student, previously trained in these types of measurement. Periodontitis was diagnosed in patients with clinical attachment loss >2 mm at least two sites and pocket depth of >3 mm.12
Outcome measurement The outcome measure of interest was occurrence of stroke, and this was diagnosed using a cerebral scanner, manipulated by a neurologist.
Ethical considerations For ethical reasons, prior to the administration of the questionnaire, the objectives and the importance of the investigation were explained to the patient and their relative/friend to obtain their consent to participate. At the end of the clinical examination, individuals benefited from a session of motivation, including oral hygiene instruction and calculus removal.
were employed, compared with 28.3% of controls. Monthly income was 100,000 CFA (152 V) for 13.3% of cases vs 18.3% of controls. Among the cases, 8.3% were smokers and 27.5% were passive smokers. More than 60% of cases had sedentary occupations and drank coffee, 8.3% drank alcohol and 41.7% chewed cola seeds, a local postprandial habit. In terms of medical history, 18.3% of cases reported diabetes and 76.7% reported hypertension. Diabetes, history of infection and hypertension were reported by 10.8%, 0.8% and 22.5% of controls, respectively. Table 1 shows that indicators of periodontal disease were more prevalent in cases than controls. Periodontitis was recorded in 73.3% of cases compared with 40.8% of controls. Clinical attachment loss was 2.0 (SD 1.657) mm in cases and 1.0 (0.913) mm in controls, and pocket depth was 2.7 (SD 0.673) mm in cases and 2.4 (SD 0.510) mm in controls. The interactions between periodontitis and age, periodontitis and smoking, and periodontitis and alcohol were tested. Only the interaction between periodontitis and age was statistically significant. Table 2 presents the final model used. Periodontitis was found to be significantly associated with stroke (P ¼ 0.021), attachment loss (P ¼ 0.005), the interaction between periodontitis and age (P ¼ 0.008), physical inactivity (P ¼ 0.01), plaque index [OR 10.732 (95% CI 3.779e30.475)], pocket depth [OR 1.776 (95% CI 1.082e4.438)] and hypertension [OR 7.711 (95% CI 3.558e16.711)].
Data analysis strategy Data were entered using EpiInfo 2000 Version 5.3.1. Univariate descriptive analysis was undertaken with the same software. This provided proportions and means or medians of the studied variables. Some variables were dichotomized. The associations between categorical variables were studied using Chi-squared test, and their importance was determined by calculating odds ratios (ORs) with confidence intervals (CIs). Modelling of multiple logistic regression with a backward selection procedure was conducted for variables with P < 0.25 on univariate analysis. At each stage, the P-value associated with the variable was eliminated from the model at the 5% threshold. The procedure was stopped when all of the variables in the model had a P-value less than the threshold. In an attempt to control confounding, the analysis checked the variation of the main independent variable ‘periodontitis’ for each output variable by calculating the relative change. Interactions with periodontitis were tested using R Version 2.9.0 (R Core Development Team, Vienna, Austria). This analytical strategy helped to identify the independent variables that were significantly associated with the dependent variable (stroke). P ¼ 0.05 was taken to indicate marginal significance, and P < 0.05 was taken to indicate significance.
Results Mean weight was 65.3 [standard deviation (SD) 13.6] kg for cases and 64.3 (SD 15.1) kg for controls. The average body mass index of cases and controls was approximately equal [23.0 (SD 3.8) and 23.2 (SD 9.3), respectively]. Of the 120 cases, 17.5%
Discussion Periodontal disease and various periodontal parameters were significantly associated with stroke, adjusted for hypertension, physical inactivity, and the interaction between periodontitis and age.
Limitations and methodological considerations The quality of this study could have been increased if ischaemic and haemorrhagic strokes had been analysed separately, or the study population had been restricted to a single type of stroke. This may have increased the precision of the results, but the adjustments made in this study should have neutralized this confounding factor. In addition, in the
Table 1 e Periodontal indicators: mean values in cases and controls. Variable
Case (SD) Control (SD) P-value
Periodontal characteristics Plaque index Bleeding index Clinical attachment loss (mm) Pocket depth (mm) CPITN Periodontitis present
1.9 (0.404) 0.5 (0.374) 2.0 (1.657) 2.7 (0.673) 2.3 (0.744) 73.3%
1.4 (0.371) 0.3 (0.287) 1.0 (0.913) 2.4 (0.510) 1.7 (0.498) 40.8%
