Anxiety a possible peripheral artery disease risk factor among older Africans: The EPIDEMCA study.

Objective During the past ten years considerable progress has been made in research on cardiac effects of anxiety. On the other hand, epidemiological research on the implication of this psychosocial factor in the development of peripheral artery disease (PAD) is lacking. The present study sought to examine the association of anxiety and PAD in elderly individuals in two sub-Saharan countries in Africa. Methods This cross-sectional study, was performed in urban and rural population in Central Africa: Republic of Congo and Central African Republic. All individuals ≥65 years living in these areas were invited to participate. Socio-demographic data, medical history, medication, clinical and biological data were collected. Anxiety symptoms were assessed with the Geriatric Mental State version B3. Peripheral artery disease was defined as ankle brachial index (ABI) ≤0.90. Results A total of 1662 subjects composed our study population (age 72.9±6.5 years, 59.3% females). Overall, the prevalence of anxiety was 7.2%, significantly higher in women (9.2% vs. 4.3% in men, p<0.001 ). Independently of all confounding factors, anxiety was strongly associated with PAD in both countries (OR: 2.56, 99%CI: .52-4.20) and more frequently associated with PAD in women than in men (OR: 3.37, 95%CI: 1.22-4.10 vs. OR: 2.34, 95%CI: 1.13-5.06). Conclusion Our study underlines the high prevalence of anxiety in African general population in the elderly, as well as a significant association with the peripheral artery disease, independently of all possible confounding factors. The higher rates of anxiety in women and stronger association with PAD as compared to men might explain the higher rates of PAD in African women.

Anxiety, a sustained mental health disorder, can be triggered by stress when the adaptive capacities of individuals to their environment are perceived as strained or exceeded. In contrast to stress, once the threat is mediated, anxiety does not fade into time and cause significant impairment in daily life and organism physiology.
Many studies have shown an interaction between anxiety and hypertension, coronary artery disease (CAD), and even outcomes after cardiac surgery [1][2][3][4] . Considerable progress has been made during the past decade in research on effects of this psychosocial factor, but data are still limited to CAD, stroke and hypertension. The pooled relative risk of acute coronary syndrome in case of prior anxiety is 3.1 in cross sectional studies 5 . In the Whitehall II study, anxiety was also associated with components of metabolic syndrome, such as higher waist circumference (central obesity), obesity, and dyslipidaemia 6 .
Peripheral artery disease (PAD) is a frequent presentation of atherosclerotic disease and shares the same major risk factors as CAD. Nevertheless, no specific information is available regarding the relationship between anxiety and PAD. Furthermore, the prevalence of PAD seems to be different according to ethnicity., Individuals of African descent in the United States are at up to 3-fold higher risk of PAD as compared to non-Hispanic Whites 7 and recent studies highlighted the higher prevalence of PAD in native African populations [8][9][10] . The Global Burden Disease project highlighted the burden of PAD in low-and-middle income countries, especially Africa, with higher prevalence of this condition among women, without a clear explanation on why this sex disparity is more pronounced in these countries than high-income ones 11 .
As findings from previous studies have shown that psychosocial factors such as anxiety may be associated with the development of cardiovascular disease, the present study sought to examine the association of anxiety and PAD in a cohort of elderly individuals in 4 two sub-Saharan countries in Africa.

Study Design and Population
The EPIDEMCA study (Epidemiology of Dementia in Central Africa) is a cross-sectional population-based study in rural and urban areas of two countries of Central Africa: Central African Republic (CAR) and Republic of Congo (ROC), focused on cardiovascular diseases and risk factors, cognitive disorders and their interrelation.
The study population included all people aged ≥65 years. Exclusion criterion was refusal to participate. Written consent was obtained or confirmed by fingerprint marks for illiterate subjects. In case of severe cognitive impairment, the purpose of the study was verbally explained to the person and his family. Sampling was proportional to the size of the main subdivisions in the urban areas (Brazzaville in ROC, Bangui in CAR) and an exhaustive door-to-door sampling was applied in two rural areas (Gamboma in ROC, Nola in CAR). Additional details pertaining to study design have been published elsewhere 12 .
Ethical committees in CAR and ROC approved the study protocol, as well as the "Comité de la Protection des Personnes Sud-Ouest Outre-Mer" in France (SOOM4-CE-3).

Data collection
Psychosocial and cognitive measurements As anxiety symptoms may span the cognitive and behavioural fields, assessment of cognitive status was achieved using the Community Screening Interview for Dementia (CSI-D) 13 adapted, back-translated and pretested in the local languages (Sango in CAR, Lari, Lingala, and Kituba in Congo).
Symptoms of anxiety and depression were studied for their potential role as CV risk factors. The psychosocial status of each participant was self-reported by a coordinator-5 administered questionnaire.
Depressive and anxiety symptoms were assessed with the Geriatric Mental State version B3 14 . Neurologists have performed neurological examination during which history of stroke and depressive disorders were sought.
Peripheral artery disease assessment Peripheral artery disease was defined as ABI ≤0.90. The ABI was measured using a manual cuff and a hand-held Doppler device (Super Dopplex II, Huntleigh Technology PLC, Luton, UK) to determine systolic blood pressure in the posterior tibial and dorsal pedis arteries and humeral arteris bilaterally. Measurements followed the protocol used in our core laboratory in France, in accordance to the American Heart Association guidelines 15,16 Covariates Epidemiological data were collected following a standardized questionnaire.
Participants' systolic blood pressure (SBP), hypertension medication use (yes/no), current smoking behaviour (never, past -former smoker who quit smoking more than one year prior to the survey/ current smoking), body-mass index (BMI), diabetes (yes/no), dyslipidaemia, and physical activity were included as covariates in the multivariable analyses. Physical activity was measured in minutes per week using the Typical Week's Physical Activity Questionnaire 18 . Definitions and measurements methods have been detailed in a previous publication 8 .
Alcohol drinking was assessed based on the frequency and amount of alcohol intake in a typical drinking week, and was categorized as never, occasional (less than 5 days/week) and regular (more than 5 days/week).
Marital status data, education (school cycles completion) and occupation were collected.
Covariates were selected because of their role as potential confounders in the association 6 between psychosocial factors and PAD 19 . These same covariates were used in the study by Ohira et al., which studied the association between anger, anxiety, and depression on subclinical atherosclerosis in non-Hispanic Whites, African American, Hispanic, and Chinese populations in the MESA study 19 .

Statistical Analysis
Descriptive statistics (chi-square, t-test) were used to compare baseline demographic data, and to compare differences between PAD and non-PAD subjects for all psychosocial measures, and covariates, in both countries. Linear regression analysis was performed to determine the relationship between psychosocial variables and PAD. The multivariate model was adjusted for age, sex, country, urban and rural areas, traditional CV risk factors (hypertension, smoking, BMI, diabetes, cholesterol, physical activity), socioeconomic factors (education, marital status, occupation), cognitive status and depression.
Statistical analyses were carried out using Statview 5.0 software (SAS Institute, Cary, USA).

Results
Among 1871 participants (939 in CAR and 932 in ROC), cognitive and psychological data were lacking for 209 subjects. A total of 1662 subjects (832 in CAR and 830 in ROC) composed our study population; 59.3 % were females ( Table 1). The mean age of the 209 excluded subjects was 74.9±7.1 years (vs 72.9±6.5 years in those included in this study).
The rates of female sex were also comparable (58.2% vs 59.3% in our study population).
Among patients excluded from the current analysis, the prevalence of hypertension, diabetes, dyslipidaemia and obesity did not differ as compared to those included in this study.
7 Mean age of the study population was significantly higher in PAD subjects but no difference was observed between CAR and ROC. The prevalence of hypertension, diabetes and obesity in our study population was higher in ROC, while subjects in CAR smoked tobacco and drank alcohol more. The general characteristics of the total population, in both countries as well as comparisons between PAD and no-PADsubjects are presented in Table 1. The prevalence of PAD in women and men was 16% and 11.7%, respectively.
Female sex, hypertension, dyslipidaemia, lower education, cognitive disorders, anxiety but not depression were more frequently associated with PAD. Overall, the prevalence of anxiety was 7.2%, significantly higher in women (9.2% vs. 4.3% in men, p<0.001)..
Prevalence of PAD and anxiety in each country and gender are presented in Figure 1.
Anxiety was more frequently associated with PAD in women than in men ( Figure 2).
In multivariate analysis, adjusted to age, sex, country, urban or rural areas, all traditional CV risk factors, socioeconomic status and associated cognitive disorders, PAD was significantly associated with anxiety but not depression. (Figure 3).

Discussion
In this population study composed of older Africans living in two sub-Saharan countries, we found a high prevalence of anxiety, especially in women and underlined the strong association between anxiety and PAD independently of age, country, all traditional CV risk factors, socioeconomic status and associated cognitive disorders.
Anxiety is known to be one of the most common mental health disorders worldwide. Data suggest that its prevalence is higher in western countries and particularly in the United States, affecting 9% of the elderly population 20 . A recent systematic review on the prevalence of anxiety disorders in adult populations 21 found the lowest prevalence in East Asia (2.8%, 95% CI: 2.2-3.4) and the highest in North America (7.7%, 95% CI: 6.8-8.8). It 8 also analysed data in North African/Middle Eastern region (7.7%, 95% CI: 6.0-10), Switzerland, Korea, Pakistan and Germany but no data were available in Sub-Saharan countries. Our study underlines the high prevalence of this disorder in general elderly African population, higher than in western countries and particularly high in women.
The association between anxiety and cardiovascular diseases, especially CAD is demonstrated in cross-sectional, longitudinal, and prognostic studies with an estimated prevalence of anxiety substantially greater in CAD populations (7.9%) than in general population (3.1%) 22 . Our study also confirms the higher prevalence of anxiety in PAD patients compared to the general population (13.4% vs. 6.2%). No geographical difference for the association of anxiety with PAD was found between in CAR or ROC.
Importantly, while previous studies described a similar prevalence of depression and anxiety in population with acute coronary syndrome 23,24 and after coronary artery bypass graft surgery 25 , depression disorders seem to be less common than anxiety in our PAD population. Anxiety disorders are comorbid with depression disorders in approximately 50 % of CAD cases 5,26 . In our study only 9.2% of PAD subjects had concomitant anxiety and depression.
Evidence is mixed on possible sex differences in CV risk though anxiety and depression are more commonly reported among females than males. Results are still controversial: studies report incident rate of CAD consistent among both males and females 27 while others found a significant association between anxiety and CV disease occurrence in men (OR = 1.10, 95 % CI 1.05-1.15) but not women 28 . In our study, interactions of PAD with anxiety were significant in both sexes but higher in women. This might explain the highest prevalence of PAD in female African population.
Pathways explaining the association between anxiety and cardiovascular disorders are less known than for depression. Findings from previous studies suggested that psychosocial factors such as anxiety and depression may be associated with the development of CV disease through unhealthy coping behaviours such as smoking, sedentary lifestyle and caloric intake 29,30 . Adverse metabolic changes are also one of the underlying plausible mechanisms, such as abdominal obesity and dyslipidaemia, that are related to psychiatric disorders 31 . Nevertheless, our study demonstrates a strong association between anxiety and PAD, independently of all CV risk factors or lifestyle behaviours and suggests a possible causal relationship.
One hypothesis that might explain this association suggests that sympathetic nervous overactivity in depressed and anxious persons can mediate the association with atherosclerosis, e.g., through elevated resting and 24-hour heart rate and reduced heart rate variability 32 . A major challenge over the next years is to incorporate anxiety processes into the mainstream of cardiovascular pathophysiological research.
Our study has several limitations. The first is related to its cross-sectional design which might lead to the assumption that anxiety was triggered by PAD. It is however not plausible that PAD induces anxiety as most participants were not aware of their condition and had PAD diagnosed through the measurement of ABI. Second, 209 subjects were excluded of the study. Nevertheless, as described above, no different characteristics were found between the study population and the subjects excluded from these analyses. A "selection bias" must also be considered in studies of older persons as those with severe atherosclerosis might not live to older age and women have a higher life expectancy.
The strengths of our study are the large sample of participants (the largest populationbased study in Central Africa), the rigours sampling covering rural and urban zones, the exhaustive data collection and particularly the mental state assessment allowing an 10 accurate diagnosis of the psychosocial factors.

Conclusions
Our study underlines the high prevalence of anxiety in African general population in the elderly, as well as a significant association with the peripheral artery disease, independently of all possible confounding factors. The higher rates of anxiety in women and stronger association with PAD as compared to men might explain the higher rates of