We conducted cross-sectional community-based study carried out thought home questionnaires, with a random sampling process of Monastir's non institutionalized elderly individuals (aged more than 65 years). We used a probabilistic multistage cluster sampling technique in two phases. The first consisted in the identification of the districts; eighteen sectors were selected, than we selected randomly a cluster of household with a probability of selection proportional to the size of their population. We identified elderly subjects within the household clusters. At the selected homes one or two resident was interviewed. The participants were previously informed about this study by social assistant of the region to maximize response rates. Three doctors performed a door-to-door survey of the participant households. Any elderly subject refusing to take part in the study will be granted the freedom to do so. The investigator also respects the decision of the old person's attendants and relatives if they express the wish not to allow the elderly person to take part in this study. In all cases, informed consent was obtained from subjects or cohabiting. No response subjects (n = 22) were replaced by others eligible subjects, a next neighbor was chosen in the same way. The parameters used to calculate the sample size were the proportion of elderly, maximum allowed 95% confidence interval with a margin of error no more than 3%. An extra of 50% were added to the sample obtained (n = 400) to cover losses, resulting in n = 600. A total of 598 individuals were investigated. The exclusion criteria were individuals aged less than 65 years. Any elderly present, the day of the visit by investigators, in the surveyed household found to be a visitor or do not usually leave there will be excluded.

Information was gathered by home-based personal interview using a structured questionnaire with 90 parameters conducted in three parts: standard questions on demographic, socioeconomic information and health behaviors, followed by a physical examination: anthropometric and blood pressure measurements. The questionnaire was translated from French into Arab by the principal investigator and another Tunisian physician and compared before being distributed. Before participation respondents were informed that participation was voluntary, and that they could quit out at any time.

Blood pressure (BP) was measured twice using standard sphygmomanometer, with participants seated after 5 minutes. For analysis we used the mean of the 2 measurements. High BP was defined as systolic blood pressure (SBP) ≥ 140 mmHg and/or diastolic pressure (DBP) ≥ 90 mm Hg 7 . Arterial hypertension (AH) was defined as having high BP or using antihypertensive drug therapy in the previous 2 weeks. Awareness of hypertension reflected a previous diagnosis of hypertension. Controlled hypertension was defined as treated hypertension with SBP < 140 mmHg and DBP < 90 mmHg. Anthropometric measurements were obtained with the participants wearing light clothing and no footwear. Height was measured without shoes, to the nearest 0.1 cm, with subjects standing fully erect on a flat surface, tocks and shoulders flat to the wall and looking straight ahead. Weight was to the nearest 0.1 kg using digital scale. Body Mass Index (BMI) was calculated as weight (in kilograms) divided by squared height (in meters squared) and further. The BMI classified individuals, using World Health Organization criteria, as normal (< 25 kg/m²), Overweight (25-29.9 kg/m²) and obese (≥ 30 kg/m²) 8 . Waist circumference was measured to the nearest centimeter at the level of the midpoint between the inferior border of the ribs and the iliac crest in the midaxillary line, at the end of expiration, using a constant tension tape, directly over light clothing. The waist circumference (WC) normal range was that recommended by the National Cholesterol Education Program ATP III. The WC < 88 cm for women and < 102 cm for men was classified as normal 9 . The diagnosis of diabetes mellitus was established by history or use of glucose-lowering drugs in the previous 2 weeks 10 . Marital status was categorized as married or single, divorced, widows, widowers and never married persons were included in the single group. Smoking status (current smokers or not), and physical activity (walking for at least 1 h per day or not). Education was recorded on four levels scale: (illiterate, kuttabs 'koranic schools', primary, secondary or more). The diagnosis of disability was made by assessing limitations on basic activities of daily living according to Katz's test, was defined as the inability to do one or more of the following without help: bathe, eat, dress, transfer from a bed to a chair, use the toilet, or walk across a small room. Depression symptoms were assessed using a mini-Geriatric Depression Scale mini-GDS, a mini-GDS Scale ≥ 1 was considered indicative of depression 11 .

We built in several quality and control measures into our survey protocol, to ensure completeness and comparability of blood pressure and anthropometric measurements and interviewee responses. For this all three doctors, underwent a common training program at the National Institute of Public Health at which a team of cardiologists instructed the investigators regarding measured of BP, weight and height. The questionnaire items were tested for clarity and validity.

The data were collected in a database and transferred to our coordinating center for statistical analysis. Every month a meeting was organised for all participating investigators to discuss the progress of the study. A total of 598 subjects were analysed using SPSS13 statistical software packages. Results are expressed as mean with the corresponding standard deviations. For statistical significance of the difference between two mean values, a Student t test was used. Comparison of frequencies was performed by chi2 tests with the Yates correction when needed. Multivariate analysis between statistically significant variables and hypertension (dichotomous dependent variable) was performed using a logistic regression model. The independent variables were modelled categorically, using dummy terms.

Table 1 gives a breakdown of the sample characteristics, there is more female than male (66% VS. 34%) and mean age was 72.3 ± 7.4 years. Most subjects were aged < 70 years (38%) but only 18% were over 80 years of age. The population under study was predominantly urban (more than three quarters). Only about 14% are from rural sites. Approximately three quarters of the subjects had no formal education and only 5% had secondary or higher education. Most of our population are still married (60%) and 61% are inactive. Only 35% are current smokers. Smoking prevalence is notably high for male (98%) compared to female (2.3%). Health insurance covered more than half of our population (57%).

The prevalence of hypertension was 52% (n = 311) and was predominant in women (women 55.5%, men 45%, p < 0.01). The prevalence of hypertension increased with age until 80 years then decrease. At age less than 70 years, there are 51.5% compared with 55% of those aged 75 to 79 years and only 45.8% of those aged more than 80 years (Table 2). The mean systolic blood pressure values were 135.2 ± 17.0 mmHg (85 - 21.5 mmHg), while mean diastolic blood pressure values were 80.0 ± 10.2 mmHg (5.5 - 12.5 mmHg). Isolated systolic hypertension was the most common form of hypertension (50.9%) of all hypertensive subjects who had elevated blood pressure at examination, followed by combined systolic and diastolic hypertension (45.7%) and isolated diastolic hypertension (3.4%).

The proportion of subjects aware of their elevated blood pressure status was 81% (n = 252), the majority (78.4%, n = 244) of these were taking pharmacological treatment for hypertension. However, only about 30.7% (n = 75) of the treated subjects had their blood pressure controlled to normal levels according to current JNC-VII recommendation 7 .

Tables 2 examine the frequency of socio-demographic and cardiovascular risk in normotensive and hypertensive participants. Schooling showed an inverse association but not significantly (p = 0.08) with hypertension, with a prevalence of 54% among analphabetic and 31.2% among those with than secondary education level. No urban/rural differences were observed. Depression, marital status, and self reported physical activity were not correlated with the prevalence of hypertension. Smoking status is negatively but not significantly associated with AH (results not shown).

Those with hypertension had higher BMI, waist circumference, and diabetes (p < 0.05). Regarding disability, the individuals who presented the highest prevalence of hypertension were more dependent. In stepwise multivariable regression analysis, the presence of obesity, diabetes, as well as disability was independently associated with hypertension (Table 3).

This study provides important new evidence on the prevalence of hypertension in a representative sample of an elderly population in a geographically well defined Mediterranean area of the Tunisia center. Moreover, it has already established socio-demographic variables and drug consumption. A peculiar aspect of our population is the low level of education (89% for illiterates plus subjects with kuttabs education).

We observed that more than 50% of elderly in Monastir city suffer from hypertension. Since the use a BP value of 140/90 mmHg as cut-off, the prevalence of hypertension has increased. These percentage values have also been found in studies carried out in North African countries such as Morocco 13 , and other Arab countries such as Egypt 14 . Laouani reported a prevalence rate of 69.3% 15 but in 2000, Kammoun et al. reported in a representative sample of the Tunisian elderly population the prevalence of 32% 16 . The increasing prevalence of hypertension among the elderly population of Monastir city may be attributed to change in lifestyle in the past decade. However, the prevalence of hypertension is still considerably lower than in other industrialised countries such as France (79.8%), England (62%), Greece (65.4%), Spain (62%), Italy (76.3%) and United States (84%) 17 18 19 20 21 22 .

This study showed clear gender differences in the prevalence of hypertension, with women being more likely to be hypertensive than men. The gender differences in the prevalence of hypertension have been reported in many studies 21 22 23 . The present study found that hypertension prevalence declined after 80 years, the difference might partly be due to the shorter survival of people with AH and a low incident in the older age group. As demonstrated in this study, isolated systolic hypertension is the most common form of hypertension found in elderly 24 . The predominance of systolic hypertension may reflect the consequences of a tendency for clinicians to treat it less aggressively than diastolic hypertension.

Only 19% of the respondents were newly diagnosed with hypertension. Although these proportions are quite similar to those reported from other developing countries 25 . Population studies have shown different rate of hypertension awareness by hypertensive individuals. In European countries, these levels range from 52.7% in Germany to 70% in Sweden. In Canada and USA the awareness rate is 83% and 87.3% respectively 26 27 . In these studies, the degree of awareness of hypertension was high (81%) than those reported by the India study 28 . Wyatt et al. reported that awareness significantly increased with age, female sex, presence of major co-morbidity and receiving preventive care 29 , a combination of these factors contributes to the high level of awareness in our study.

Among treated hypertensive individuals in this study, about 30% had blood pressure controlled. These data are also similar to those reported by Liau et al. (34.3%) 23 , but lower than those reported by Cipullo et al. (52.4%). In Europe the blood pressure controls rate range from 22.9% in Spain to 37.7% in England 26 . The favourable results in our study may be due to adequate public information about hypertension and availability of free government-supplied anti hypertensive drugs with relatively high access to health care and insurance coverage.

Regarding the educational level, there are an inverse but not significantly association between schooling and AH, the small number of educated population may influence this result. Although, this finding was consistent with previous reports 30 31 . Cipullo et al. reported that the risk of hypertension was 2.8-times higher for those with lower schooling 26 . There is an intricate relationship between physical health and psychological status. Depression is known to have diverse effects on body functions. A French study of the elderly in the community revealed that hypertension was associated with anxiety but not depression 32 . Our study showed the lack of association between depression symptoms and incident hypertension.

The benefits of physical activity in the prevention and treatment of high blood pressure have been very well described 33 . In the present study, physical activity was not at a reduced risk for hypertension, but self-reported bias and interviewer bias can result in no differential misclassification.

The higher, but not significant, prevalence of hypertension in non smokers should be interpreted with caution since it is possible that part of the existing difference may be due to the high percentage of current smokers for male compared to female. No association between AH and marital status, these results are in accordance to what has been reported in other study 28 . In Martinique and China populations, the urban population exhibit a risk profile of hypertension 34 35 , but not urban/rural differences were observed in our study, it is important to note that in our study in Monastir City, only 14% are from rural sites, the small number may influence statistics. The higher prevalence rates in hypertension among the illiterate and in individuals living in urban area are often being attributed to differences in life style, modernization, a shift from an agricultural to non agricultural economy, physical activity and occupation. Socioeconomic difference play, also, an important role in health conditions influencing different factors such as access to the health system, degree of information and understanding of medical conditions 36 37 .

Similar to several population-based studies 26 38 39 , it was observed that there is an important association between BMI and AH, as well as waist circumference. Overweight and obesity are actual risk factors for hypertension. The positive correlation between waist circumference and AH identifies a simple, low cost and easy-to-apply measure as an important marker for AH.

The strong association between diabetes mellitus and hypertension has been reported by numerous studies 28 40 . It is widely appreciated that hypertension increase the risk of activities of daily living and instrumental activities of daily living decline and have an important impact on the dependency categories 41 , our data support this observation.

Some limitations of this study should be mentioned: first the use of a single visit to ascertain hypertension status can result in an over-estimation of its prevalence. Other potential sources of bias include the self-reported hypertensive treatment by the participants. Treatment status was assessed by the question of whether or not the person was currently using antihypertensive drugs (within the last two weeks). No information on type and dosage of antihypertensive medication being used or medication compliance was collected from participants, nor were they asked about AH control measures other than pharmacotherapy. Therefore, "treatment" was restricted to the use of medication, without consideration of other, non pharmaceutical strategies such as dietary sodium restriction.

In this study, we did not assess the lipid profile and blood glucose measurement, a fact that did not allow us to evaluate the association between the metabolic changes and high BP.

We used BMI as a simple method to categorize people as "healthy weight," "overweight," or "obese". BMI formula may overestimates fatness in elderly because of loss of height resulting from vertebral compression (a problem commonly associated with aging). Knee-heel length and weight/knee height as alternative measurements in the elderly were not measured.

On the other hand, we described the prevalence of hypertension among the elderly in Monastir city, in order to provide estimates of the problem, this adding information about a group that is under researched. The use of a door-to door survey, the large sample size, adequate representation of women and the very old and a health professional for blood pressure measurement are strengths of our study.