Analysis on associated factors of uncontrolled hypertension among elderly hypertensive patients in Southern China: a community-based, cross-sectional survey

This study conducted in Zhejiang province was a cross-sectional survey administrated in 6 community health centers across 12 counties. A multistage stratified random cluster sampling was used to include the subjects. In the first stage, 12 administrative districts were divided into Type 1 urban districts, Type 2 suburban districts and Type 3 rural districts based on economic levels. From each of these three groups, a district was systematically selected. The first stage included 3 districts. Two communities were randomly chosen from each district for the second stage. Subjects aged 60 years or more with hypertension living in the selected communities were invited to participate for the third stage. The cluster intra-group correlation intraclass correlation coefficient (ICC) is 0.02. Ethics approval was obtained from the Ethics Committee of Zhejiang Hospital, and all participants signed an informed consent.

Data collection of this field survey included a questionnaire interview, physical examinations and biochemical examinations. A standardized questionnaire was administered by trained general practitioners in community public health service centers during a face-to-face individual interview. The questionnaire included demographic information such as age, areas of residence, education and health behaviors such as history of smoking, alcohol consumption, diet and physical activity. Physical examinations performed by trained nurses included height, weight, waist circumference, blood pressure (BP) and biochemical examinations including fasting plasma glucose (FPG) and lipid profile.

Weight and height were measured with subjects standing without shoes and wearing light clothing. Participants stood upright with the head in Frankfort plane for height measurement. Height was recorded to the nearest 0.5 cm, and weight was recorded to the nearest 100 g. Body-mass-index (BMI) was calculated as weight in kilograms over height in meters squared [weight(kg)/(height(m))²] [7]. Waist circumference was measured at the level of the iliac crest using a non-elastic tape measure.

Blood specimens were collected in vacuum tubes after at least 10 hours of overnight fasting, for the measurement of plasma glucose concentration and serum concentrations of total cholesterol and triglycerides. Plasma glucose was measured with the glucose oxidase method. Lipids were measured using Beckman auto analyzer at the biochemistry laboratory of Zhejiang Hospital [7].

BP measurements were performed using standardized mercury sphygmomanometers in this study. Subjects were advised to avoid consuming alcohol or tobacco, ingesting tea or coffee, or engaging in exercise for at least 30 minutes before BP measurement. In addition, subjects were asked to rest for at least 15 minutes before BP testing in a quiet room. The average of two measurements taken 5 minutes apart will be used, but if systolic BP (SBP) measurements differed by 10 mmHg or greater, or diastolic BP (DBP) differed by 5 mmHg or greater, a third measurement will be taken. In this instance the average of the two closest SBP and two closest DBP measurements will be used to define the BP [8].

Knowledge about hypertension was tested by questionnaire including the normal blood pressure values, and some healthy lifestyles related to hypertension. Awareness of hypertension was defined as subjects knowing their own blood pressure values. Treatment of hypertension was defined as taking a prescribed medicine for management of hypertension.

According to the Chinese Guidelines on Prevention and Control of Hypertension [9], hypertension is defined as SBP ≥ 140 mmHg and/or DBP ≥ 90 mmHg, or self-reported treatment of hypertension with antihypertensive medication. Pre-hypertension is defined as 120 mmHg ≤ SBP ≤ 139 mmHg and/or 80 mmHg ≤ DBP ≤ 89 mmHg, Stage II hypertension is defined as 160 mmHg ≤ SBP ≤ 179 and/or 100 mmHg ≤ DBP ≤ 109 mmHg, and stage III hypertension is defined as SBP ≥ 180 mmHg and/or DBP ≥ 110 mmHg. According to the recent JNC 8 Hypertension Guidelines, a BP goal of less than 150/90 mmHg for hypertensive persons aged 60 years or older is defined as control of hypertension [10].

According to the Chinese criteria, overweight is defined as 24 kg/m² ≤ body mass index (BMI) < 28 kg/m²; obesity was defined as BMI ≥ 28 kg/m²[11]. Fasting plasma glucose (FPG) is classified as impaired fasting glucose (IFG) with 6.1 mmol/L ≤ FBG ≤ 6.99 mmol/L, and diabetes mellitus (DM) as FBG ≥ 7.0 mmol/L [11]. Serum triglycerides (TG) are classified as moderately high TG (1.7 mmol/L ≤ TG ≤ 2.25 mmol/L); high TG (TG ≥ 2.26 mmol/L). Serum total cholesterol (TC) is categorized as moderately high TC (5.18 mmol/L ≤ TC ≤ 6.19 mmol/L); high TC (TC ≥ 6.22 mmol/L). Low serum high-density lipoprotein cholesterol (HDL-C) are defined as HDL-C < 1.04 mmol/L [11].

Subjects who smoked one cigarette or more per day for over 6 months were defined as smokers, and alcohol drinkers were assessed by asking subjects whether they had consumed more than once every week in the last 12 months. Subjects who consumed more than 10 grams salt per day for over 6 months were defined as excessive salt users.

The sampling design including stratification, clustering and sampling weights (accounting for differential probabilities of selection and also a post-stratification on gender, age) was taken into account in all estimates and analyses using the specific SAS commands.

Epidata 3.0 was used for data entry and validation and SAS 9.2 for data management and analysis. Socio-demographic characteristics, physical measurements and hypertension status of participants were summarized using frequencies (percentages) or means and standard deviations. Mean values and proportions were compared by the Student’s t-test and chi-square test, respectively. The trends in prevalence of hypertension associated factors across categories were analyzed using chi-square test. And the strength of associations of socio-demographic associated factors of hypertension was assessed for binary response variable (uncontrolled hypertension yes/no) by Odds-Ratios (OR) estimated in logistic regression models. Crude associations were first assessed using univariate models, then associations were assessed using multivariate models, and 95% confidence intervals were calculated. The final model was evaluated by goodness of fit. Significance level was set at p < 0.05 for all hypothesis tests.

The 10644 participants (mean age of 70.33 years) included 5527 women (51.93%), 5117 men (48.1%), 3148 overweight subjects (29.57%), 846 (7.95%) obese subjects, 1654 smokers (15.54%) and 1750 drinkers (16.44%). Men, compared with women, had a lower BMI, FBG, TG, TC and higher level of education; reported significantly higher proportions of current smoking, alcohol use, and had a lower prevalence of overweight and obesity (Table 1). The demographic characteristics of participants in different clusters such as gender, age and education level were showed in the Additional file 1.

Control of hypertension for participants of this survey is presented in Table 2. Of the 10644 hypertensive patients aged 60 years or older, 4747 achieved a BP under 150/90 mmHg. The control rate of hypertension was 44.6% according to the JNC 8 Guidelines, the control rate was 34.5% if the BP goal of less than 140/90 mmHg was defined as control of hypertension. And 27.3% participants had BP of more than 160/100 mmHg (stage II and stage III hypertension).

Compared with men, women had slightly higher control rate of hypertension (46.8% vs. 42.6%, P = 0.003), and the control rate decreased with older age (P < 0.001). Participants living in suburban areas had lower control rates of hypertension compared with living in rural and urban areas (31.4% vs. 34.2% vs. 51.1%, P < 0.0001). The control rate was higher among those with lower education levels and those with higher BMI’s (all P < 0.001). The rate of uncontrolled hypertension was increased with age and BMI, and decreased with the level of education. Smokers, drinkers, and the groups with family history of hypertension, lower HDL-C, high FBG or TC had lower control rate of hypertension (Table 2).

Awareness of hypertension was defined as subjects knowing their own blood pressure values. In the 10644 hypertensive subjects, 4924 (46.27%) were aware of their normal blood pressure values, 5357 (50.33%) were taking antihypertensive drugs. Men, compared with women, had slightly lower rates of awareness (42.05% vs. 48.69%, P = 0.003) and treatment (47.51% vs. 52.06%, P < 0.001). In men as well as women, the awareness and treatment rates of hypertension were higher with older age (all P < 0.001, Figures 1 and 2).

Table 3 showed the use of antihypertensive medications according to areas among the 5357 treated hypertensive patients. The results showed that 68.57% of treated hypertensives used combination therapy, and 32.36% chose monotherapy. Urban participants, compared with suburban and rural participants, more frequently used a combination of antihypertensive drugs including single pill combination treatment (P < 0.001, Table 3). The control rate was higher among subjects using combination of antihypertensive medications compared with that of using monotherapy (62.2% vs. 52.4%, P < 0.001).

Logistic regression models including crude model and adjusted model were used to assess factors associated with uncontrolled hypertension, respectively. The model adjusted for demographic factors such as gender, age, areas, showed that elderly, living in suburban and rural areas compared with living in urban areas, low level of education, family history of hypertension, smoking, alcohol use, excessive salt consumption, lack of physical activity, overweight and obesity and high FPG were associated with uncontrolled hypertension (Table 4).

In the adjusted model, the associations between uncontrolled hypertension and smoking, alcohol use, lack of physical activity and higher FPG were more pronounced in men than women. The associations between uncontrolled hypertension and TG and TC were less pronounced for men (Table 4).