|Year : 2012 | Volume
| Issue : 3 | Page : 320-329
Screening high-risk population for hypertension and type 2 diabetes among Thais
Kulaya Narksawat1, Natkamol Chansatitporn1, Panuwat Panket2, Jariya Hangsantea1
1 Faculty of Public Health, Mahidol University, Bangkok, Thailand
2 Bureau of Non Communicable Diseases, Department of Disease Control, Ministry of Public Health, Bangkok, Thailand
|Date of Web Publication||25-May-2017|
Faculty of Public Health, Mahidol University, Bangkok
Source of Support: None, Conflict of Interest: None
Background: waist circumference (WC) and body mass index (BMI) are simple screening tools for hypertension (HT) and type 2 diabetes (DM). Cutoffs of WC for BMI for Asians have been discussed. This study aimed to assess the accuracy of screening tools and associations of WC, BMI with HT and DM.
Methods: Data from the national screening programme for metabolic syndrome conducted in 2010 in 21 provinces in the central region of Thailand were analysed. A total of 10 748 participants aged >35 years were included in the analysis with cutoffs of WC set at 90 cm for men, 80 cm for women, and BMI at 23 kg/m2 for both sexes.
Results: WC produced low sensitivity and high specificity among male participants, and moderate sensitivity and specificity among female participants, while BMI produced moderate sensitivity and specificity in both sexes. Significant associations were found among those who had high WC only, high BMI only, and both high WC and BMI with HT and DM in both sexes. (males for HT, OR=1.63, 95%CI: 1.15–2.33, OR=1.22, 95%CI: 1.03–1.44 and OR=2.03, 95%CI: 1.07–2.42; males for DM, OR=1.39, 95%CI: 1.05–1.83), OR=1.77, 95%CI: 1.07–2.94 and OR= 2.05, 95%CI: 1.57–2.69, females for HT, OR=1.69: 95%CI 1.38–2.07, OR=1.32; 95%CI: 1.09–1.60 and OR=2.54, 95%CI: 2.11–2.91; females for DM, OR=1.45, 95%CI: 1.08–1.94, OR=1.45, 95%CI: 1.09–1.91 and OR=1.70, 95%CI: 1.39–2.09). When the cutoff WC was lowered among male participants to 85 cm, sensitivity increased, and significant strengths of associations with HT and DM were nearly the same.
Conclusion: For Thailand, WC and BMI with appropriate cutoffs can be effective screening tools to recruit high-risk populations into health promotion programmes. However, WC and BMI should be implemented with other screening tools for other risk factors because of their moderate accuracy.
Keywords: Waist circumference, body mass index, cutoffs, screening accuracy, hypertension, type 2 diabetes, Thailand.
|How to cite this article:|
Narksawat K, Chansatitporn N, Panket P, Hangsantea J. Screening high-risk population for hypertension and type 2 diabetes among Thais. WHO South-East Asia J Public Health 2012;1:320-9
|How to cite this URL:|
Narksawat K, Chansatitporn N, Panket P, Hangsantea J. Screening high-risk population for hypertension and type 2 diabetes among Thais. WHO South-East Asia J Public Health [serial online] 2012 [cited 2022 Oct 4];1:320-9. Available from: http://www.who-seajph.org/text.asp?2012/1/3/320/207028
| Introduction|| |
Screening for risk factors of hypertension (HT) and type 2 diabetes (DM) is very useful to recruit high-risk populations into health promotion programmes as primary prevention in developing countries like Thailand. Since 2009, Thailand has had an increasing trend of morbidity due to hypertension and type 2 diabetes during the last five years thus dramatically increasing the economic burden for treatment of these two diseases. HT and DM usually have the same risk factors pertaining to unhealthy lifestyle or behavioural risk factors such as lack of exercise, stress, smoking and poor eating habits which lead to overweight and obesity., Many anthropometric methods have been employed for measuring overweight and obesity. The simple methods which are easy to obtain for screening programmes and for self- monitoring are waist circumference (WC) and body mass index (BMI) which are calculated from weight and height., The World Health Organization (WHO) reported that the cutoff point of BMI for observed risk of HT and DM varied from 22 kg/m2 to 25 kg/m2 in different Asian populations, and high risk of HT and DM varied from 26 kg/m2 to 31 kg/m2. WHO also suggested that BMI for each country should be defined for its own population. The WHO Regional Office for the Western Pacific (WPRO) proposed that for Asian adults, overweight and obesity should have BMI>23 and BMI>25 kg/m2, while the Ministry of Public Health (MoPH), Thailand has used a cutoff of BMI at 23 kg/m2 for many national health survey programmes.,
Visceral obesity is known to be closely linked to HT and DM, but the appropriate cutoff of WC for being a risk factor of HT and DM is still controversial. Thailand, i.e., the Ministry of Public Health (MoPH), has specified cutoffs of WC for Thais at 90 cm for males and 80 cm for females. High WC or excess body fat around the waist is one among several clusters of symptoms of metabolic syndrome including dyslipiddemia as well as elevated levels of fasting blood glucose, increased blood pressure, low level of high density of lipoprotein (HDL), and high level of triglyceride. These are already known to be cardiovascular risk factors. The Ministry of Public Health (MoPH), Thailand, through the National Health Security Office (NHSO) has implemented screening programmes for metabolic syndrome and provided modified risk behaviour programmes for the Thai population all over Thailand since 2005. The screening tools for metabolic syndrome in these programmes include BMI, waist circumference and blood pressure, as well as fasting blood sugar.
Thus, the aims of this study were to assess the accuracy of WC and BMI when screened for HT and DM and the strength of associations between BMI, WC with HT and DM when using the cutoffs as recommended for Asian populations by WHO, and the MoPH, Thailand. In addition, the study aimed to identify new and appropriate cutoffs of WC and BMI to improve the sensitivity of screening programmes to recruit more higher-risk populations to participate in health promotion programmes to reduce the occurrence of HT and DM among Thais effectively.
| Methods|| |
The analysis was based partially on secondary data from the fifth national screening programme for metabolic syndrome conducted in 21 provinces of the central region of Thailand by the NHSO, MPH, from 1 October 2009 to 31 July 2010. Stratified random sampling technique was used to select participants for each province. Simple random sampling with proportion to size of population using a computer programme was used to select subjects for each province. The number of subjects (nh) taken from each province was calculated by nh = Nh×[n/N]; when Nh represents the total number of participants of each province, n is the estimated sample size and N is the total number of target population (N=673 952). The inclusion criteria of participants for analysis were aged 35 years or older with complete and plausible data for analysis. The total number of study population included in the analysis was 10 748.
Sites of study
Data collection was performed within one day at primary care units at health care centres or health promotion hospitals at village and district levels by trained health officers under the supervision of provincial health officers.
| Methods of data collection|| |
After 30 minutes of resting, blood pressure was measured by standard mercury sphygmomanometers and standardized digital equipment (Omron IA1) by trained health officers twice at a gap of 1–2 minutes. The average of the two readings was used to classify high blood pressure or hypertension.
Weight and height were measured by balanced beam scale. Body mass index (BMI) w as calculated by weight in kilograms divided by height in meters squared (kg/m2).
Fasting blood glucose
Fasting blood glucose level was detected from blood samples obtained from the veins of participants in the morning after at least eight hours of overnight fasting. Fasting blood glucose (FPG) was measured using the standard enzymatic method by health care officers.
Interviewing for behavioural risk factors
Interviews were conducted to identify covariate factors comprising general characteristics such as age, sex, living area, past history of other chronic diseases and having family members with hypertension and diabetes. Health risk behaviours including smoking, alcohol consumption, exercise habits and dietary habits during the last year were also discussed.
High blood pressure was defined as having a SBP ≥ 140 mmHg, and/or DBP ≥ 90 mmHg or currently taking antihypertensive drugs. Type 2 diabetes was defined when the level of FPG ≥ 126 mg/dl with confirmation and/ or taking antidiabetes drugs. BMI categories for Asian populations were defined as < 18.5 for underweight, 18.5–22.9 kg/m2 for normal, 23.0–24.9 kg/m2 for overweight, 25.0–29.9 kg/m2 for obese I, and ≥30 kg/ for obese II. Abdominal obesity was defined as WC ≥90 cm for men and ≥ 80cm for women.,
Pearson's correlation coefficient was used to demonstrate the relationships between WC, BMI, FBS, SBP and DBP. Receiver Operating Characteristic (ROC) curve was used to identify the areas under the curve (AUCroc) with 95% confidence interval. When AUCroc was greater than 0.5, it meant the screening tool was better than chance alone, and the bigger the AUCroc, the better the accuracy of the screening tool. Sensitivity, and specificity were calculated for accuracy of WC and BMI when screened for HT and DM. Multiple logistic regression analysis stratified by sex was used to determine the strength of associations between having high WC only, high BMI only, a combination of high WC and high BMI ( above cutoff points) with hypertension, type 2 diabetes by adjusted odds ratio (OR) and 95% confidence interval (95%CI). The confounding variables were age in years, living area (urban/rural), having family members with hypertension or type 2 diabetes (yes/ no), having been diagnosed for and using antihypertensive drugs or antidiabetes drugs (yes/no), smoking (current/ex/no), alcohol consumption (current/ex/no), exercise habits (never/irregular/regular), and dietary habits (healthy food/unhealthy food).
| Results|| |
The mean age of the study population was 53.55 (±12.56) years for male participants and 53.48 (±12.39) years for female participants. The average waist circumference was 82.07 ± 9.62 cm among female and 83.73 ± 9.15 cm among male participants. Moreover, the average BMI of female participants was higher than that of male participants which was 24.16 ± 3.98 and 23.08 ± 3.48 kg/m2, respectively [Table 1].
|Table 1: Means and standard division of biological characteristics of participants|
Click here to view
Results showed a significantly moderate positive correlation between WC and BMI, when positive correlation between WC and BMI with FBS, SBP, DPP among these group of participants was low (correlation coefficient for males; WC:BMI = 0.619, WC:FBS=1.46, WC:SBP = 0.167, WC:DBP= 0.142, BMI:FBS = 0.167, BMI:SBP = 0.155, BMI:0.137, correlation coefficient for females; WC:BMI = 0.657, WC:FBS=1.33, WC:SBP = 0.184, WC:DBP = 0.168, BMI:FBS = 0.119, BMI:SBP = 0.156, BMI:DBP=0.169) (not shown in the table).
Areas under the receiver operating characteristic curve (AUCroc)
Cutoffs were conducted for WC at 90 cm for male participants and 80 cm for female participants, and BMI of 23.00 kg/m2 for both sexes for screening HT and DM.
When screening for HT, AUCroc of WC produced significantly better accuracy than BMI by non overlapping of 95% confidence interval of AUCroc in both sexes (males: AUCroc of WC = 0.61 95%CI 0.58–0.62 and AUC of roc BMI = 0.57, 95% CI 0.55–0.58, females: AUC roc of WC =0.62, 95%CI 0.60–0.63 and AUCroc of BMI 0.59, 95%CI 0.57–0.60, respectively). However, when screening for diabetes, WC produced nonsignificant difference accuracy by AUC and 95% confidence interval with BMI roc by overlapping of 95% confidence interval of AUCroc in both sexes (males: AUCroc of WC = 0.62 95%CI 0.59–0.64 and AUC of BMI = roc 0.59, 95% CI 0.57–0.62, females: AUC of roc WC =0.59, 95%CI 0.57–0.62 and AUC of roc BMI 0.57, 95%CI 0.52–0.59, respectively) [Table 2].
|Table 2: Area under receiver operating characteristics curve, sensitivity, specificity of WC and BMI by sex and by different cutoffs when screened for hypertension and type 2 diabetes|
Click here to view
Sensitivity and specificity
When screening for HT and DM among male participants, WC demonstrated low sensitivity but high specificity (sensitivity: HT=30.12% and DM = 34.65%, specificity: HT=83.45% and DM=80.62%, respectively). In addition, among female participants, WC demonstrated low to moderate sensitivity and specificity (sensitivity: HT=66.23% and DM=66.85%, specificity: HT=52.54% and DM=48.23%, respectively).
When the cutoff of BMI was set at 23 kg/m2 screening for HT and DM in both sexes, results demonstrated low to moderate levels for both sensitivity and specificity (males: sensitivity for HT=56.91% and sensitivity for DM=62.73%, specificity for HT=52.92% and specificity for DM = 51.18%, females: sensitivity for HT=66.01% and sensitivity for DM=67.48%, specificity for HT=47.45% and specificity for DM=44.24%, respectively) [Table 2].
Suggested new cutoff for WC for screening hypertension and diabetes among male Thais
In order that more of the high-risk population can be recruited to participate in health promotion programmes, the sensitivity for screening of WC among male participants should be increased. When the cutoff of WC was lowered to 85 cm, it was found that sensitivity increased from 30.12% to 51.13% when screening for HT, and from 34.65% to 52.25% when screening for DM, and it made the specificity decrease from 83.45% to 63.31% and from 80.62% to 64.46% when screening for DM and HT, respectively [Table 2].
Strength of associations between WC, BMI and hypertension and type 2 diabetes
Results from logistic regression analysis when stratified by sex, demonstrated significant strength of associations. Male participants, who had WC >90 cm only, had BMI >23 kg/m2 only, and had combination of WC > 90 cm and BMI >23 kg/m2, were more likely to develop hypertension 1.63 times (OR=1.63; 95%CI = 1.15–2.33), 1.22 times (OR=1.22; 95%CI = 1.03–1.44) and 2.03 times (OR=2.03; 95%CI=1.70–2.42) times, respectively, when compared with those with WC≤90 cm, and BMI≤23 kg/m2. Moreover, they were more likely to develop type 2 diabetes 1.77 times (OR=1.77; 95%CI= 1.07–2.94), 1.39 times (OR=1.39; 95%CI = 1.05–1.83), and 2.05 times (OR=2.05; 95%CI=1.57–2.69), respectively, when compared with those with WC≤90 cm, and BMI≤23 kg/m2 [Table 3].
|Table 3: Adjusted odds ratio (OR) and 95% confidence interval (95%CI) for hypertension and diabetes by cutoffs of WC and BMI among male participants|
Click here to view
Female participants, who had WC >80 cm only, had BMI >23 kg/m2 only and had a combination of WC >80 cm and BMI >23 kg/m2, were more likely to develop hypertension 1.69 times (OR=1.69; 95%CI=1.38–2.07), 1.32 times (OR= 1.32; 95%CI = 1.09–1.60) and 2.54 times (OR=2.54; 95%CI = 2.21–2.91), respectively, when compared with those with WC≤80 cm, and BMI≤23 kg/m2 . Furthermore, they were more likely to develop type 2 diabetes 1.45 times (OR=1.45; 95%CI = 1.08–1.94), 1.45 times (OR=1.45; 95%CI = 1.09–1.91), and 1.70 times (OR=1.70; 95%CI=1.39–2.02), respectively, when compared with those with WC≤80 cm, and BMI≤23 kg/m2 [Table 4].
|Table 4: Adjusted odds ratio (OR) and 95% confidence interval ( 95%CI) for hypertension and diabetes by cutoffs of WC and BMI among female participants|
Click here to view
Among male participants, when lowered cutoff of WC to 85 cm, it was found that those who had WC >85 cm only, having both WC>85 cm and BMI>23 kg/m2, still had significant associations with HT and DM (for HT: OR=1.64:95%CI 1.15–2.32, and OR = 2.09: 95%CI 1.55–2.83, for DM: OR=1.64: 95%CI 1.15–2.32; and OR=2.09;95%CI 1.55–2.83, respectively). However, BMI held nonsignificant associations with HT and DM. [Table 3]
The results of this study also demonstrated higher effects of genetic determinants with HT and DM than WC and BMI. Male participants, who had family members with hypertension and diabetes, were more likely to develop hypertension and type 2 diabetes 2.66 times (OR=2.66; 95%CI=2.16–3.26) and 4.06 times (OR=4.06; 95%CI 3.06–5.38), respectively, compared with those who did not. Female participants, those who had family members with hypertension and type 2 diabetes were more likely to develop these two diseases 2.15 times (OR=1.87:95%CI=1.87–2.48) and 3.20 times (OR=3.70: 95%CI=2.68–3.93), respectively, compared with those who did not [Table 3],[Table 4].
Among this study population, independently high WC and high BMI demonstrated lower significant strength of associations with HT and DM than the combination of high WC and BMI , even though, moderate correlation between WC and BMI was found. These findings were consistent with the reported analysis among Chinese adults, Japanese men, and young adults in Jamaica,, and also in Thailand.
The high specificity (83.45%) and low sensitivity (30.12%) of WC among male participants implied that WC at 90 cm could screen those who did not have hypertension and type 2 diabetes better than screening for those who have hypertension and type 2 diabetes. In the meantime, high specificity can also produce high false negatives which lead to the exclusion of the high risk population to participate in health promotion programmes. In fact, it is desirable to choose a screening tool that has high values for both sensitivity and specificity but it is rather difficult to find such screening tools. If not, screening tools with high sensitivity are preferable for screening of these two chronic diseases when a false positive is better than a false negative. This is because including participants with false positives in a high risk population to attend health promotion programmes can do no harm and is better than to leave them out with other biological risk factors.
When this analysis lowered the cutoff of WC to 85 cm among male participants, the sensitivity increased when using screens for both HT and DM; this will also increase the number of participants in health promotion programmes, even though the strength of association (ORs) did not change much.
From this analysis, it was also found that WC had significantly better accuracy for screening for HT than BMI both in male and female participants, but both WC and BMI still had quite low screening accuracy (AUCroc= 0.57–0.62). This suggests that those with normal WC and normal BMI should be taken into consideration by other means to screen for other risk factors for HT and DM. However, both WC and BMI are quite good as monthly self monitoring tools for HT and DM. In some provinces in Thailand, they have used risk scores to assess the risk to develop hypertension and type 2 diabetes of which WC and BMI were the two components among others.
Hypertension and diabetes have genetic determinants,, and the results from this analysis also showed a higher strength of associations of genetic factors compared with environmental factors or lifestyle behavioural factors among this group of study population. It confirmed that having genetic factors with HT and DM should be one among other screening criteria.
Studies have reported that HT and DM had an ubiquitous association, and diabetes may be linked to increased peripheral vascular resistance of hypertension, or anti-hypertensive drugs can influence the probability of the occurrence of metabolic syndrome.,22 This analysis found 21.92% of the study population had both HT and DM. This association needs further follow-up studies to conclude that HT and DM might be risk factors for each other.
Cross-sectional associations usually have limitations to explain the causal relationship between risk factors and disease. This study only intended to suggest appropriate screening and self-monitoring tools, and to be concerned for those who had normal WC and BMI with high biological intermediate risk factors by leaving them out of the health promotion programmes.
However, in developing countries where the national budget is quite limited, health screening programmes with simple, inexpensive, easy to implement screening tools have been usually provided. Screening programmes for biological factors such as blood cholesterol level, triglyceride, fasting blood glucose, and good health education with health promotion programmes to modify health behaviours should also be continuously provided in order to reduce the occurrence and budget for treatment of these chronic diseases.
| Conclusion|| |
Accuracy of WC with cutoff 90 cm for males and 80 cm for females and BMI at 23 kg/m2 when screened for HT and DM is not high. This study suggests lowering the cutoff for WC from 90 cm to 85 cm for male participants in order to include more participants in health promotion programmes. Strength of association between undesirable WC and BMI with HT and DM was significantly high, but not as high as genetic predisposing. Annual health checkups to identify biological factors resulting from unhealthy lifestyle are highly recommended as not to solely depend on anthropometric indicators for developing countries.
| Acknowledgement|| |
The authors wish to thank the National Health Security Office, Ministry of Public Health, Thailand for allowing the use secondary data for analysis.
| References|| |
Duangtep Y, Narksawat K, Chongsuwat R, Rojanavipart P. Association between an unhealthy lifestyle and other factors with hypertension among hill tribe populations of the Mae Fah Luang District, Chiang Rai Province, Thailand. Southeast Asian J Trop Med Public Health
. 2010; 41(3): 726-734.
Pradhan A, Manson J, Rifai N, Buring J, Ridker J. C-reactive protein, interleukin 6, and risk developing type 2 diabetes mellitus. JAMA
. 2001; 286: 327-334.
World Health Organization. WHO expert consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. The Lancet
. 2004; 363: 157-163.
Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the national cholesterol education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA: the Journal of the American Medical Association
. 2001; 285(19): 2486–2497.
Romero-Corral, A, Somers V K, Sierra-Johnson J, Thomas RJ, Collazo-Clavell ML, Korinek J, et al. Accuracy of body mass index in diagnosing obesity in the adult general population. Int J Obes
. 2008; 32(6): 959–966.
World Health Organization, Regional Office for the Western Pacific. The Asia-Pacific perspective: redefining obesity and its treatment. Melbourne: Health Communications Australia, 2000.Ministry of Public Health. National Health Security Office. Manual for national screening program for metabolic syndrome among Thai people aged more than 15 years for the year 2010. Accessed 5/1/2012. http:// ucapps.nhso.go.th/allpp2010/FrmLoginMain- .jsp.
Ministry of Public Health. National health examination survey office. Report of 4th national health examination survey in Thailand 2008-2009. Nonthaburi, 2010.
Tuan N, Adair L, Suchindran C, He K, Popkin B. The association between body mass index and hypertension is different between East and Southeast Asians. Am J Cli Nutr
. 2009; 89: 1905-1912.
Okosun I, Osotimehin B, Cooper R, Forrester T, Rotimi C. Association of waist circumference with risk of hypertension and type 2 diabetes in Nigerians, Jamaicans, and African-Americans. Diabetes Care
. 1998; 21: 1836-1842.
Lwanga SK, Lemshow S. Sample size determination in health studies: a practical manual
. Geneva: World Health Organization, 1991.
Bewick V, Cheek L, Ball J. Statistics review 13 : receiver operating characteristic curves. Critical Care
. 2004; 8(6): 508-512.
Wildman RP, Gu D, Reynolds K, Duan X, He J. Appropriate body mass index and waist circumference cutoffs for categorization of overweight and central adiposity among Chinese adults. Am J Cli Nutr
. 2004; 80: 1129-1136.
Tabata S, Yoshimisu S, Hamachi T, Abe H, Ohnaka K, Kono S. Waist circumference and insulin resistance: a cross-sectional study of Japanese men. BMC Endocrine Disorder
. 2009; 9: 1.
Tulloch M, Ferguson T, Younger N, Broeck V, Boyne M, Knight-Madden J, et al. Appropriate waist circumference cut points for identifying insulin resistance in black youths: a cross sectional analysis of 1986 Jamaican birth cohort. Diabetology & Metabolic Syndrome
. 2010; 2: 68.
Thaikruea L, Seetamanotch W, Seetamanotch S. Appropriate cut-off level of bmi for screening in Thais adults. J Med Assoc Thai
. 2006; 89(12): 2123-2128.
Zhu S, Heshka S, Wang Z, Shen W, Allison D, Ross R, Heymsfield S. Combination of BMI and waist circumference for identifying cardiovascular risk factors in whites. Obesity research
. 2004; 12(4): 633-645.
Kotchen T, Kochen J, Grim C, George V, Kaldunski M, Cowley A, et al. Genetic determinants of hypertension: identification of candidate phenotypes. Hypertension
. 2000; 36: 7-13.
O’Rahilly S, Barroso I, Wareham N. Genetic factors in type 2 diabetes: the end of the beginning? Science
. 2005; 307: 370-373.
Modan M, Halkin H, Almog S, Lusky A, Eshkol A, Shefi M. Hyperinsulinemia: a link between hypertension obesity and glucose intolerance. J Clin Invest
. 1985; 75: 809-817.
Mancia G. The association of hypertension and diabetes: prevalence, cardiovascular risk and protection by blood pressure reduction. Acta Diabetol
. 2005; 42: S17-S25.
[Table 1], [Table 2], [Table 3], [Table 4]