PDF version

Tahseen Kazmi,¹ MLF Nagi,² Shama Razzaq,³ Shamaila Hussnain,¹ Noor Shahid¹ and Unsa Athar²

¹Community Medicine, Central Park Medical College, Lahore, Pakistan. ²Community Medicine, Shalamar Medical and Dental College, Lahore, Pakistan. ³Health Services Academy, Islamabad, Pakistan. (Correspondence to: Tahseen Kazmi: This e-mail address is being protected from spambots. You need JavaScript enabled to view it ).

Abstract

Background: Cases of noncommunicable (NCD) diseases are increasing in Pakistan. They are responsible for 58% of all deaths and information on the risk factors is lacking.

Aims: To determine the prevalence of diabetes, hypertension and abnormal heart rate and their risk factors in a multiethnic, low-income area of Lahore.

Methods: This was a cross-sectional study conducted in an urban settlement of Lahore in 2018–2019. Eligible participants were aged ≥ 30 years and resident in selected settlements. We used a modified World Health Organization stepwise approach to NCD risk factor surveillance (STEPS) questionnaire to collect data from a random sample of 906 residents. The modified questionnaire had a Cronbach alpha of 80.0%. Participants were interviewed in their homes and their height, weight, heart rate, and hip circumference were measured.

Results: Most of the respondents were women (64.5%). The greatest proportion of the participants were aged 30–39 years (42.4%). Of all the respondents, 40.1% had hypertension, 15.8% had diabetes and 17.0% had ischaemic heart disease. Of the risk factors examined, 68.8% of the respondents were overweight or obese, 37.0% had pre-hypertension, 13.6% used tobacco, and 1.8% used alcohol. Age was the most significant risk factor for noncommunicable diseases. Body mass index was significantly associated with diabetes in women, while family history of diabetes was significantly associated with diabetes in men.

Conclusion: The prevalence of hypertension and diabetes was higher among our study participants in Lahore than reported in the national statistics. Significant proportions of the respondents had pre-hypertension and pre-diabetes, making them vulnerable to serious complications of hypertension and diabetes.

Keywords: noncommunicable diseases, hypertension, diabetes, prevalence, risk factors, Pakistan.

Citation: Kazmi T; Nagi MLF; Razzaq S; Hussnain S; Shahid N; Athar U. Burden of noncommunicable diseases in Pakistan. East Mediterr Health J. 2022;28(11):798–804. https://doi.org/10.26719/emhj.22.083
Received: 29/03/22; accepted: 11/08/22

Copyright © Authors 2022; Licensee: World Health Organization. EMHJ is an open access journal. This paper is available under the Creative Commons Attribution Non-Commercial ShareAlike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO; https://creativecommons.org/licenses/by-nc-sa/3.0/igo).

---

Introduction

The burden of noncommunicable diseases (NCDs) continues to rise and they have become the leading cause of morbidity and mortality globally, accounting for almost 70.0% of all the deaths worldwide (1). The leading NCDs such as diabetes, chronic respiratory disease, cardiovascular disease and cancer jointly accounted for about 79.0% of all NCD deaths all over the world in 2018 (2). The estimated number of deaths globally caused by NCDs was 26.6 million in 1990 (3), which increased to 34.5 million in 2010 and to about 40.5 million in 2016.

Because of the rapid change in lifestyle and advanced urbanization, low- and middle-income countries suffer from a great burden of NCDs (4–6). Deaths caused by NCDs in low- and middle-income countries was 78.0% in 2016. This high burden of NCDs is due to the limited number of facilities and limited resources available for this kind of diseases (3).

Pakistan is the fifth most populated country worldwide with a population of 207.7 million (7). The country is facing a double burden of communicable diseases and NCDs (8). Adoption of a sedentary lifestyle and changes in diet have resulted in a surge in NCDs in Pakistan (9). Hypertension is the most common NCD with a prevalence of 38.7%, followed by diabetes with a prevalence of 14.6% (10). The National Health Survey of Pakistan reported a diabetes prevalence of 11.0% in the general population (11). In 2013–2014, a large-scale survey was conducted in Sindh and Punjab provinces, to determine the prevalence of NCDs and risk factors using the WHO STEPS instrument (12). Hypertension was seen in 53% of the sample which is higher than other South Asian countries (12). The main factors affecting the level of hypertension in Pakistan were: lack of proper healthcare guidelines; absence of patient safety measures; drug misuse; limited public health research; lack of enforcement of policies; and limited resources (13).

Several other risk factors for common NCDs exist, the most common being obesity, pre-hypertension, use of tobacco products and poor diet. Obesity was a risk factor for hypertension and diabetes in a study conducted using WHO STEPS instrument (14), with the prevalence of obesity reported as 31.2% (14). Another study in Kathmandu reported the obesity or over-weight prevalence as 52.2% (15).

The main objective of our study was to determine the overall prevalence of common NCDs in the general population in Lahore, Pakistan. For data collection, we modified the WHO STEPS questionnaire and determined its internal validity as the secondary objective.

Methods

Study design and setting

This was a cross-sectional study carried out in urban settlements of Union Councils 120 and 122 of Lahore, Punjab from September 2018 to September 2019. Union councils are the fifth tier of government in Pakistan; each union council has 13 elected members or councilors. The two union councils were selected from Shalamar and Gulberg zones.

Study sample

The minimum sample size calculated was 882 using the WHO sample size calculator with 95% confidence interval (CI) and 49.0% as the prevalence of NCDs in Pakistan (10).

We used a two-stage cluster sampling method to collect data. The union councils were the primary unit. Blocks were the secondary units and were made up of about 2000 houses in the two union councils. One block was randomly selected from each union council. We started with a randomly selected household within the block and then selected every fifth house in that block. Eligibility criteria were: age ≥ 30 years and resident in these union councils. Exclusion criteria were: refusal to participate; self-reported severe comorbid conditions with life expectancy of less than 1 year; or other self-reported serious conditions likely to interfere with study participation or with the ability to complete the study (e.g. terminal cancer, HIV or tuberculosis). All eligible individuals present in each household were interviewed. If any individual declined to participate, the whole household was excluded and the next household was selected.

Data collection

Data were collected through face-to-face, door-to-door interviews with all the eligible participants. The WHO STEPS questionnaire, modified to suit the setting with more detailed items excluded, was used to collect the data. The modified version was developed in English. Data were collected by nurses and doctors trained to conduct the interviews and complete the questionnaires. If any data were later found missing, the household was revisited to complete the information.

The demographic information collected was age, sex, marital status, educational level and occupation. A few response categories were changed for some variables. The other sections of the WHO STEPS questionnaire were limited to basic information such as consumption of tobacco, alcohol, fruits and vegetables, and intensity of physical activity (mild, moderate or intense). Physical characteristics (height, weight, heart rate and hip circumference) were measured by trained nurses (male and female). Information on history of hypertension, diabetes, stroke and ischaemic heart disease (chest pain, angina and shortness of breath) was also collected. A few items on history of raised blood pressure, diabetes, raised cholesterol and cardiovascular disease were included in the modified version.

Statistical analysis

Data analysis was done using SPSS, version 26. Socioeconomic and demographic factors were cross-tabulated with age group and sex. The chi-square test or Fisher’s exact test was used to evaluate significant differences.

The reliability of the questionnaire was assessed and the Cronbach alpha determined was 80%. We used exploratory factor analysis to assess the internal structure of the questionnaire. Principal component analysis was used with varimax rotation. Factors with Eigen values > 1.0 were extracted.

The prevalence of the main NCDs was determined and the association between various risk factors and common NCDs was examined.

Ethical considerations

The study was approved by the institutional review board of Shalamar Institute of Health Sciences (SMDC/IRB/11-12/140). Verbal consent was obtained before entering any household and written consent was obtained before starting the interview.

Results

Data were collected from 906 participants who lived in Union Councils 120 and 121. The internal reliability of 34 items was 80% using Cronbach alpha. These items were related to tobacco use, dietary pattern, physical activity, raised blood pressure and physical measurements. With regard to history of NCDs, 38.2% (346/906) of the participants had a family history of high blood pressure, 41.9% (380/906) had a family history of diabetes and 7.9% (72/906) had a history of stroke. With regard to the risk factors, 68.8% (623/906) of the sample was overweight or obese, 37.0% (335/906) had pre-hypertension, 13.6% (123/906) were tobacco users, 5.6% (51/906) had pre-diabetes and 1.8% drank alcohol regularly (16/906). The prevalence of hypertension, diabetes and ischaemic heart disease was 40.1% (363/906), 15.8% (143/906) and 17.0% (154/906), respectively. The prevalence of ischaemic heart disease was the same in male and female respondents.

The greatest proportion of the participants were in the age group 30–39 years (42.5% [385/906]), and 64.5% (584/906) were female (Table 1). Age was significantly associated with sex (P < 0.001). Marital status, educational level and occupation were all significantly associated with age and sex (P < 0.001; Table 1).

With regard to the internal structure of the questionnaire, the Kaiser–Meyer–Olkin measure of sampling adequacy and the Bartlett test of sphericity had values of 0.49 and 1248.07, respectively, (P < 0.001), allowing for exploratory factor analysis. The threshold for factor loadings was 0.3. Maximum factor loading for each item in the principal component analysis was used to assign that item in a particular factor. Twelve factors were extracted with about 76.1% of the variance explained (Table 2).

Age was the most significant risk factor for NCDs. A statistically significant association was observed for body mass index and diabetes in females (P = 0.02). Family history of diabetes was significantly associated with diabetes in males (P < 0.001). In male participants, occupation was significantly associated with abnormal heart rate (less than 60 beats/minute (bradycardia) or more than 100 beats/minute (tachycardia; P = 0.05). Income, family history of ischaemic heart disease, family history of hypertension and regular walking were not significantly associated with common NCDs (Table 3). The prevalence of diabetes was same for male and female participants but varied with age: 4.5% in participants aged 30–39 years, 16.0% in participants aged 40–49 years, 19.6% in participants aged 50–59 years and 23.8% in participants aged 60 years and older.

Discussion

We used exploratory factor analysis of the modified WHO STEPS questionnaire to assess the validity of the modified version. We could not use the original version of the STEPS questionnaire because it included questions on behavioural measurements, medical check-ups, tradition and background. The WHO STEPS questionnaire was designed to be applied to any community, we modified it according to our socioeconomic and demographic conditions. Another study on the burden of NCDs developed and validated the tool extracted from a past research questionnaire (16). In this study, the questionnaire was based on an 85-item questionnaire designed to assess cardiovascular disease risk (17). Exploratory factor analysis was used to determine the validity of the newly designed 21-item questionnaire together with confirmatory factor analysis (16).

Our study also aimed to assess the prevalence of NCDs by ethnicity and other demographic characteristics. Of all the NCDs examined, hypertension had the highest prevalence, followed by diabetes. This finding concurs with other studies in Pakistan on NCDs where hypertension or raised blood pressure was the most common NCD (10).

The prevalence of hypertension in our study was 40.1%, which is similar to a 2016 study in Pakistan (10). A study in Myanmar found an inverse association between years of schooling and hypertension (16), which is similar to our finding that the prevalence of hypertension differed by educational level. A study based on samples from 29 countries concluded that an additional year of education was associated with a 0.13 mmHg reduction in systolic blood pressure for men (18). A cross-sectional study conducted in Swaziland found an overall hypertension prevalence of 48.3% (19).

In our study the prevalence of diabetes was same for male and female participants. In other studies, the prevalence of diabetes ranged from 4.6% to 15% (10,20). In general, the prevalence of diabetes was higher in older respondents. However, diabetes was most commonly seen among women (14). The prevalence of diabetes varied by ethnicity, and this warrants further investigation; and a qualitative study to investigate the differences and interventions to reduce the prevalence of NCDs. We found no difference in the prevalence of diabetes according to occupation, which differs from a study in Sweden, which found that diabetes prevalence was higher among labourers, clerks, heavy machinery operators and drivers (21).

The most common risk factor was pre-hypertension, with a prevalence of 37.0%. The prevalence of tobacco use was low in our study at 13.6%. Tobacco use is a leading risk factor for morbidity and mortality globally (22). A study in different districts of Pakistan reported that 37.7% of participants used tobacco (23), a higher figure than we found. About 1.8% of our sample drank alcohol regularly, which was considerably lower than the other risk factors. The risk of hypertension has been shown to increase with excessive use of alcohol (24). Overweight and obesity is another risk factor for many NCDs (14) and it was high in our study (68.8%).

The prevalence of ischaemic heart disease was 17.0% in our study. A large-scale study in 53 cities of Punjab, Pakistan showed that 17.5% of the population had cardiovascular disease (25), which concurs with our finding. However, another study reported a lower prevalence of 6.2% for ischaemic heart disease and among those most were male participants (26). The prevalence of ischaemic heart disease was same in male and female respondents.

A strength of our study is that it was conducted in a multi-ethnic setting and covered almost all the economic classes. Two union councils were selected to remove the confounding of socioeconomic class. While the study was limited to one city, the sample size was large and representative. The important risk factors found in the study can be used to guide the development of policies to control NCDs.

Funding: None.

Competing interests: None declared.

Charge des maladies non transmissibles au Pakistan

Résumé

Contexte : Les cas de maladies non transmissibles (MNT) sont en augmentation au Pakistan. Ces dernières sont responsables de  58 % de tous les décès ; par ailleurs, les informations sur les facteurs de risque font défaut.

Objectifs : Déterminer la prévalence du diabète, de l'hypertension et de l'arythmie cardiaque ainsi que leurs facteurs de risque dans une zone multiethnique à faible revenu de Lahore.

Méthodes : Il s'agissait d'une étude transversale menée dans un établissement urbain de Lahore en 2018-2019. Les participants admissibles étaient âgés de 30 ans ou plus et résidaient dans les établissements sélectionnés. Nous avons utilisé une approche par étapes modifiée de l'Organisation mondiale de la Santé pour le questionnaire de surveillance des facteurs de risque des maladies non transmissibles (STEPS) afin de recueillir des données auprès d'un échantillon aléatoire de 906 résidents. Le questionnaire modifié présentait un alpha de Cronbach de 80,0 %. Les participants ont été interrogés à leur domicile, leur taille et leur poids ont été consignés et leur fréquence cardiaque et la circonférence de leurs hanches ont été mesurées.

Résultats : La plupart des répondants étaient des femmes (64,5 %). La plus grande proportion des participants étaient âgés de 30 à 39 ans (42,4 %). Sur l'ensemble des personnes interrogées, 40,1 % présentaient une hypertension, 15,8 % un diabète et 17,0 % des cardiopathies ischémiques. Parmi les facteurs de risque examinés, 68,8 % des répondants étaient en surpoids ou obèses, 37,0 % avaient une préhypertension, 13,6 % consommaient du tabac et 1,8 % de l'alcool. L'âge était le facteur de risque le plus important pour les maladies non transmissibles. L'indice de masse corporelle était significativement associé au diabète chez les femmes, tandis que les antécédents familiaux de diabète étaient significativement associés au diabète chez les hommes.

Conclusion : La prévalence de l'hypertension et du diabète était plus élevée parmi les participants de notre étude à Lahore que celle rapportée dans les statistiques nationales. Des proportions significatives des personnes interrogées présentaient une préhypertension et un prédiabète, ce qui les rendait plus à risque de développer de graves complications liées à l'hypertension et au diabète.

عبء الأمراض غير السارية في باكستان

تحسين كاظمي، ملف ناجي، شاما رزّاق، شمايلا حسانين، نور شاهد، آنسة آثار

الخلاصة

الخلفية: تتزايد الأمراض غير السارية في باكستان، وهي السبب في 58% من جميع الوفيات، ولكن لا تتوفَّر معلومات عن عوامل الخطر للإصابة بها.

الأهداف: هدفت هذه الدراسة الى تحديد معدل انتشار أمراض السكري وارتفاع ضغط الدم واضطراب سرعة القلب وعوامل الخطر المرتبطة بها، في منطقة لاهور التي تتسم بتعدُّد الأعراق وانخفاض مستوى الدخل.

طرق البحث: أُجريت هذه الدراسة المقطعية في الحضر في لاهور في عامي 2018-2019. ولقد اختير المشاركون المؤهلون ممن لا تقل أعمارهم عن 30 عامًا من سكان المدن المختارة. وقد استخدمنا استبيان نهج منظمة الصحة العالمية التدريجي لترصُّد عوامل الخطر للأمراض غير السارية، بعد تعديله لجمع البيانات من عينة عشوائية ضمت 906 من السكان. وقد بلغ معامل ألفا كرونباخ للاستبيان المعدل 80.0%. وقد أُجريت مقابلات مع المشاركين في منازلهم، مع قياس الطول والوزن وسرعة القلب ومحيط الجسم عند الورك.

النتائج: كان معظم من أجاب عن الاستبيان من النساء (64.5%)، وتراوحت أعمار النسبة الأكبر من المشاركين (42.4%) بين 30 و 39 عامًا. ومن بين من أجابوا عن الاستبيان، كان 40.1% مصابين بارتفاع ضغط الدم، و 15.8% مصابين بالسكري، و 17% مصابين بمرض القلب الإقفاري. ومن بين عوامل الخطر التي بحثتها الدراسة، كان 68.8% ممن أجابوا عن الاستبيان يعانون من زيادة الوزن أو السمنة، و 37.0% منهم يعانون من مقدمات ارتفاع ضغط الدم، و 13.6% يتعاطون التبغ، و 1.8% يتعاطون المشروبات الكحولية. وكان العمر هو أهم عامل خطر للإصابة بالأمراض غير السارية. أما السكري تحديدًا، فقد كان هناك ارتباط كبير بين منسب كتلة الجسم والإصابة به لدى النساء، وارتباط كبير بين وجود إصابات بالسكري في العائلة والإصابة به لدى الرجال.

الاستنتاجات: كان معدل انتشار ارتفاع ضغط الدم والسكري بين المشاركين في الدراسة في لاهور أعلى من معدلاته الواردة في الإحصاءات الوطنية. وكانت نسبة كبيرة ممن أجابوا على الاستبيان مصابين بمقدمات ارتفاع ضغط الدم ومقدمات السكري، مما يجعلهم عرضةً للمضاعفات الخطيرة لارتفاع ضغط الدم والسكري.

References

1. Noncommunicable diseases [Internet]. Geneva: World Health Organization (https://www.who.int/health-topics/noncommunicable-diseases#tab=tab_1, accessed 13 September 2021).
2. Noncommunicable diseases country profiles 2018. Geneva: World Health Organization; 2018 (https://apps.who.int/iris/handle/10665/274512, accessed 3 October 2022).
3. Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380(9859):2095–128. https://doi.org/10.1016/S0140-6736(12)61728-0
4. Khorrami Z, Etemad K, Yarahmadi S, Khodakarim S, Kameli M, Hezaveh AM, et al. Urbanization and noncommunicable disease (NCD) risk factors: WHO STEPwise Iranian NCD risk factors surveillance in 2011. East Mediterr Health J. 2017;23(7):469–79. https://doi.org/10.26719/2017.23.7.469
5. Bricas N. Urbanization issues affecting food system sustainability. In: Brand B, Bricas N, Conaré D, Daviron B, Debru J, Michel L, et al., editors. Designing urban food policies. Cram: Springer; 2019.
6. Panda R, Mahapatra S, Persai D. Health system preparedness in noncommunicable diseases: findings from two states Odisha and Kerala in India. J Family Med Prim Care. 2018;7(3):565–70. https://doi.org/10.4103%2Fjfmpc.jfmpc_111_17
7. Final results (Census-2017) [Internet]. Islamabad: Pakistan Bureau of Statistics; 2017 (https://www.pbs.gov.pk/content/final-results-census-2017, accessed 15 September 2021).
8. Naeem Z. Pakistan – double burden of diseases. J Shifa Tameer-e-Millat University. 2019;2(1):1–2.
9. Roglic G, Unwin N, Bennett PH, Mathers C, Tuomilehto J, Nag S, et al. The burden of mortality attributable to diabetes: realistic estimates for the year 2000. Diabetes Care. 2005;28(9):2130–5. https://doi.org/10.2337/diacare.28.9.2130
10. Naseem S, Khattak UK, Ghazanfar H, Irfan A. Prevalence of non-communicable diseases and their risk factors at a semi-urban community. Pan Afr Med J. 2016;23(1):151. https://doi.org/10.11604%2Fpamj.2016.23.151.8974
11. Shera AS, Rafique G, Khwaja IA, Ara J, Baqai S, King H. Pakistan national diabetes survey: prevalence of glucose intolerance and associated factors in Shikarpur, Sindh Province. Diabet Med. 1995;12(12):1116–21. https://doi.org/10.1111/j.1464-5491.1995.tb00430.x
12. Rafique I, Saqib MAN, Munir MA, Qureshi H, Rizwanullah, Khan SA, et al. Prevalence of risk factors for noncommunicable diseases in adults: key findings from the Pakistan STEPS survey. East Mediterr Health J. 2018;24(1):33–41. https://doi.org/10.26719/2018.24.1.
13. Riaz M, Shah G, Asif M, Shah A, Adhikari K, Abu-Shaheen A. Factors associated with hypertension in Pakistan: a systematic review and meta-analysis. PLoS One. 2021;16(1):e0246085. https://doi.org/10.1371/journal.pone.0246085
14. Saeed KM. Prevalence of risk factors for non-communicable diseases in the adult population of urban areas in Kabul City, Afghanistan. Cent Asian J Glob Health. 2014;2(2):69. https://doi.org/10.5195%2Fcajgh.2013.69
15. Dahal S, Sah RB, Niraula SR, Karkee R, Chakravartty A. Prevalence and determinants of non-communicable disease risk factors among adult population of Kathmandu. PLoS One. 2021;16(9):e0257037. https://doi.org/10.1371/journal.pone.0257037
16. Mya KS, Zaw KK, Mya KM. Developing and validating a questionnaire to assess an individual’s perceived risk of four major non-communicable diseases in Myanmar. PLoS One. 2021;16(4):e0234281. https://doi.org/10.1371/journal.pone.0234281
17. Woringer M, Nielsen JJ, Zibarras L, Evason J, Kassianos AP, Harris M, et al. Development of a questionnaire to evaluate patients’ awareness of cardiovascular disease risk in England’s National Health Service Health Check preventive cardiovascular programme. BMJ Open. 2017;7(9):e014413. https://doi.org/10.1136/bmjopen-2016-014413
18. Stamler R, Shipley M, Elliott P, Dyer A, Sans S, Stamler J. Higher blood pressure in adults with less education. Some explanations from INTERSALT. Hypertension. 1992;19(3):237–41. https://doi.org/10.1161/01.hyp.19.3.237
19. Gbadamosi MA, Tlou B. Modifiable risk factors associated with non-communicable diseases among adult outpatients in Manzini, Swaziland: a cross-sectional study. BMC Public Health. 2020;20(1):1–2. https://doi.org/10.1186/s12889-020-08816-0
20. Idris IO, Oguntade AS, Mensah EA, Kitamura N. Prevalence of non-communicable diseases and its risk factors among Ijegun-Isheri Osun residents in Lagos State, Nigeria: a community based cross-sectional study. BMC Public Health. 2020;20(1):1258. https://doi.org/10.1186/s12889-020-09349-2
21. Carlsson S, Andersson T, Talbäck M, Feychting M. Incidence and prevalence of type 2 diabetes by occupation: results from all Swedish employees. Diabetologia. 2020;63(1):95–103. https://doi.org/10.1007/s00125-019-04997-5
22. Global status report on noncommunicable diseases 2014. Geneva: World Health Organization; 2014 (https://apps.who.int/iris/handle/10665/148114, accessed 11 October 2021).
23. Alam AY, Iqbal A, Mohamud KB, Laporte RE, Ahmed A, Nishtar S. Investigating socio-economic-demographic determinants of tobacco use in Rawalpindi, Pakistan. BMC Public Health. 2008;8(1):1–9. https://doi.org/10.1186/1471-2458-8-50
24. Rao N, Eastwood SV, Jain A, Shah M, Leurent B, Harvey D, et al. Cardiovascular risk assessment of South Asians in a religious setting: a feasibility study. Int J Clin Pract. 2012;66(3):262–9. https://doi.org/10.1186%2F1471-2458-13-160
25. Zubair F, Nawaz SK, Nawaz A, Nangyal H, Amjad N, Khan MS. Prevalence of cardiovascular diseases in Punjab, Pakistan: a cross-sectional study. J Public Health. 2018;26(5):523–9. https://doi.org/10.1007/s10389-018-0898-4
26. Abbas S, Kitchlew AR, Abbas S. Disease burden of ischemic heart disease in Pakistan and its risk factors. Ann Pak Inst Med Sci. 2009;5(3):145–50.