Abstract
Background
Non communicable diseases (NCDs) are taking their toll in most low and middle income countries; incidentally, in the same populations that are struggling to deal with communicable diseases, hence presenting a picture of “double tragedy”. Most of the researches conducted on NCDs are facility-based; often in urban locations.
Objective
To determine the prevalence of high BP in a semi-urban community in Nigeria.
Methods
It was a cross-sectional study, where all the adult residents of the community were enrolled. Data was collected using a multi-section questionnaire, including anthropometric measurements. Data entry and analysis was done using SPSS 17.0 for windows and STATA 10.
Results
The results showed that 47.0% of the study population had a raised BP ≥140/90mmHg. At a univariate level, age, sex, higher income, more people in the household, daily cigarette smoking, daily alcohol intake and BMI, showed increased risk for high BP. However, sleeping for at least 8 hours a day showed protective influence against raised BP among the participants.
In the adjusted model, only age, sex and BMI remained statistically significant with Odds Ratio (OR) of 1.04 (95%CI; 1.00, 1.08), 0.45 (95%CI; 0.22, 0.90) and 1.08(95%CI; 1.03, 1.13) respectively. The model had an area under curve of 71.9%,
Conclusion
It was concluded that there is a high prevalence of raised BP in this semi-urban community, thus the need for intervention and preventive services to curb the looming epidemic of hypertension in this community in particular, and Nigeria as a whole, cannot be overemphasized.
Keywords: High BP, BMI, prevalence
Introduction
High blood pressure (BP) is a leading cause of global burden of disease, and most of it occurs in the developing world1. When selected major risk factors were assessed, the leading global risks for mortality in the world were found to be high BP (which accounted for 13% of global deaths), tobacco use (9%), high blood glucose and physical inactivity (6% each), overweight and obesity (5%). These were the risk factors for developing chronic diseases such as heart disease, diabetes mellitus and cancers2.
When the burden of high blood pressure disease was further assessed, it was found that globally it accounted for 7.6million premature deaths, 92 million DALYs (6% of global total), 54% stroke and 47% ischemic heart diseases. Furthermore, about 80% of disease burden caused by high BP was found to be borne by low and middle income countries, by people in middle ages and those with prehypertension3. Globally, from 1980 to 2008, the number of people with uncontrolled hypertension rose from 600million to nearly 1 billion, and overweight, obesity, high BP and high cholesterol had shifted their presence from western wealthy nations towards low and middle income countries. Regional differences showed highest systolic BP levels in Baltic, East and West African countries with higher figures for men in most regions of the world4.
In Africa, hypertension is thought to be the foundation for epidemic cardiovascular diseases5, and it had been identified as the most powerful, highly prevalent, independent, modifiable risk factor for death from heart disease and stroke at the population level6. In Soweto, approximately 46% of participants in a study had systolic/diastolic BP values ≥ 140/90 mmHg7. In the West African sub-region, hypertension prevalence rates between 1970 and 2009 in Ghana, revealed a range of 19% to 48%, and in 2007, hypertension was said to be the 2nd leading cause of outpatient morbidity in the Greater Accra region8, 9.
It is an established fact that most countries of the world are going through different phases of risk transition. Most low and middle income countries including Nigeria are in the interface between traditional and modern risk, otherwise explained as the phase of double burden of diseases, where the challenge of controlling infectious diseases have been caught up by increasing burden of noncommunicable chronic diseases10. Nigeria is not immune to the trend of global health, it is even more vulnerable considering her population growth rate and socio-economic changes, which has led to ageing populations, with dietary and lifestyle modifications, all of which tend to favour the development of non-communicable diseases. Surveillance of key modifiable risk factors is needed to monitor the magnitude of the problem of chronic noncommunicable diseases and to study the effect of interventions. This study therefore aimed at assessing the prevalence of high BP and the factors that influence BP among residents of Ibawa military barracks. The result will also serve as baseline information for evaluation of a planned intervention in the community.
Methods
It was a descriptive cross-sectional study. The study location was 6th battalion, Nigerian Army Barracks, Ibawa, Abak, Akwa Ibom State, Nigeria. The 6th Battalion army barracks is a well defined community within Abak local government area. It has the administrative block (which houses all the offices), the residential quarters, a primary school, a secondary school, church, mosque and a market. The residential quarters where the army officers live with their families, are arranged in platoons. There are a total of 21 platoons in the barracks.
All eligible adults that reside at the barracks were enrolled in the study, however, pregnant women, people with obvious skeletal deformities, especially those affecting the spine and the long bones of the lower limbs, and those whose ages could not be determined were excluded from the study.
Data was collected by the researchers, six research assistants (2 Community health resident doctors and 4 Science graduates) and six community guides, some of who helped to interpret atimes. The participants were interviewed in the homes or offices, depending on where the data collectors met them.
Data collection was done using an interviewer administered semi-structured questionnaire, which was pre-tested and re-designed based on the feedback from the pre-test. It consisted of five main sections:- section A was on the socio-demographic characteristics of the participants, section B was information on the self reported health status, section C was information on activity and sleep pattern, section D was food frequency questions, section E was measurements: anthropometry and blood pressure. Weight and height measurements were taken while observing standard measurement procedures11. The body mass index (BMI) was calculated using the formula, weight in kilogram divided by the square of the height in meters.
Blood pressure measurement was done using aneroid mercury sphygmomanometer, on the left mid-upper arm rested on a table, with the subject in a sitting on a chair with feet flat on the floor. Blood pressure was said to be high if two or more readings were ≥140/90mmHg, systolic/diastolic, at phase I and V Korotkoff sounds. In such cases the blood pressure was re-checked by another observer in the team, and the lowest value obtained would be recorded.
Data entry, validation and analysis were done using SPSS version 17.0 for windows and STATA 10 software. Statistical significance was decided at p value < 0.05.
Results
Four hundred and forty-four people took part in the study, however two of them could not provide information on some of the socio-demographic parameters and therefore they were excluded from the analysis.
Of the 442 whose data were analysed, for some reason, measurements of some of the variables were not taken. This explains why the number “n” is different for the different variables.
The socio-demographic characteristics of the participants show that they were aged 16 to 65 years with a mean age of 31.7 ± 7.6 years, 228 (51.6%) were males and 214 (48.4%) females. Most of them (67.5%) had at least a secondary education. One hundred and sixteen (26.1%) had no sources of income at all, and 95.7% of these were females. However, majority of the participants (44.6%) earned between N30, 000 and N 50,000 per month (equivalent of $200 and $333). One hundred and eighty-one (approximately 41%) of the participants had a maximum of two children, and 45.6% had a maximum total of 4 people in their household.
The males were older, had more abdominal fat and higher BP, both systolic and diastolic (table 1).
Table 1.
Some characteristics of the study population
| Variable | Male | Female | Total | p-value on difference | |||
| No. | Mean (sd) | No. | Mean (sd) | No. Mean (sd) | |||
| Age | 228 | 34.33 (7.62) | 214 | 28.94 (6.58) | 442 | 31.72 (7.62) | < 0.001 |
| BMI | 219 | 26.03 (3.78) | 210 | 26.10 (5.82) | 429 | 26.06 (4.88) | 0.88 |
| Abd. girth | 220 | 90.28 (9.44) | 201 | 87.97 (12.87) | 421 | 89.18 (11.26) | 0.04 |
| SBP | 219 | 134.38 (17.11) | 209 | 123.60 (18.71) | 428 | 129.12 (18.62) | < 0.001 |
| DBP | 219 | 87.56 (12.19) | 209 | 80.69 (13.03) | 428 | 84.21 (13.06) | < 0.001 |
Sd= standard deviation Abd.girth= abdominal girth
Only 200 (46.6%) had normal BMIs, out of these, 103 (24%) were females, X2=17.36 p=0.001 Only 200 (46.6%) had normal BMIs, out of these, 103(24%) were females, X2= 17.36, p=0.001 (table 2). Forty seven percent of the study population had high BP. Difference between the sexes was statistically significant. X2=25.68, p<0.001 (table 3). There was consistently increased BP; both systolic and diastolic, with increase in age, except for those aged >60years. However, there was only one participant in this category (table 4).
Table 2.
Distribution of participants according to BMI and sex (n=429)
| Sex | BMI categories | |||
| <18.5 | 18.5–24.99 | 25.0–29.99 | >30.0 | |
| Male | 0 (0%) | 97 (22.6%) | 92 (21.4%) | 30 (7.0%) |
| Female | 5(1.2%) | 103 (24.0%) | 58 (13.5%) | 44 (10.3%) |
| Total | 5(1.2%) | 200 (46.6%) | 150 (35.0)% | 74 (17.2%) |
Table 3.
Distribution of participants according to blood pressure and sex (n=428)
| BP in mmHg | ||
| Sex | <140/90 | ≥140/90 |
| Male | 90 (210%) | 129 (30.1%) |
| Female | 137 (32.0%) | 72 (16.8%) |
| Total | 227 (53.0%) | 201 (47.0%) |
Forty-seven percent of the study population had high BP
Table 4.
Distribution of participants according to mean blood pressure and age group
| Age group in years |
No. | Mean systolic BP ±SD |
Mean diastolic BP ±SD |
| < 20 | 3 | 107 ±6 | 70±0 |
| 20–30 | 224 | 125±16 | 81±12 |
| 31–40 | 148 | 130±18 | 85±11 |
| 41–50 | 44 | 143±23 | 93±14 |
| 51–60 | 6 | 162±22 | 106±12 |
| >60 | 1 | 140 | 90 |
Table 4 shows that both systolic and diastolic BP increased with increase in age from < 20 years up to 60 years of age.
At a univariate level, for every one year increase in age there appeared to be an increased risk of high BP of 8%, (5–11% at 95% CI), male sex had an OR of 0.38 for having high BP compared to females, income appeared to have a negligible influence, but the greater the number of people in the household the higher the risk of high BP. For every kg/m2 increase in BMI, there is 9% increased risk of having high BP (table 5).
Table 5.
Univariate regression analysis of risk factors for high BP
| High BP | OR | 95%CI | p |
| Age | 1.0831 | 1.05, 1.11 | <0.001 |
| Female Sex | 0.3761 | 0.25, 0.55 | <0.001 |
| Income | 1 | 1.00, 1.00 | <0.001 |
| No. in house | 1.1017 | 1.00, 1.21 | 0.04 |
| Religion | 0.6657 | 0.44, 1.00 | 0.05 |
| Atleast 8 hours of sleep | 0.3767 | 0.15, 0.95 | 0.04 |
| BMI | 1.0946 | 1.05, 1.14 | <0.001 |
| Daily Cigarette smoking | 2.3992 | 1.24, 4.62 | 0.01 |
| Daily alcohol intake | 3.5559 | 1.45, 8.71 | 0.01 |
OR=Odds ratio
In the multivariate analysis, after adjusting for cigarette smoking, alcohol intake and participants' total income, only age, sex and BMI remained as statistically significant predictive factors that influenced high BP with OR: 1.04, 0.45 and 1.08 respectively (table 6). The model had an area under the Receiver Operative Characteristic Curve of 0.719.
Table 6.
Multivariate (adjusted model) regression for risk factors for high BP
| High BP | OR | 95%CI | p |
| Age | 1.0428 | 1.00, 1.08 | 0.030 |
| Sex | 0.4475 | 0.22, 0.90 | 0.025 |
| income | 1.0000 | 1.00, 1.00 | 0.349 |
| No. in house | 1.0646 | 0.94, 1.21 | 0.344 |
| Religion | 0.7769 | 0.48, 1.26 | 0.304 |
| Hours of exercise | 0.9359 | 0.78, 1.13 | 0.488 |
| Hours of sleep | 0.8546 | 0.66, 1.10 | 0.226 |
| BMI | 1.0808 | 1.03, 1.13 | 0.001 |
| Cigarette | 0.9684 | 0.90, 1.05 | 0.420 |
| Alcohol | 0.9667 | 0.90, 1.04 | 0.365 |
At multivariate level, age, sex and BMI remained statistically significant predictors of high BP.
Discussion
For over four decades now, several researchers have studied blood pressure pattern of different populations in different countries of the world, initially using the cut-off of 160/95mmHg to decide on hypertension, but currently by adopting the 140/90 mmHg cut-off. It is expected that the prevalence of high BP that were observed with the 160/95mmHg cut-off would be higher if the current 140/90mmHg cut-off were to be adopted11.
Prevalence of high BP
The prevalence of high BP in this study was 47%, this is slightly higher than the 42.2% observed among a market population in Enugu in 200612, but obviously very much higher than the 18.4% and 28.2% reported from facility based studies of the prevalence of hypertension among medical admissions at UNTH, Enugu13, and UPTH Port Harcourt14 respectively. Even in a semi-urban community in South -Western zone of the country15, a lower prevalence (36.6%) was observed. Several other studies have recorded lower prevalence rates of high BP in Nigeria,16–22 other West African countries 23, 24 and even East Africa25. However, in a facility based study in south west Nigeria, 50.5% of the adults who attended the health facility were found to have high BP26, also economically developed countries have documented prevalence as high as 50%27, and 49.4% of residents in a rural community in Spain had high BP25. The high prevalence of high BP noticed in this study is a reflection of the increasing prevalence of hypertension that is occurring in all regions of the world. In addition, the fact that some of the earlier studies referenced used the 160/95 mmHg cut off, and reported figures were extrapolations from the observed prevalence, explains the difference observed. Furthermore, the high prevalence of overweight/obesity (52.2%) in this study may explain partly the high prevalence of high BP in this population. It is an established fact that obesity is directly related to hypertension; a similar observation was made by Ulasi et al 12 and Maher et al25. However, it may be difficult to compare the findings of different studies, as the definitions and characteristics of settings differ between studies.
Predictors of high BP
The findings from this study showed that BP increased with increase in age. A univariate analysis showed that for every year increase in age, there was an increased risk of high BP to the tune of 8%, p<0.001. The risk was however reduced at a multivariate adjusted level to 4%, p=0.03 (Tables 4 &5). This finding is similar to the finding of several other studies in Nigeria 15, 17, 18 Uganda25 and Ashanti, Ghana29. Sex was seen as one of the predictors of high BP in this study, with the male sex having an odds ratio of 0.45 (p=0.03) at the multivariate level; 30.2% males versus 16.8% females (p<0.001) had high BP in the study. There has been conflicting reports on the association between sex and BP, but review of several articles generally show that men have higher BP than women in most regions of the world4. The finding of this study is corroborated by the findings of most other studies conducted in Nigeria 12, 15–19. However, while some studies in Ghana showed no gender difference 23, 24, 29, in the USA, the prevalence of high BP was higher in females27 and worldwide data as at 2003 revealed that prevalence of hypertension was lowest (3.4%) among rural Indian men and highest (72.5%) among Polish women30. The fact that the men in the present study were older; mean age 34.3 ±7.6 years versus 28.9 ± 6.6 years in females (p<0.001), and more overweight/obese; 28.4% versus 23.8% of females (p=0.001), explains why more men had high BP in the study.
Apart from age and sex, BMI was also observed as a predictor of BP from the data. The prevalence of overweight and obesity in this study was 52.2%; overweight 35%, obesity 17.2%. The obesity prevalence of 17.2% is lower than the 19.2 % and 22.6% observed by Ulasi et al in a semi-urban community and among market workers respectively12. However, the data showed that BMI is a positive predictor of high BP, OR 1.08, (CI 1.03, 1.13. p=0.001), which is similar to the finding from several other studies 12,16,17,25, 26. Several other factors such as cigarette smoking, poor diets (eating of unhealthy/salt-laden diets) and physical inactivity (sedentary lifestyle) have been reported as contributors to high BP 12, but the present study did not establish the relationship between these factors and high BP. It was surprising to note that cigarette smoking was a statistically significant predictor of high BP at the univariate level only. Also, it is not clear if the availability of social amenities in the barracks; electricity and borehole water, may have encouraged the residents to spend hours watching films and TV, leading to relatively sedentary lifestyle, and thus increase in BP; or to what extent their BP at recruitment could have contributed to the observed BPs, noting that it has been documented that that even among army recruits, high BP was seen as the commonest condition28. Further epidemiological research will be required to confirm or refute such associations in this population. In conclusion, the prevalence of high BP in this semi-urban community is quite high (47%), the predictive factors were age, sex and BMI. This implies that there is an impending epidemic of cardiovascular diseases in this community, hence the need to develop health delivery tools for early identification and prompt treatment of hypertension. Additionally, public health measures to combat the growing risks should be instituted. Lifestyle modification in addition to treatment have been proven to decrease high BP and prevent its complications, the time to act is now.
References
- 1.Lawes CM, Vander Hoorn S, Law MR, Elliott P, MacMahon S, Rodgers A. Blood pressure and the global burden of disease 2000. Part II: estimates of attributable burden. J Hypertens. 2006 Mar;24(3):423–430. doi: 10.1097/01.hjh.0000209973.67746.f0. [DOI] [PubMed] [Google Scholar]
- 2.WHO, author. Global health risks: mortality and burden of disease attributable to selected major risks. 2009 [Google Scholar]
- 3.Lawes CM, Vander Hoorn S, Rodgers A. Global burden of blood-pressure-related disease, 2001. Lancet. 2008 May 3;371(9623):1513–1518. doi: 10.1016/S0140-6736(08)60655-8. [DOI] [PubMed] [Google Scholar]
- 4.Danaei G, Finucane MM, Lin JK, et al. on behalf of the Global Burden of Metabolic Risk Factors of Chronic Diseases Collaborating Group (Blood Pressure). National, regional, and global trends in systolic blood pressure since 1980: systematic analysis of health examination surveys and epidemiological studies with 786 country-years and 5.4 million participants. Lancet. 2011 doi: 10.1016/S01406736(10)62036-3. published online Feb 4. [DOI] [PubMed] [Google Scholar]
- 5.Cooper RS, Amoah AG, Mensah GA. High blood pressure: the foundation for epidemic cardiovascular disease in African populations. Ethn Dis. 2003 summer;13(2 Suppl 2):S48–S52. [PubMed] [Google Scholar]
- 6.Kannel WB, Wolf PA, Verter J, et al. Epidemiologic assessment of the role of blood pressure in stroke. The Framinghan study. JAMA. 1970;214:301–310. [PubMed] [Google Scholar]
- 7.Maseko MJ, Norton GR, Majane OH, Molebatsi N, Woodiwiss AJ. Global cardiovascular risk profiles of untreated hypertensives in an urban, developing community in Africa. Cardiovasc J Afr. 2010 Dec 15;21:1–7. doi: 10.5830/CVJA-2010-094. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bosu WK. Epidemic of hypertension in Ghana: a systematic review. BMC Public Health. 2010 Jul 14;10:418. doi: 10.1186/1471-2458-10-418. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Greater Accra regional Health Directorate, author. Annual Report 2007. Accra: Ghana Health Service; 2008. [Google Scholar]
- 10.Alan DL, Colin DM, Majid E, Dea TJ, Christopher JM. Global burden of disease and risk factors. Oxford University Press and the World Bank; p. 3. Chapter 1. [Google Scholar]
- 11.The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure (The JNC 7 Report) JAMA. 2003;289(19):2560–2572. doi: 10.1001/jama.289.19.2560. [DOI] [PubMed] [Google Scholar]
- 12.Ulasi II, Ijoma CK, Onwubere BJ, Arodiwe E, Onodugo O, Okafor C. High Prevalence and Low Awareness of Hypertension in a Market Population in Enugu, Nigeria. International Journal of Hypertension. 2011 doi: 10.4061/2011/869675. Article ID 869675, 5 pages. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ike SO. Prevalence of hypertension and its complications among medical admissions at the University of Nigeria Teaching Hospital, Enugu (study 2) Niger J Med. 2009;18(1):68–72. [PubMed] [Google Scholar]
- 14.Onwuchekwa AC, Chineye S. Clinical profile of hypertension at a University Teaching Hospital in Nigeria. Vasc Health Risk Manag. 2010;9(6):511–516. doi: 10.2147/vhrm.s10245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Rufus AA, Chidozie EM, Michael OB, Tanimola M, Rasaaq AA, Anthony A, Patience OA. Prevalence and pattern of hypertension in a semi-urban community in Nigeria. Eur J Cardiovasc Prev Rehabil. 2008;15:683–687. doi: 10.1097/HJR.0b013e32830edc32. [DOI] [PubMed] [Google Scholar]
- 16.Clareann HB, Flora AU, Martin UN, Jackson AO, Glenn WC, Linda HK, et al. Factors associated with hypertension in Nigerian civil servants. Preventive Medicine. 1992;21(6):710–722. doi: 10.1016/0091-7435(92)90078-v. [DOI] [PubMed] [Google Scholar]
- 17.Oghagbon EK, Okesina AB, Biliaminu SA. Prevalence of hypertension and associated variables in paid workers in Ilorin, Nigeria. Nig J of Cl Practice. 2008;11(4):342–346. [PubMed] [Google Scholar]
- 18.Obinna IE, Patrick OU, Izuchukwu LN. Prevalence, awareness, treatment and control of hypertension in a Nigerian population. doi: 10.4236/health.2010.27111. [DOI] [Google Scholar]
- 19.Olatunbosun ST, Kaufunan JS, Cooper RS, Bella AF. Hypertension in a black population: prevalence and biosocial determinants of high BP in a group of urban Nigerians. Journal of Human Hypertension. 2000;14(4):249–257. doi: 10.1038/sj.jhh.1000975. [DOI] [PubMed] [Google Scholar]
- 20.Non-Communicable Diseases (NCD) in Nigeria - final report of a national survey. Federal Ministry of Health-National Expert Committee on NCD; 1997. [Google Scholar]
- 21.Lawoyin TO, Asuzu MC, Kaufman J, et al. Prevalence of cardiovascular risk factors in an African urban, inner city community. West Afr J Med. 2002;21:208–211. doi: 10.4314/wajm.v21i3.28031. [DOI] [PubMed] [Google Scholar]
- 22.Kadiri S, Walker O, Salako BL, Akinkugbe OO. Blood pressure, hypertension and correlates in urbanized workers in Ibadan, Nigeria - a revisit. J Hum Hypertens. 1999;13:23–27. doi: 10.1038/sj.jhh.1000722. [DOI] [PubMed] [Google Scholar]
- 23.Amoah AB. Hypertension in Ghana: a cross-sectional community prevalence study in Greater Accra. Ethn Dis. 2003;13:310–315. [PubMed] [Google Scholar]
- 24.Cooper R, Rotimi C, Ataman S, McGee D, Osotmehin B, Kadiri S, Muna W, Kingue S, Fraser H, Forrester T, Bennett F, Wilks R. The prevalence of hypertension in seven populations of West African origin. Am J Public Health. 1997;87:160–168. doi: 10.2105/ajph.87.2.160. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Maher D, Waswa L, Baisley K, Karabarinde A, Unwin N. Epidemiology of hypertension in low-income countries: a cross sectional population-based survey in rural Uganda. doi: 10.1097/HJH.Ob013e3283466e90. [DOI] [PubMed] [Google Scholar]
- 26.Amole IO, Olaolorun AD, Odeigah LO, Adesina SA. The prevalence of abdominal obesity and hypertension among adults in Ogbomoso, Nigeria. Afr J Prim Health care Fam Med. 2011;3(1) doi: 10.4102/phcfm.v3. Art 188, 5pages. 188. [DOI] [Google Scholar]
- 27.Hall WD, Ferrario CM, Moore MA, Hall JE, Flack JM, Cooper W, et al. hypertension-related morbidity and mortality in the southeastern United States. Am J Med Sci. 1997;313(4):195–209. doi: 10.1097/00000441-199704000-00002. [DOI] [PubMed] [Google Scholar]
- 28.Awoyemi AO, Osagbemi GK, Ogunleye VA. Medical examination findings among army recruits in Ilorin. West Afr J Med. 2001;20(3):256–258. [PubMed] [Google Scholar]
- 29.Cappuccio Francesco P, Micah Frank B, Emmett Lynsey, Kerry Sally M, Antwi Samson, Martin-Peprah Ruby, Phillips Richard O, Plange-Rhule Jacob, Eastwood John B. Prevalence, Detection, Management, and Control of Hypertension in Ashanti, West Africa. Hypertension. 2004;43:1017–1022. doi: 10.1161/01.HYP.0000126176.03319.d8. [DOI] [PubMed] [Google Scholar]
- 30.Kearney PM, Whelton M, Reynolds K, Whellton PK, He J. Worldwide prevalence of hypertension: a systematic review. J Hypertens. 2004;22:11–19. doi: 10.1097/00004872-200401000-00003. [DOI] [PubMed] [Google Scholar]