ABSTRACT

Objective: To assess and compare the prevalence of comorbidities and risk factors for cardiovascular disease (CVD) in COPD patients according to disease severity. Methods: The study included 25 patients with mild-to-moderate COPD (68% male; mean age, 65 ± 8 years; mean FEV1, 73 ± 15% of predicted) and 25 with severe-to-very severe COPD (males, 56%; mean age, 69 ± 9 years; mean FEV1, 40 ± 18% of predicted). Comorbidities were recorded on the basis of data obtained from medical charts and clinical evaluations. Comorbidities were registered on the basis of data obtained from medical charts and clinical evaluations. The Charlson comorbidity index was calculated, and the Hospital Anxiety and Depression Scale (HADS) score was determined. Results: Of the 50 patients evaluated, 38 (76%) had been diagnosed with at least one comorbidity, 21 (42%) having been diagnosed with at least one CVD. Twenty-four patients (48%) had more than one CVD. Eighteen (36%) of the patients were current smokers, 10 (20%) had depression, 7 (14%) had dyslipidemia, and 7 (14%) had diabetes mellitus. Current smoking, depression, and dyslipidemia were more prevalent among the patients with mild-to-moderate COPD than among those with severe-to-very severe COPD (p < 0.001, p = 0.008, and p = 0.02, respectively). The prevalence of high blood pressure, diabetes mellitus, alcoholism, ischemic heart disease, and chronic heart failure was comparable between the two groups. The Charlson comorbidity index and HADS scores did not differ between the groups. Conclusions: Comorbidities are highly prevalent in COPD, regardless of its severity. Certain risk factors for CVD, themselves classified as diseases (including smoking, dyslipidemia, and depression), appear to be more prevalent in patients with mild-to-moderate COPD.

Keywords: Pulmonary disease, chronic obstructive; Spirometry; Cardiovascular diseases; Risk factors.

RESUMO

Objetivo: Avaliar e comparar a prevalência de comorbidades e de fatores de risco de doença cardiovascular (DCV) em pacientes com DPOC de acordo com a gravidade da doença. Métodos: O estudo incluiu 25 pacientes com DPOC leve/moderada (homens: 68%; média de idade: 65 ± 8 anos; média de VEF1: 73 ± 15% do previsto) e 25 com DPOC grave/muito grave (homens: 56%; média de idade: 69 ± 9 anos; média de VEF1, 40 ± 18% do previsto). As comorbidades foram registradas com base nos dados dos prontuários médicos e avaliações clínicas. O índice de comorbidades de Charlson foi calculado, e a pontuação na Hospital Anxiety and Depression Scale (HADS) foi determinada. Resultados: Dos 50 pacientes avaliados, 38 (76%) receberam diagnóstico de pelo menos uma comorbidade, sendo que 21 (42%) receberam diagnóstico de pelo menos uma DCV. Vinte e quatro pacientes (48%) apresentavam mais de uma DCV. Dezoito pacientes (36%) eram fumantes, 10 (20%) tinham depressão, 7 (14%) apresentavam dislipidemia, e 7 (14%) tinham diabetes mellitus. Tabagismo atual, depressão e dislipidemia foram mais prevalentes nos pacientes com DPOC leve/moderada que naqueles com DPOC grave/muito grave (p < 0,001, p = 0,008 e p = 0,02, respectivamente). A prevalência de pressão arterial elevada, diabetes mellitus, alcoolismo, doença isquêmica do coração e insuficiência cardíaca crônica foi semelhante nos dois grupos. O índice de comorbidades de Charlson e a pontuação na HADS não diferiram entre os grupos. Conclusões: Comorbidades são muito prevalentes na DPOC, independentemente da gravidade da doença. Certos fatores de risco de DCV, eles próprios considerados doenças (incluindo tabagismo, dislipidemia e depressão), parecem ser mais prevalentes nos pacientes com DPOC leve/moderada.

Palavras-chave: Doença pulmonar obstrutiva crônica; Espirometria; Doenças cardiovasculares; Fatores de risco.

INTRODUCTION

COPD is characterized by chronic airflow limitation, a range of pathological changes in the lungs, significant ex-trapulmonary effects, and major comorbidities that can contribute to increasing the severity of the disease. (1) Com-mon comorbidities in patients with COPD include cardiovascular disease (CVD), anemia, lung cancer, diabetes, oste-oporosis, anxiety, and depression.(2) Among COPD patients, CVDs are responsible for approximately 50% of all hospitalizations and 20% of all deaths.(3)

The majority (94%) of all COPD patients have at least one comorbidity and up to 46% have three or more,(4) the most prevalent being high blood pressure (HBP), coronary artery disease (CAD), congestive heart failure (CHF), dyslipidemia, and diabetes mellitus (DM).(2,5) Two previous studies evaluated the prevalence of comorbidities ac-cording to COPD severity,(5,6) although neither evaluated all of the risk factors for CVD. Therefore, the aim of the present study was to assess the prevalence of comorbidities and risk factors for CVD in COPD patients, comparing them by COPD severity-mild-to-moderate versus severe-to-very severe.

METHODS

Seventy patients with COPD were recruited from the Pulmonary Outpatient Unit at the Botucatu Hospital das Clí-nicas of the São Paulo State University Botucatu School of Medicine. Patients were included if they met the Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria for a diagnosis of COPD(1): a post-bronchodilator FEV1/FVC ratio < 0.70 and a < 15% or 200-mL increase in FEV1 following inhalation of a β2 agonist; age ≥ 40 years; and a ≥ 10 pack-year smoking history. Patients with a primary diagnosis of another respiratory disease, such as asthma, a restrictive disorder (tuberculosis sequelae or interstitial fibrosis), obstructive sleep apnea-hypopnea syn-drome, and lung cancer, were excluded, as were those with a primary diagnosis of unstable angina, CHF (New York Heart Association functional class III or IV), or any other chronic disease, such as uncontrolled DM, kidney failure, liver failure, and cancer. We took into consideration the post-bronchodilator values of FEV1 (% of the predicted value) and arterial blood gas parameters. We categorized the severity of COPD according to the GOLD stages,(1) as mild-to-moderate (GOLD stage I or II) or severe-to-very severe (GOLD stage III or IV).

The study was approved by the Research Ethics Committee of the Botucatu Hospital das Clínicas. All participating patients gave written informed consent.

Laboratory tests, pulmonary function tests, and pulse oximetry

Laboratory tests included complete blood counts, as well as determination of the following: lipid profile; total and fractional protein levels; fasting glucose level; hepatic function; and renal function. The biochemical tests were per-formed according to the methods applied by the Clinical Analysis Laboratory of the institution.

Pre- and post-bronchodilator spirometry were performed with a portable spirometer (Koko; Ferraris Respiratory, Louisville, CO, USA), in accordance with the American Thoracic Society criteria.(7) Values of FEV1 were expressed in liters, as percentages of the FVC, and as percentages of the reference values.(8) The SpO2 was assessed with an oximeter (Onyx 9500; Nonin Medical Inc., Plymouth, MN, USA) while the patients were breathing room air.

Comorbidities

Comorbidities were identified in medical records and confirmed at clinical evaluation. Current smoking was consid-ered a comorbidity, in accordance with the International Statistical Classification of Diseases, 10th Revision, in which nicotine dependence is coded as F17. The smoking status was confirmed by measuring carbon monoxide (CO) in exhaled air with a CO analyzer (Micro CO Meter, Cardinal Health, Chatham, UK). An exhaled CO level > 6.0 ppm was considered indicative of current smoking.(9)

For each patient, we calculated the Charlson comorbidity index.(10) The Hospital Anxiety and Depression Scale (HADS), which has been translated to Portuguese and validated for use in Brazil,(11) was used in order to evaluate symptoms related to anxiety and depression.

Nutritional assessment

We determined body weight and height using a calibrated platform scale with a stadiometer (Filizola, São Paulo, Brazil). The body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared (kg/m2). Body composition was evaluated by bioelectrical impedance (BIA 101A; RJL Systems Inc., Clinton Township, MI, USA), in accordance with the guidelines established by the European Society for Parenteral and Enteral Nutri-tion.(12) Fat-free mass (FFM, in kg) was calculated through the use of a group-specific regression equation developed by Kyle et al.(13) We also calculated the FFM index (FFMI), using the following equation:
FFMI = FFM/height2

Statistical analyses

Descriptive statistics were calculated for the features of all participants. Data are expressed as mean ± standard deviation or as median and interquartile range (25-75%), depending on their distribution. Categorical variables are expressed as absolute and relative frequency. The chi-square test and Fisher's exact test were used in order to com-pare categorical variables. For comparisons between the two groups, an unpaired t-test was used for continuous variables and the Mann-Whitney test was used for ordinal variables. The level of significance was set at 5%. All analyses were performed with the SigmaStat program, version 3.2 (Systat Software Inc., San Jose, CA, USA).

RESULTS

Twenty patients were excluded from the final analysis, 15 because their primary diagnosis was not COPD and 5 because they did not complete the protocol. Therefore, 50 patients were included in the final analysis: 25 with mild-to-moderate COPD and 25 with severe-to-very severe COPD. Age, gender, smoking history (pack-years), BMI, and FFMI were not statistically different between the groups (Table 1). Of the 50 patients evaluated, 9 (18%) were obese (BMI ≥ 30 kg/m2), 3 in the mild-to-moderate COPD group and 6 in the severe-to-very severe COPD group (p = 0.39).(14) Five patients had very severe COPD and were on long-term oxygen therapy. Figure 1 shows the mainte-nance medications used by the patients.
 

 





Thirty-eight patients (76%) presented with at least one comorbidity, 21 (42%) presenting with at least one cardio-vascular comorbidity: HBP, in 40% of the sample; CAD, in 10%; or New York Heart Association functional class I CHF, in 6%. Twenty-four patients (48%) presented more than one CVD. Of the 50 patients evaluated, 18 (36%) were clas-sified as active smokers, 10 (20%) were diagnosed with depression, 7 (14%) were diagnosed with dyslipidemia, 7 (14%) were diagnosed with DM, and 4 (8%) were diagnosed with alcoholism.

Of the 25 patients in the mild-to-moderate COPD group, 14 (56%) were classified as current smokers, 5 (20%) had been diagnosed with dyslipidemia, and 7 (28%) had been diagnosed with depression. Of the 25 patients in the se-vere-to-very severe COPD group, only 4 (16%) were classified as current smokers, only 2 (8%) had been diagnosed with dyslipidemia, and only 3 (12%) had been diagnosed with depression.

Current smoking and dyslipidemia were more prevalent in the mild-to-moderate COPD group than in the severe-to-very severe COPD group (p < 0.001 and p = 0.02, respectively), whereas the prevalence of HBP, DM, alcoholism, CAD, and CHF was comparable between the two groups (Table 2). Although the prevalence of depression was signifi-cantly higher in the mild-to-moderate COPD group (p=0.008), the HADS scores were not significantly different be-tween the groups (p = 0.93 and p = 0.89 for anxiety and depression, respectively). All 7 of the patients who had previously been diagnosed with depression were receiving pharmacological treatment.
 



In the majority of the patients diagnosed with dyslipidemia, the concentrations of lipids (total cholesterol, HDL, LDL, and triglycerides) were within normal values, and the lipid profile, according to the Brazilian Society of Cardiology IV Brazilian Guidelines on Dyslipidemia and Prevention of Atherosclerosis,(15) was similar between the two groups. All 7 of the patients who had previously been diagnosed with dyslipidemia were receiving pharmacological treatment.

The Charlson comorbidity index did not differ between the mild-to-moderate COPD and severe-to-very severe COPD groups. For the sample as a whole, the mean Charlson comorbidity index was 3.5 ± 1.3.

DISCUSSION

The main finding of the present study is that risk factors for CVD, including smoking and dyslipidemia, were more prevalent among patients with mild-to-moderate COPD than among those with severe-to-very severe COPD. The prevalence of depression was also higher in the former group.

The prevalence of current smoking in our sample was 36%. Shahab et al.(16) evaluated 1,093 COPD patients and found a similar (34.9%) prevalence of current smoking. In the Projeto Latino-Americano de Investigação em Ob-strução Pulmonar (PLATINO, Latin American Project for the Investigation of Obstructive Lung Disease) study,(17) the observed prevalence of smoking among patients with COPD categorized as GOLD stage II-IV was 26.2%, which, albeit high, is lower than that identified in our sample. That can be explained, at least in part, by the fact that the PLATINO study sample included subjects who lived in the greater metropolitan area of the city of São Paulo, whereas our sample was composed of subjects living in small towns or rural areas of the state.(17) Black-Shinn et al.(18) also showed that the prevalence of current smoking was higher among patients in the early stages of COPD than among those with severe COPD, which is in line with the findings of another study, which reported a positive association between being a former smoker and having severe airflow limitation.(19) Similarly, in the Evaluation of COPD Longi-tudinally to Identify Predictive Surrogate Endpoints (ECLIPSE) study, patients with GOLD stage II, III, and IV COPD were compared, and the proportion of current smokers decreased in parallel with increased disease severity.(5) How-ever, there are some data showing that the rates of current smoking actually increase as COPD becomes more se-vere.(16)

Dyslipidemia was present in 14% of the patients in our sample. This is considerably lower than the 40% prevalence of dyslipidemia (38% among males and 42% among females) reported for the general population of Brazil(15) and the 39% reported in an international study.(20) Two previous studies reported the prevalence of dyslipidemia among COPD patients to be 26.5% and 28.0%, respectively.(21,22) In our study, a diagnosis of dyslipidemia was more preva-lent among the patients with mild-to-moderate COPD than among those with severe-to-very severe COPD. That finding could be attributable to the fact that the number of current smokers is usually higher among the patients with mild-to-moderate COPD.(23) To our knowledge, there have been no previous studies comparing the prevalence of dyslipidemia among the stages of COPD severity. Our findings regarding dyslipidemia merit attention, especially in view of the fact that one cross-sectional study of COPD patients showed that the rate of first cardiovascular event (myocardial infarction and stroke), which is usually associated with dyslipidemia, was higher among younger sub-jects than among those of more advanced age.(24) The authors of that study suggested that the risk of cardiovascular mortality is higher in patients with mild-to-moderate COPD than in those with severe-to-very severe COPD. It has also been demonstrated that lung cancer and cardiovascular comorbidities constitute the leading cause of death in patients with mild-to-moderate COPD, accounting for up to two thirds of all deaths among such patients, whereas respiratory failure is the predominant cause of death in patients with advanced COPD.(4)

Depression was diagnosed in 20% of our patients. That is in agreement with the findings of two previous studies, in which the reported prevalence of depression among COPD patients was 23.1% and 35.0%, respectively.(25,26) In the ECLIPSE study, the overall prevalence of depression in COPD patients was 17%, and there was a statistically signifi-cant difference between GOLD stages among females.(5) In another study of COPD patients, Echave-Sustaeta et al.(27) reported that the prevalence of depression was 11.1% and found no association between depression and COPD severity. The higher number of current smokers among patients with mild-to-moderate COPD is a possible explana-tion for the higher prevalence of depression in such patients. In fact, it has previously been shown that anxiety and depression scores are associated with current smoking.(28,29)

In the present study, the mean Charlson comorbidity index did not differ between the mild-to-moderate COPD and severe-to-very severe COPD groups. For the sample as a whole, the mean Charlson comorbidity index was 3.5, similar to the 3.9 reported by Díez-Manglano et al.,(22) although higher than the 2.5 reported by Echave-Sustaeta et al.(27) and the 2.7 reported by Almagro et al.(2) Although COPD, myocardial infarction, CHF, and DM are included in the Charlson comorbidity index, other highly prevalent diseases observed in our study, including current smoking, HBP, dyslipidemia, alcoholism, and depression, are not. Therefore, the exclusive use of the Charlson comorbidity index could result in an underestimation of the prevalence of comorbidities and of their influence on the prognosis of COPD patients. After age and chronic symptoms, comorbidities are the most important predictive factors of the future health care costs and direct costs of COPD.(4) In addition, our findings underscore the need for a new comorbidity index to be applied in patients with COPD, given that some highly prevalent comorbidities are not included in the index currently available.

Our study has certain limitations. Although the comorbidities observed in our study are similar to those identified in the current literature, the prospective design and characteristics of the patient sample could explain some of our findings. In addition, the small size of our sample might impose some limitations on the interpretation of our data.

In conclusion, the prevalence of comorbidities is high among patients with COPD, regardless of disease severity. In addition, smoking and dyslipidemia appear to be more prevalent in mild-to-moderate COPD. Therefore, controlling these comorbidities might be a key measure in the early phases of the disease, in order to decrease mortality due to cardiovascular events. Our findings show the importance of therapeutic measures to promote smoking cessation and of early diagnosis to prevent the progression of airflow obstruction.

REFERENCES

1. Global Initiative for Chronic Obstructive Lung Disease [homepage on the Internet]. Bethesda: Global Initiative for Chronic Obstructive Lung Disease. [cited 2014 Oct 1]. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD Executive Summary, update 2014. Available from: http://www.goldcopd.org/
2. Almagro P, López García F, Cabrera F, Montero L, Morchón D, Díez J, et al. Comorbidity and gender-related differences in patients hospitalized for COPD. The ECCO study. Respir Med. 2010;104(2):253-9. http://dx.doi.org/10.1016/j.rmed.2009.09.019
3. Sin DD, Man SF. Impact of cancers and cardiovascular disease in chronic obstructive pulmonary disease. Curr Opin Pulm Med. 2008;14(2):115-21. http://dx.doi.org/10.1097/MCP.0b013e3282f45ffb
4. Hillas G, Perlikos F, Tsiligianni I, Tzanakis N. Managing comorbidities in COPD. Int J Chron Obstruct Pulmon Dis. 2015;10:95-109. http://dx.doi.org/10.2147/COPD.S54473
5. Agusti A, Calverley PM, Celli B, Coxson HO, Edwards LD, Lomas DA, et al. Characterisation of COPD heterogeneity in the ECLIPSE cohort. Respir Res. 2010;11:122. http://dx.doi.org/10.1186/1465-9921-11-122
6. Pedone C, Scarlata S, Sorino C, Forastiere F, Bellia V, Antonelli Incalzi R. Does mild COPD affect prognosis in the elderly? BMC Pulm Med. 2010;10:35. http://dx.doi.org/10.1186/1471-2466-10-35
7. Standardization of spirometry--1987 update. Statement of the American Thoracic Society. Am Rev Respir Dis. 1987;136(5):1285-98. http://dx.doi.org/10.1164/ajrccm/136.5.1285
8. Pereira CA, Barreto SP, Simões JG, Pereira FW, Gerstler JG, Nakatani J. Valores de referência para a espirometria em uma amostra da população brasileira adulta. J Pneumol. 1992;18(1):10-22.
9. Middleton ET, Morice AH. Breath carbon monoxide as an indication of smoking habit. Chest. 2000;117(3):758-63. http://dx.doi.org/10.1378/chest.117.3.758
10. Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol. 1994;47(11):1245-51. http://dx.doi.org/10.1016/0895-4356(94)90129-5
11. Botega NJ, Bio MR, Zomignani MA, Garcia C Jr, Pereira WA. Mood disorders among medical in-patients: a validation study of the hospital anxiety and depression scale (HAD) [Article in Portuguese]. Rev Saude Publica. 1995;29(5):359-63.
12. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Gómez JM, et al. Bioelectrical impedance analysis--part I: review of principles and methods. Clin Nutr. 2004;23(5):1226-43. http://dx.doi.org/10.1016/j.clnu.2004.06.004
13. Kyle UG, Pichard C, Rochat T, Slosman DO, Fitting JW, Thiebaud D. New bioelectrical impedance formula for patients with respiratory insufficiency: comparison to dual-energy X-ray absorptiometry. Eur Respir J. 1998;12(4):960-6. http://dx.doi.org/10.1183/09031936.98.12040960
14. World Health Organization [homepage on the Internet]. Geneva: World Health Organization; c2015 [cited 2015 May 10]. Overweight and obesity fact sheet 2015. Available from: http://www.searo.who.int/linkfiles/non_communicable_diseases_obesety-fs
15. Sposito AC, Caramelli B, Fonseca FA, Bertolami MC, Afiune Neto A, Souza AD, et al. IV Brazilian Guideline for Dyslipidemia and Atherosclerosis prevention: Department of Atherosclerosis of Brazilian Society of Cardiology [Article in Portuguese]. Arq Bras Cardiol. 2007;88 Suppl 1:2-19. http://dx.doi.org/10.1590/S0066-782X2007000700002
16. Shahab L, Jarvis MJ, Britton J, West R. Prevalence, diagnosis and relation to tobacco dependence of chronic obstructive pulmonary disease in a nationally representative sample. Thorax. 2006;61(12):1043-47. http://dx.doi.org/10.1136/thx.2006.064410
17. Menezes AM, Jardim JR, Pérez-Padilla R, Camelier A, Rosa F, Nascimento O, et al. Prevalence of chronic obstructive pulmonary disease and associated factors: the PLATINO Study in São Paulo, Brazil. Cad Saude Publica. 2005;21(5):1565-73. http://dx.doi.org/10.1590/S0102-311X2005000500030
18. Black-Shinn JL, Kinney GL, Wise AL, Regan EA, Make B, Krantz MJ, et al. Cardiovascular disease is associated with COPD severity and reduced functional status and quality of life. COPD. 2014;11(5):546-51. http://dx.doi.org/10.3109/15412555.2014.898029
19. Mitsiki E, Bania E, Varounis C, Gourgoulianis KI, Alexopoulos EC. Characteristics of prevalent and new COPD cases in Greece: the GOLDEN study. Int J Chron Obstruct Pulmon Dis. 2015;10:1371-82. http://dx.doi.org/10.2147/COPD.S81468
20. Leiter LA, Lundman P, da Silva PM, Drexel H, Jünger C, Gitt AK, et al. Persistent lipid abnormalities in statin-treated patients with diabetes mellitus in Europe and Canada: results of the Dyslipidemia International Study. Diabet Med. 2011;28(11):1343-51. http://dx.doi.org/10.1111/j.1464-5491.2011.03360.x
21. Nagorni-Obradovic LM, Vukovic DS. The prevalence of COPD co-morbidities in Serbia: results of a national survey. NPJ Prim Care Respir Med. 2014;24:14008. http://dx.doi.org/10.1038/npjpcrm.2014.8
22. Díez-Manglano J, Bernabeu-Wittel M, Escalera-Zalvide A, Sánchez-Ledesma M, Mora-Rufeti A, Nieto-Martín D, et al. Comorbidity discapacity and mortality in patients with multiple conditions and chronic obstructive pulmonary disease [Article in Spanish]. Rev Clin Esp. 2011;211(10):504-10. http://dx.doi.org/10.1016/j.rce.2011.04.006
23. Forti N, Diament J. High-density lipoproteins: metabolic, clinical, epidemiological and therapeutic intervention aspects. An update for clinicians. Arq Bras Cardiol. 2006;87(5):671-9. http://dx.doi.org/10.1590/S0066-782X2006001800019
24. Feary JR, Rodrigues LC, Smith CJ, Hubbard RB, Gibson JE. Prevalence of major comorbidities in subjects with COPD and incidence of myocardial infarction and stroke: a comprehensive analysis using data from primary care. Thorax. 2010;65(11):956-62. http://dx.doi.org/10.1136/thx.2009.128082
25. Schneider C, Jick SS, Bothner U, Meier CR. COPD and the risk of depression. Chest. 2010;137(2):341-7. http://dx.doi.org/10.1378/chest.09-0614
26. Wagena EJ, Arrindell WA, Wouters EF, van Schayck CP. Are patients with COPD psychologically distressed? Eur Respir J. 2005;26(2):242-8. http://dx.doi.org/10.1183/09031936.05.00010604
27. Echave-Sustaeta JM, Comeche Casanova L, Cosio BG, Soler-Cataluña JJ, Garcia-Lujan R, Ribera X. Comorbidity in chronic obstructive pulmonary disease. Related to disease severity? Int J Chron Obstruct Pulmon Dis. 2014;9:1307-14. http://dx.doi.org/10.2147/COPD.S71849
28. Northrop-Clewes CA, Thurnham DI. Monitoring micronutrients in cigarette smokers. Clin Chim Acta. 2007;377(1-2):14-38. http://dx.doi.org/10.1016/j.cca.2006.08.028
29. Saravanan C, Heidhy I. Psychological problems and psychosocial predictors of cigarette smoking behavior among undergraduate students in Malaysia. Asian Pac J Cancer Prev. 2014;15(18):7629-34. http://dx.doi.org/10.7314/APJCP.2014.15.18.7629