Abstract
Background: Obesity is an important risk factor for decrease in high-density lipoprotein cholesterol (HDL-C) levels, which predisposes to cardiovascular diseases. But, the relative contribution of obesity toward decreased HDL-C and the risk estimates of decreased HDL-C attributable to obesity are unavailable. Such measures will help in understanding the extent by which the burden of decreased HDL-C can be reduced by tackling obesity. Objectives: The objectives of this study were to (a) determine the association between decreased HDL-C and obesity and (b) estimate the attributable risk proportion (ARP) and population attributable risk proportion (PARP) for decreased HDL-C due to obesity. Methods: Body mass index (BMI) and waist circumference (WC) were measured as indices of overweight (or generalized obesity) and central obesity, respectively in 190 subjects (95 cases with low HDL-C and 95 healthy controls with normal HDL-C) from Guwahati city. Crude odds ratio (OR) and adjusted OR with 95% confidence interval (CI) were calculated along with the risk estimates (ARP and PARP). Results: People with overweight or generlized obesity (adjusted OR = 4.90, 95% CI = 3.59-6.68), and people with central obesity (adjusted OR = 3.33, 95% CI = 2.39-4.64) had significantly greater odds of developing decreased HDL-C. Among the exposed, 79.8% of the decreased HDL-C cases could be attributed to overweight (or generalized obesity), while 72.8% cases could be attributed to central obesity. In the overall population, the corresponding figures were 57.1% and 36%, respectively. Conclusion: Decreased HDL-C is strongly associated with and largely attributable to obesity.
Keywords: body mass index, waist circumference, attributable risk proportion, population attributable risk proportion, high-density lipoprotein, Guwahati, Assam
Introduction
High-density lipoprotein cholesterol (HDL-C) occupies a special position among the different lipid fractions. Also dubbed as “good cholesterol,” decreased HDL-C concentration is a major risk factor for cardiovascular diseases (CVDs).1 Asian Indians commonly have decreased HDL-C levels.2,3 This assumes significance because Asian Indians are also one of the worst affected ethnic groups afflicted by CVDs.4 Obesity/overweight is an important risk factor for decreased HDL-C.5 The inverse correlation between obesity and HDL-C is well documented.6-9 But the magnitude to which obesity is responsible for contributing toward low HDL-C is unclear. Measure of the relative significance of a risk factor in producing an adverse outcome has public health importance. It helps in understanding the extent by which the outcome (decreased HDL-C) can be reduced if the risk factor (obesity) can be eliminated. Estimates like attributable risk proportion (ARP) and population attributable risk proportion (PARP) may offer valuable insights in this regard.10 The ARP and PARP values reflect the extent to which the outcome can be attributed to the exposure (risk factor) in the exposed and general populations, respectively.10,11
Guwahati city is located in the state of Assam, northeast India. Decreased HDL-C concentration was found to be the commonest lipid abnormality in Guwahati.7 Our primary objective was to determine the strength of association between obesity and decreased HDL-C in a sample from Guwahati. In addition, we estimated the risk (ARP and PARP) for decreased HDL-C that was attributable to obesity.
Methods
Study Design
This analytical cross-sectional study was conducted between January 2014 and June 2014 in Gauhati Medical College and Hospital (GMCH), Guwahati, Assam, with Institutional Ethical Committee approval and participants’ informed written consent. All the subjects were residents of Guwahati. Obesity was treated as the exposure (risk factor) and decreased concentration of HDL-C as the outcome. Subjects were enrolled into 1 of the 2 groups—cases (decreased HDL-C group) and controls (normal HDL-C group), using the following selection criteria.
The case group included individuals with decreased HDL-C concentrations (<40 mg/dL).5 Subjects with pregnancy, liver disease, on medications (beta-blockers, oral contraceptives, steroids, etc), high triglycerides (>200 mg/dL), diabetes mellitus, thyroid, and other endocrinal disorders were excluded. The control group included healthy individuals undergoing routine health checkup, who exhibited HDL-C and other lipid profile fractions within normal limits (ie, HDL-C ≥40 mg/dL, total cholesterol (TC) <200 mg/dL, triglycerides (TGL) <150 mg/dL, low-density lipoprotein cholesterol (LDL-C) <130 mg/dL).5
Sample Size
The sample size was calculated in G*POWER v3.1.9.2. The reported prevalence of obesity/overweight in the urban population of Assam is 16.3% to 23.4%.12 The study was powered at 80% with a 2-sided α = .05. Assuming the prevalence of obesity in the control population as 20% and expecting an odds ratio of 2.5, we found that at least 93 individuals would be needed for each group if cases and controls were to be selected at 1:1 ratio. Thus, we recruited 95 individuals each for the case and the control groups (total sample size = 190).
Biochemical Estimations and Obesity Measurements
Quantification of TC, TGL and HDL-C was done in fasting venous samples using VITROS 5600 Autoanalyser (Ortho-Clinical Diagnostics Inc. USA). LDL-C was estimated using Friedewald’s equation.13 Results were validated by quality control materials (Bio-Rad and Christian Medical College, Vellore, India).
Increased body mass index (BMI) was used to identify overweight or generalized obesity, and increased waist circumference (WC) was used to identify central obesity. Measurements were performed according to World Health Organization (WHO) protocol.14 Increased BMI was defined using WHO-specified Asian-specific cutoff (≥23 kg/m2),15 which was subsequently endorsed for Indians by a consensus statement.16 For increased WC, International Diabetic Federation (IDF)–recommended values for South Asians were used (≥80 cm in females, ≥90 cm in males).17
Statistics
Statistical analysis was done in OpenEpi v3.0.1. Data were compared between the case and control groups using unpaired t test or Fischer’s exact test, as appropriate. Crude association between obesity and decreased HDL-C was estimated by unadjusted odds ratios (ORs) with 95% confidence intervals (CIs). These were complemented by adjusted ORs, calculated using Mantel-Haenszel approach18 after controlling for potential confounders (like gender, smoking, and alcohol use) which could not be eliminated in the design phase. Two-tailed P value of less than .05 was considered significant. The risk estimates, ARP and PARP were calculated as previously described.19,20
Results
The baseline characteristics of the subjects are presented in Table 1. The 2 groups were comparable (P > .05) with respect to age and sex composition, and smoking and alcohol use. The obesity measures were significantly higher (P < .01) in the case group (BMI = 27.52 ± 3.91 kg/m2, WC = 91.28 ± 6.32 cm) than in the control group (BMI = 22.09 ± 3.12 kg/m2, WC = 80.52 ± 5.92 cm).
Table 1.
Baseline Characteristics of the Case and the Control Groups.a
| Variable | Study Subjects (N = 190) |
P | |
|---|---|---|---|
| Case Group (n = 95) | Control Group (n = 95) | ||
| Age (years) | 47.5 ± 11.24 | 45.04 ± 10.28 | .12 |
| Sex | |||
| Male | 63 (66.32) | 59 (62.11) | .65 |
| Female | 32 (33.68) | 36 (37.89) | |
| Smoking | |||
| Yes | 33 (34.74) | 24 (22.11) | .21 |
| No | 62 (65.26) | 71 (77.89) | |
| Alcohol | |||
| Yes | 42 (44.21) | 32 (33.68) | .18 |
| No | 53 (55.79) | 63 (66.32) | |
| BMI (kg/m2) | 27.52 ± 3.91 | 22.09 ± 3.12 | <.01 |
| WC (cm) | 91.28 ± 6.32 | 80.52 ± 5.92 | <.01 |
| HDL-C (mg/dL) | 33.16 ± 4.14 | 51.24 ± 7.16 | <.01 |
| TC (mg/dL) | 186.18 ± 35.74 | 179.83 ± 15.23 | .11 |
| TGL (mg/dL) | 130.81 ± 20.58 | 127.45 ± 11.05 | .16 |
| LDL-C (mg/dL) | 127.13 ± 28.8 | 103.1 ± 13.05 | <.01 |
Abbreviations: BMI, body mass index; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TGL, triglycerides; WC, waist circumference.
Values are expressed as mean ± SD or n (%).
Among cases, 68 individuals (71.58%) were overweight or generally obese, and 47 individuals (49.47%) were centrally obese; while among controls, the respective numbers were 32 (33.68%) and 20 (21.05%) (Table 2). Decreased HDL-C was strongly associated with obesity. Both increased BMI (unadjusted OR = 4.96, 95% CI = 2.68-9.18) and increased WC (unadjusted OR = 3.67, 95% CI = 1.94-6.97) augmented the risk for decreased HDL-C. Even after adjusting for confounders, increased BMI (adjusted OR = 4.90, 95% CI = 3.59-6.68) and increased WC (adjusted OR = 3.33, 95% CI = 2.39-4.64) continued to be strong determinants for decreased HDL-C.
Table 2.
Association of Decreased HDL-C Levels With Obesity and Risk Estimates of Decreased HDL-C Attributable to Obesity.
| Exposure (Risk Factor) | Case Group, n (%) | Control Group, n (%) | Unadjusted OR (95% CI) | Adjusted OR (95% CI) | ARP (95% CI) | PARP (95% CI) |
|---|---|---|---|---|---|---|
| Overweight/general obesity | ||||||
| Increased BMI | 68 (71.58) | 32 (33.68) | 4.96 (2.68-9.18) | 4.90 (3.59-6.68) | 0.798 (0.726-0.852) | 0.571 (0.497-0.646) |
| Normal BMIa | 27 (28.42) | 63 (66.32) | Reference | Reference | — | — |
| Central obesity | ||||||
| Increased WC | 47 (49.47) | 20 (21.05) | 3.67 (1.94-6.94) | 3.33 (2.39-4.64) | 0.728 (0.485-0.856) | 0.36 (0.216-0.504) |
| Normal WCa | 48 (50.53) | 75 (78.95) | Reference | Reference | — | — |
The ARP and PARP values were also considerable (Table 2). The ARP associated with decreased HDL-C in individuals with increased BMI was 0.798 (95% CI = 0.726-0.852). Whereas, in the overall population, the PARP on comparing individuals having increased BMI and those with normal BMI was 0.571 (95% CI = 0.497 0.646). Similarly, the ARP associated with decreased HDL-C in individuals with increased WC was 0.728 (95% CI = 0.485-0.856). The PARP on comparing individuals having increased WC and those with normal WC was 0.36 (95% CI = 0.216-0.504).
Discussion
A large proportion of decreased HDL-C cases in our study were obese. Overweight or generally obese individuals were at nearly 5 times, and centrally obese individuals were at nearly 4 times greater odds for developing decreased HDL-C than their nonobese counterparts.
A large proportion of the disease burden of low HDL-C dyslipidemia could be attributed to obesity. To our knowledge, no study has previously described the magnitude to which decreased HDL-C is attributable to obesity. We found that 79.8% of decreased HDL-C was attributable to increased BMI in the individuals with overweight or generalized obesity, while 72.8% of decreased HDL-C among the centrally obese individuals was due to increased WC. The burden of decreased HDL-C owing to obesity in the overall population was also quite high: 57.1% of low HDL-C was attributable to overweight or generalized obesity, whereas, 36% of low HDL-C was attributable to central obesity.
Our findings have considerable public health relevance because they suggest that it will be possible to reduce the burden of decreased HDL-C substantially if obesity can be combated. Obesity is a modifiable risk factor. Clinical studies have reported that weight reduction in obese individuals brought about appreciable increments in the HDL-C levels.21-23 Therefore, in view of our findings, the prospective benefits of eliminating or modifying obesity in terms of elevated HDL-C would be enormous. Guwahati (located at 26°18′ N latitude and 91°73′ E longitude) is the capital city of Assam and the most populated one in northeast India. Its total population was 962 334 individuals (498 450 males and 463 884 females) as per the Census of India 2011 report.24 Guwahati is undergoing fast transition and modernization. It is likely that this rapid urbanization would contribute to obesity even further.25 In fact, the National Family Health Survey (NFHS-3) indicated that the prevalence of obesity/overweight in the urban population of Assam (16.3% to 23.4%) was almost 4 times than that in the rural population (4.5% to 6%).12 Thus, it is necessary that efforts to challenge obesity be initiated as a major strategy to reduce the burden of decreased HDL-C and improve cardiovascular health in our population.
Our study had some limitations. Since the study was conducted in a single center, we cannot rule out selection bias and coverage bias. We conducted the study in GMCH—a premier hospital that caters to all sections of people from Guwahati. Though it was desirable to conduct a population-based study, yet it was operationally and monetarily not feasible for carrying out random lipid testing in the community to identify appropriate subjects who satisfied the selection criteria. Although GMCH is a tertiary care center, people attend GMCH for routine health checkup and also for primary and secondary health services as it is the chief source of quality and subsidized health care in Guwahati. Because of these aspects, the beneficiaries are not a highly selected group of people with the most severe ailments, but we think that they represent a cross section from the general population at large. Secondly, our study was conducted in one city. Therefore, the internal validity is high. But, the findings need to be confirmed in other cities and populations.
Conclusion
Obesity was a strong determinant for decreased HDL-C levels. Decreased HDL-C levels could be largely attributed to obesity. This has important connotations because interventions targeting a modifiable exposure like obesity have the potential of tremendously improving HDL-C levels and promoting cardiovascular health both in the exposed and the general population.
Acknowledgments
The authors thank Dr Chandana Bhattacharjee, PhD (anthropology), Women’s College, Shillong, Meghalaya and Dr Madhur Borah, MD (community medicine), Jorhat Medical College and Hospital, Jorhat, Assam for their valuable inputs.
Author Biographies
Kaustubh Bora is a medical doctor (MBBS) with specialization in biochemistry (MD). His research interests include laboratory medicine, genetic epidemiology, community health and biomedical informatics, particularly with respect to chronic disorders. He is presently working as a Senior Resident Doctor in the Department of Biochemistry, North Eastern Indira Gandhi Regional Institute of Health and Medical Sciences, Shillong, India.
Mauchumi Saikia Pathak, MBBS, MD, is a professor of Biochemistry and Principal Investigator, Department of Biotechnology (DBT) project in Gauhati Medical College & Hospital, Guwahati, Assam. She has been supervising and coordinating research aimed at understanding genetic and environmental factors influencing haemoglobinopathies and dyslipidemias in the state of Assam.
Probodh Borah is a PhD. in Veterinary Microbiology and presently working as Professor & Head of the Department of Animal Biotechnology at Assam Agricultural University, Guwahati. As the coordinator of the Bioinformatics Centre and State Biotech Hub of the state of Assam, India, he has been mentoring and supervising multidisciplinary research on biomedical sciences.
Dulmoni Das is a board certified nurse, currently working as a Clinical Instructor at Army Institute of Nursing, Guwahati. She has specialized in mental health nursing (MSc) and is further pursuing a master’s in psychology. Her research interests are chronic physical ailments and their relationship with mental health.
Footnotes
Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
References
- 1. Rader DJ, Hobbs HH. Disorders of lipoprotein metabolism. In: Longo DL, Fauci AS, Kasper DK, Hauser SL, Jameson JL, Loscalzo J, eds. Harrison’s Principles of Internal Medicine. 18th ed New York, NY: McGraw-Hill; 2012:3145-3161. [Google Scholar]
- 2. Enas EA, Yusuf S, Mehta JL. Prevalence of coronary artery disease in Asian Indians. Am J Cardiol. 1992;70:945-949. [DOI] [PubMed] [Google Scholar]
- 3. Anand SS, Yusuf S, Vuksan V, et al. Differences in risk factors, atherosclerosis, and cardiovascular disease between ethnic groups in Canada: the Study of Health Assessment and Risk in Ethnic groups (SHARE). Lancet. 2000;356:279-284. [DOI] [PubMed] [Google Scholar]
- 4. Reddy KS, Shah B, Varghese C, Ramadoss A. Responding to the threat of chronic diseases in India. Lancet. 2005;366:1744-1749. [DOI] [PubMed] [Google Scholar]
- 5. National Cholesterol Education Program. Third and final 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). Circulation. 2002;106:3143-3421. [PubMed] [Google Scholar]
- 6. Petrenya N, Brustad M, Dobrodeeva L, et al. Obesity and obesity-associated cardiometabolic risk factors in indigenous Nenets women from the rural Nenets Autonomous Area and Russian women from Arkhangelsk city. Int J Circumpolar Health. 2014;73:23859. doi: 10.3402/ijch.v73.23859. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Bora K, Pathak MS, Borah P, Das D. Variation in lipid profile across different patterns of obesity—observations from Guwahati, Assam. J Clin Diagn Res. 2015;9:OC17-OC21. doi: 10.7860/JCDR/2015/15334.6787. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Rocha FL, de Menezes TN, de Melo RLM, Pedraza DF. Correlation between indicators of abdominal obesity and serum lipids in the elderly. Rev Assoc Med Bras. 2013;59:48-55. [PubMed] [Google Scholar]
- 9. Lichtash CT, Cui J, Guo X, et al. Body adiposity index versus body mass index and other anthropometric traits as correlates of cardiometabolic risk factors. PLoS One. 2013;8:e65954. doi: 10.1371/journal.pone.0065954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Bonita R, Beaglehole R, Kjellström T. Measuring health and disease. In: Basic Epidemiology. 2nd ed Beijing, China: World Health Organization; 2006:15-38. [Google Scholar]
- 11. Tyler CW, Jr, Last JM. Epidemiology. In: Wallace RB, ed. Maxcy-Rosenau-Last: Public Health and Preventive Medicine. 14th ed New York, NY: McGraw-Hill; 1998:5-34. [Google Scholar]
- 12. International Institute for Population Sciences and Macro International. NFHS-3 fact sheets for key indicators based on final data: Assam. In: National Family Health Survey (NFHS-3), India, 2005-06: Assam. Mumbai, India: International Institute for Population Sciences; 2008. http://rchiips.org/nfhs/assam_report_for_website.pdf. Accessed on July 15, 2016. [Google Scholar]
- 13. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the low-density lipoprotein cholesterol concentration in plasma without the use of preparative centrifuge. Clin Chem. 1972;18:499-502. [PubMed] [Google Scholar]
- 14. World Health Organization. Annex 2: Recommended measurement protocols and derivation of indices. In: World Health Organization Technical Report Series 854—Physical Status: The Use and Interpretation of Anthropometry. Geneva, Switzerland: World Health Organization; 1995:424-438. [Google Scholar]
- 15. World Health Organization, the International Association for the Study of Obesity, and the International Obesity Task Force. Assessment/diagnosis. In: The Asia-Pacific Perspective: Redefining Obesity and Its Treatment. Manila, Philippines: WHO Regional Office for the West Pacific; 2000:15-21. [Google Scholar]
- 16. Misra A, Chowbey P, Makkar BM, et al. Consensus statement for diagnosis of obesity, abdominal obesity and the metabolic syndrome for Asian Indians and recommendations for physical activity, medical and surgical management. J Assoc Physicians India. 2009;57:163-170. [PubMed] [Google Scholar]
- 17. International Diabetic Federation. The IDF Consensus Worldwide Definition of the Metabolic Syndrome. Brussels, Belgium: International Diabetic Federation; 2006. [Google Scholar]
- 18. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst. 1959;22:719-748. [PubMed] [Google Scholar]
- 19. Cole P, MacMahon B. Attributable risk percent in case-control studies. Br J Prev Soc Med. 1971;25:242-244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Taylor JW. Simple estimation of population attributable risk from case-control studies. Am J Epidemiol. 1977;106:260. [DOI] [PubMed] [Google Scholar]
- 21. Streja DA, Boyko E, Rabkin SW. Changes in plasma high-density lipoprotein cholesterol concentration after weight reduction in grossly obese subjects. Br Med J. 1980;281:770-772. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Mandai N, Akazawa K, Hara N, Ide Y, Dazai U, Chishaki A, Chisaki H. Body weight reduction results in favourable changes in blood pressure, serum lipids, and blood sugar in middle-aged Japanese persons: a 5-year interval observational study of 26,824 cases. Global J Health Sci. 2015;7:159-170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Finelli C, Crispino P, Gioia S, et al. The improvement of high-density lipoprotein (HDL) particle levels, and presumably HDL metabolism, depend on effects of low-carbohydrate diet and weight loss. EXCLI J. 2016;15:166-176. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
- 24. Directorate of Census Operations Assam. Section 1: Primary census abstract (PCA). In: District Census Handbook Kamrup Metropolitan. Village and Townise Primary Census Abstract. Census of India 2011. http://www.censusindia.gov.in/2011census/dchb/1823_PART_B_DCHB_KAMRUP%20METROPOLITAN.pdf. Accessed July 15, 2016.
- 25. Goryakin Y, Suhrcke M. Economic development, urbanization, technological change and overweight: what do we learn from 244 demographic and health surveys? Econ Hum Biol. 2014;14:109-127. [DOI] [PMC free article] [PubMed] [Google Scholar]