The Impact of Aerobic Exercise and Badminton on HDL Cholesterol Levels in Adult Taiwanese

Abstract

Elevated levels of high-density lipoprotein cholesterol (HDL-C) have been associated with a decreased risk of coronary heart disease (CHD). An active lifestyle is necessary in order to improve lipid HDL-C, including (but not limited to) physical exercise. Research on the association between badminton, an intermittent exercise, and HDL-C is limited. We investigated the impact of aerobic exercise and badminton on HDL-C levels in Taiwanese adults. The sociodemographic data of 7797 participants comprising 3559 men and 4238 women aged between 30 to 70 years were retrieved from the Taiwan Biobank. The participants were grouped into three exercise categories—no exercise, aerobic exercise, and badminton exercise. The HDL-C levels were compared using an analysis of variance (ANOVA). The multivariate linear regression models were used to determine the associations between HDL and exercise. Comparing the other two groups to the no-exercise group, the individuals who were engaged in aerobic and badminton exercise were significantly associated with a higher HDL-C (β =1.3154; p <0.0001 and β = 6.5954; p = 0.0027, respectively). Aerobic exercise and badminton were also associated with higher HDL-C levels among carriers of the lipoprotein lipase (LPL) rs328 genotypes. Aerobic exercise and regular badminton were associated with higher levels of HDL-C, with the badminton group being more significant.

1. Introduction

Substantial epidemiologic evidence suggests a negative linear correlation between of high-density lipoprotein cholesterol (HDL-C) levels and the incidence of coronary heart disease (CHD); an inverse relationship between HDL-C and cardiovascular disease was not well established until the Framingham study in the 1970s, which identified HDL-C as a powerful risk factor inversely associated with the incidence of CHD [1]. High-density lipoprotein (HDL) is positively associated with a decreased risk of coronary heart disease (CHD). As defined by the United States National Cholesterol Education Program Adult Treatment Panel III guidelines, an HDL-C level of 60 mg/dL or greater is a negative (protective) risk factor. On the other hand, a high-risk HDL-C level is described as one that is less than 40 mg/dL.

The major apolipoproteins of HDL are apolipoprotein (apo) A-I and apo A-II, the alpha-lipoproteins. Elevated concentrations of apo A-I and apo A-II are called hyperalphalipoproteinemia (HALP), which is associated with lower risk CHD. Conversely, hypoalphalipoproteinemia increases the risk of CHD. The levels at which HDL-C confers benefit or risk are not discrete, and the cut points are somewhat arbitrary, especially considering that HDL-C levels are, on average, higher in United States women compared with men [2,3]. Hyperalphalipoproteinemia (HALP) is caused by a variety of genetic and environmental factors. Among these, plasma cholesteryl ester transfer protein (CETP) deficiency is the most important and frequent cause of HALP in Asian populations. CETP facilitates the transfer of cholesteryl ester (CE) from a high-density lipoprotein (HDL) to apolipoprotein (apo) B-containing lipoproteins, and is a key protein in the reverse cholesterol transport system [4].

However, environmental factors also have a significant impact on HDL-C. Smoking and obesity are the most significant risk factors associated with a lower HDL-C [5]. Besides these factors, genetic variants also have an impact on HDL-C. Certain genes play an essential role in the synthesis and metabolism of serum lipids. One of these genes is the lipoprotein lipase (LPL) gene, whose variant (Rs 328) has been associated with HDL-C and triglyceride [6,7]. LPL rs328 GG and CG genotypes were found to be significantly related to a higher HDL-C and triglyceride [8,9].

Randomized controlled clinical trials have demonstrated that interventions to raise HDL-C levels are associated with reduced CHD events. Exercise is one of the lifestyle integrations that have been recommended for improving lipid fractions such as HDL cholesterol [10]. Several studies have shown that aerobic exercise is associated with higher HDL-C. Among them is Dr. Satoru Kodama (Ochanomizu University, Tokyo, Japan) and colleagues, who showed that aerobic training resulted in a 2.53-mg/dL increase in HDL-C levels, so, by rough estimates, it could result in a 5.1% and 7.6% reduction in cardiovascular disease risk in men and women, respectively [11,12,13,14,15]. The most important element of an exercise program is the duration per session [11,14]. Aerobic exercise has also been associated with a better prognosis of cardiovascular disease [16]. Based on a previous study, intermittent exercise programs were associated with significantly improvements in lipid profiles following eight weeks of training in obese children [17].

The effects of exercise behavior on the predicted CVD risks were found to vary depending on different factors [18]. Badminton, an indoor intermittent exercise most popular in Asia, has been shown to improve the maximum power output of regular practitioners, so it should be considered as a strategy for improving the health and well-being of untrained females who are currently not meeting the physical activity guidelines [19]. Outdoor exercises have been linked to air pollution and associated health issues. The respiratory physiology of exercise suggests that athletes and other exercisers may experience magnified exposure to ambient air pollution in outdoor exercises, hence should avoid exercising by the road side, as ozone (O₃) is particularly damaging to athletes [20]. As badminton is an indoor sport, playing it might reduce the harmful health effects associated with air pollution. For instance, in sedentary United Kingdom females, badminton significantly lowered some cardiovascular health markers, including the mean arterial pressure, systolic and diastolic blood pressure, and resting heart rate [19]. The findings from another study revealed that playing badminton can reduce all-cause mortality by 47% and CVD mortality risk by 59% [21].

Both aerobic exercise and badminton have positive effects on health. Several investigations have been made regarding HDL-C and aerobic exercise [16]. The results show that HDL-C levels compared to other lipid fractions are more sensitive to aerobic exercise. As far as research on HDL-C and exercise is concerned, hardly any has been done with regards to badminton exercise [3,19]. Because of this, we investigated the association between badminton, aerobic exercise, and HDL-C among adult Taiwanese.

2. Methods

2.1. Data Source

The data were obtained from the Taiwan Biobank, a national health resource. The Biobank contains the genetic information of over 200,000 ethnic Taiwanese residents aged 30 to 70 years [22]. Presently, there are 29 recruitment centers, with each city or county having at least one. The recruitment methods in the Taiwan Biobank are in accordance with the relevant guidelines and regulations. Written informed consent is obtained from all of the participants prior to data collection. The data are collected through questionnaires as well as physical and biochemical examinations. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Institutional Review Board of Chung Shan Medical University.

2.2. Study Participants

Overall, 7797 individuals consisting of 3559 men and 4238 women aged 30–70 years were recruited. Their demographic (age, sex, body mass index (BMI), waist-hip ratio (WHR), and body fat), biochemical (high-density lipoprotein cholesterol (HDL-C]), and lifestyle (physical activity, coffee drinking, smoking, alcohol consumption, and betel nut chewing) data were retrieved from the database. The participants were categorized based on exercise status—no exercise (did not exercise at all during the last three months), aerobic exercise (three different types of any regular exercise (excluding badminton) three times a week for at least 30 min each session), and badminton (only regular badminton in the last three months).

2.3. Statistical Analysis

The data were managed and analyzed using the SAS 9.4 software (SAS Institute, Cary, NC). One-way analysis of variance (ANOVA) was used to compare the HDL-C levels in the various exercise groups. Multivariate linear regression models were used to determine the association between HDL-c and exercise. The data were presented as mean ± standard error (SE) for the continuous variables.

3. Results

Table 1. Basic characteristics of participants stratified by exercise type.

Variable	No Exercise	Aerobic Exercise	Badminton	p-Value
	(n = 8345)	(n = 4111)	(n = 49)
	Mean ± SE	Mean ± SE	Mean ± SE
rs328
CC	52.41 ± 0.17	53.93 ± 0.24	52.55 ± 2.65	<0.0001
CG + GG	54.79 ± 0.35	56.41 ± 0.52	63.82 ± 2.79	0.0036
CG	54.67 ± 0.35	56.39 ± 0.53	63.82 ± 2.79
GG	57.29 ± 1.78	56.86 ± 2.49	-
Sex
Female	57.61 ± 0.20	58.93 ± 0.31	67.91 ± 5.98	<0.0001
Male	47.14 ± 0.18	49.45 ± 0.27	51.18 ± 1.71	<0.0001
Waist-hip ratio
Male <0.9; female <0.8	54.54 ± 0.25	56.01 ± 0.39	55.09 ± 2.97	0.0056
Male ≥0.9; female ≥0.8	51.87 ± 0.19	53.53 ± 0.27	55.64±3.38	<0.0001
Body fat (%)
Male <25; female <30	55.07 ± 0.22	55.72 ± 0.31	57.41 ± 3.17	0.1567
Male ≥25; female ≥30	50.53 ± 0.20	52.70 ± 0.31	52.12 ± 2.70	<0.0001
Coffee drinking
Yes	54.47 ± 0.28	55.16 ± 0.40	55.29 ± 4.41	0.3484
No	52.11 ± 0.18	54.05 ± 0.26	55.40 ± 2.58	<0.0001
Age
30–40	53.43 ± 0.25	54.68 ± 0.61	55.00 ± 3.75	0.1440
41–50	52.84 ± 0.27	54.79 ± 0.46	60.33 ± 4.85	0.0002
51–60	52.53 ± 0.31	54.11 ± 0.35	48.67 ± 2.92	0.0022
61–70	51.71 ± 0.46	54.33 ± 0.44	56.80 ± 5.50	0.0002
BMI
Normal	57.34 ± 0.22	58.28 ± 0.32	59.79 ± 3.99	0.0416
Underweight	64.59 ± 0.91	67.73 ± 1.97	-	0.1162
Overweight	49.41 ± 0.24	51.91 ± 0.36	54.50 ± 3.02	<0.0001
Obese	46.22 ± 0.25	46.85 ± 0.38	48.82 ± 3.53	0.2906
Smoking
Never	54.56 ± 0.17	55.79 ± 0.25	57.44 ± 2.68	0.0001
Quit	48.38 ± 0.41	49.84 ± 0.52	51.71 ± 3.02	0.0726
Current	46.39 ± 0.37	48.14 ± 0.73	40.33 ± 1.20	0.0541
Drinking
Never	53.21 ± 0.16	54.89 ± 0.23	56.87 ± 2.48	<0.0001
Quit	44.31 ± 0.79	46.90 ± 0.97	-	0.0366
Current	51.26 ± 0.56	51.97 ± 0.78	45.83 ± 1.64	0.4426
Vegetarian
No	52.97 ± 0.16	54.73 ± 0.23	55.44 ± 2.50	<0.0001
Former	53.92 ± 0.76	52.41 ± 1.08	54.67 ± 2.40	0.5507
Current	50.00 ± 0.59	49.32 ± 0.86	55.00 ± 3.00	0.6621

shows the baseline characteristics of the participants in different exercise groups. The participants comprised 8345 (no exercise), 4111 (aerobic exercise), and 49 (badminton). The mean (± SE) HDL-C in various exercise groups for no exercise, aerobic exercise, and badminton were 47.14 ± 0.18 (men) and 57.61 ± 0.20 (women), 49.45 ± 0.27 (men) and 58.93 ± 0.31 (women), and 51.18 ± 1.17 (men) and 67.91 ± 5.98 (women), respectively. Among both the men and women, there were significant differences in HDL-C between the exercise groups.

Table 2. Association of high-density lipoprotein (HDL) with aerobic exercise (model 1) and badminton (model 2).

Variables	Model 1			Model 2
	β	SE	p-Value	β	SE	p-Value
Exercise (Ref: no exercise)
Aerobic exercise	1.2788	0.2384	<0.0001	-	-	-
Badminton	-	-	-	4.7663	1.6698	0.0043
rs328 (Ref: CC)
CG + GG	2.5070	0.2698	<0.0001	2.4938	0.3221	<0.0001
Sex (Ref: female)
Male	−9.9469	0.2815	<0.0001	−9.9028	0.3356	<0.0001
Waist-hip ratio (Ref: Male<0.9; female<0.8)
Male ≥0.9; Female ≥0.8	−3.2221	0.2541	<0.0001	−2.8504	0.3032	<0.0001
Body fat rate (Ref: Male <25; female <30)
Male ≥25; female ≥30	−2.0892	0.2786	<0.0001	−2.2202	0.3347	<0.0001
Coffee (Ref: No)
Yes	1.3160	0.2299	<0.0001	1.3593	0.2752	<0.0001
Age (Ref: 30–40)
41–50	0.7584	0.2861	0.0080	0.8970	0.3144	0.0043
51–60	1.0165	0.3004	0.0007	1.1284	0.3494	0.0012
61–70	1.0464	0.3644	0.0041	0.7587	0.4697	0.1062
BMI (Ref: Normal)
Underweight	5.5083	0.6821	<0.0001	4.9897	0.7633	<0.0001
Overweight	−4.2453	0.2803	<0.0001	−4.4967	0.3416	<0.0001
Obese	−6.5750	0.3600	<0.0001	−6.4469	0.4288	<0.0001
Smoking (Ref: Never)
Former	−0.0113	0.3731	0.9758	−0.1053	0.4615	0.8195
Current	−3.1084	0.3817	<0.0001	−3.1845	0.4343	<0.0001
Drinking (Ref: Never)
Former	−1.2749	0.6757	0.0592	−1.0467	0.8938	0.2416
Current	4.1397	0.4367	<0.0001	4.3372	0.5098	<0.0001
Vegetarian (Ref: No)
Former	−0.1944	0.5036	0.6996	0.4353	0.5798	0.4528
Yes	−5.7091	0.5016	<0.0001	−5.0560	0.5868	<0.0001

SE = standard error; β = beta value; ref. = reference; model 1 = association between HDL-cholesterol (C) and aerobic exercise; model 2 = association between HDL-C and badminton.

shows two separate models demonstrating the association of HDL-C with aerobic exercise (model 1) and badminton exercise (model 2). After adjusting for confounders, HDL-C was positively associated with aerobic exercise (β = 1.2788, p ≤ 0.0001) and badminton (β = 4.7663, p = 0.0043) when compared to no exercise.

Table 3. Association between HDL-C and exercise type (aerobic exercise and badminton are included in the same model).

	β	SE	p-Value
Exercise (Ref: no exercise)
Aerobic exercise	1.2839	0.2385	<0.0001
Badminton	4.7697	1.7072	0.0052
rs328 (Ref: CC)
CG + GG	2.5368	0.2693	<0.0001
Sex (Ref: Female)
Male	−9.9687	0.2811	<0.0001
WHR (Ref: male<0.9; female<0.8)
Male ≥0.9; Female ≥0.8	−3.2070	0.2538	<0.0001
Body fat rate (Ref: Male <25; female <30)
Male ≥25; female ≥30	−2.1104	0.2783	<0.0001
Coffee (Ref: No)
Yes	1.3058	0.2295	<0.0001
Age (Ref:30–40)
41–50	0.7659	0.2855	0.0073
51–60	0.9964	0.2999	0.0009
61–70	1.0307	0.3638	0.0046
BMI (Ref: Normal)
Underweight	5.4986	0.6823	<0.0001
Overweight	−4.2299	0.2799	<0.0001
Obese	−6.5719	0.3594	<0.0001
Smoking (Ref: Never)
Former	−0.0184	0.3721	0.9607
Current	−3.0920	0.3810	<0.0001
Drinking (Ref: Never)
Former	−1.2715	0.6759	0.0600
Current	4.0625	0.4351	<0.0001
Vegetarian (Ref: No)
Former	−0.2088	0.5024	0.6778
Current	−5.7038	0.5009	<0.0001

shows the association between HDL-C and exercise, with both aerobic exercise and badminton included in the same model. There were positive associations of HDL-C with aerobic exercise and badminton (β = 1.2839, p < 0.0001 and β = 4.7697, p = 0.0052, respectively). The Rs328 CG/GG genotype was associated with increased levels of HDL-C (β = 2.5070, p <0.0001) when aerobic exercise and badminton were both included in the model. HDL-C was negatively associated with male sex (β = −9.9687, p < 0.0001), WHR (β = −3.2070, p < 0.0001), body fat (β = −2.1104, p < 0.0001), overweight (β = −4.2299, p < 0.0001), obesity (β = −6.5719, p < 0.0001), current smoking (β = −3.0920, p < 0.0001), and current vegetarian (β = −5.7038, p < 0.0001). However, it was positively associated with coffee drinking, age, being underweight, and current drinking.

Compared with no exercise, aerobic exercise, and badminton—associated with higher HDL-C levels among carriers of rs328 genotypes (

Table 4. Association between HDL-C and exercise type (aerobic exercise and badminton are included in the same model) stratified by rs328 polymorphism.

	CC			CG + GG
	β	SE	p-value	β	SE	p-value
Exercise (Ref: no exercise)
Aerobic exercise	1.2294	0.2642	<0.0001	1.5777	0.5552	0.0045
Badminton	3.2045	1.9680	0.1035	9.3738	3.4357	0.0064
p for trend	<0.0001			0.0006
Sex (Ref: Female)
Male	−9.8037	0.3121	<0.0001	−10.8237	0.6480	<0.0001
WHR (Ref: male <0.9; female <0.8)
Male ≥ 0.9; Female ≥ 0.8	−3.2158	0.2808	<0.0001	−3.2330	0.5926	<0.0001
Body fat rate (Ref: male <25; female <30)
Male ≥25; female ≥30	−1.9110	0.3094	<0.0001	−2.9933	0.6370	<0.0001
Coffee (Ref: No)
Yes	1.4443	0.2563	<0.0001	0.9175	0.5177	0.0765
Age (Ref:30–40)
41–50	1.1310	0.3167	0.0004	−0.8803	0.6606	0.1828
51–60	1.0243	0.3330	0.0021	0.8815	0.6905	0.2019
61–70	1.5765	0.4054	0.0001	−1.2202	0.8259	0.1397
BMI (Ref: Normal)
Underweight	6.0939	0.7653	<0.0001	3.0483	1.5028	0.0426
Overweight	−4.1994	0.3098	<0.0001	−4.3507	0.6516	<0.0001
Obese	−6.5343	0.4007	<0.0001	−6.6934	0.8135	<0.0001
Smoking (Ref: Never)
Former	−0.0216	0.4124	0.9582	0.0822	0.8647	0.9243
Current	−3.0946	0.4245	<0.0001	−3.0652	0.8632	0.0004
Drinking (Ref: Never)
Former	−1.6490	0.7507	0.0281	0.5759	1.5541	0.7110
Current	4.0044	0.4859	<0.0001	4.4941	0.9772	<0.0001
Vegetarian (Ref: No)
Former	0.6089	0.5602	0.2771	−3.5493	1.1375	0.0018
Current	−5.8054	0.5563	<0.0001	−5.0515	1.1531	<0.0001

). The β-values were higher for badminton than for aerobic exercise (i.e., 3.2045 vs. 1.2294 for the CC genotype, and 9.3738 vs. 1.5777 for the CG + GG genotype) and the tests for trend were statistically significant (p < 0.05).

4. Discussion

To our knowledge, this is the first Asian study that has investigated the effect of badminton on HDL-C. There were significant associations of the HDL-C level with aerobic exercise and badminton. Several studies have been carried out on the association between HDL-C and exercise only. However, the results have not been consistent. While some demonstrated that regular exercise could significantly raise the serum levels of HDL-C [13,14,23], others showed no significant changes [24,25,26,27]. In our study, the badminton effect on HDL-C was stronger than the aerobic exercise. The mechanisms underlying these associations are still unclear. However, these effects might be linked to a higher expression of liver ATP-binding cassette transporters A-1(ABCA1) [28], caused by the upregulation of the liver X receptor (LXR) [16]. This improves the reverse cholesterol transport (RCT) process, hence, resulting in more cholesterol being transported to the liver via HDL.

Several studies have reported that different exercise types can influence cholesterol and may change personal health status [14,15,23,26,29]. As far as the association of badminton with HDL-C is concerned, more research has not yet been done. A study conducted in the United Kingdom showed significant associations between badminton and cardiovascular health markers [19]. Nonetheless, how different exercise types influence the risk of cardiovascular diseases is yet to be fully understood. Similar to our results, Sasaki and colleagues found that long term aerobic exercise was associated with an increase in HDL-C and weight reduction in obese children [30]. Research focused on Taiwanese adults also discovered that regular weekly exercise durations of <2.5 and ≥2.5 h were both positively associated with HDL-C in both sexes. However, the associations were stronger in males than females [11].

It is worth mentioning that significant associations were found between exercise and HDL-C based on age, gender, obesity, blood pressure, blood cholesterol, smoking, and alcohol drinking, some of which have served as risk factors for cardiovascular diseases [31,32,33,34]. Compared to several previous studies, this study had a larger sample size. To better understand the relationship between exercise and HDL-C, participants were grouped into different exercise categories—no exercise (did not do exercise at all during the last three months), aerobic exercise (three different types of any exercise at least, regularly performed three times a week for at least 30 min each session, where badminton was not one of these three exercises), and badminton (only practicing badminton regularly in the last three months, and not any other type of sports). So far, such stratifications had not been made in studies conducted in Asia, particularly in Taiwan. The addition of LPL Rs328 in the model did not modulate the association between HDL-C and exercise modality. However, we found a significant association between its genotypes and HDL-C levels. For instance, rs328CG/GG polymorphism influenced the effect of current alcohol drinking and smoking on HDL-C levels. That is, among the rs328CG/GG carriers, current alcohol drinking was associated with increased HDL-C levels (β = 4.4941, p < 0.0001), while current smoking was linked to lower HDL-C (β = −3.0652, p = 0.0004). The possible mechanisms underlying these associations remain to be clarified. The effects of age and coffee drinking on HDL-C of rs328CG/GG carriers were not significant. The study is limited in that the sample size for the badminton players was small. This is because these players included those that were restricted only to badminton and nothing else in the last three months.

In conclusion, aerobic exercise and regular badminton were associated with higher levels of HDL-C, with the badminton group being more significant. Further investigations with large-scale sample sizes are recommended in order to make a stronger and definite conclusion regarding the link between HDL-C and badminton in particular.