After obtaining informed written consent, subjects diagnosed with hyperlipidemia and on statin medication were recruited from the Klinik Rawatan Keluarga at Hospital Universiti Sains Malaysia (HUSM) between May 2018 and October 2020. The recruited subjects included individuals aged between 18 and 75 who had been diagnosed with hyperlipidemia and prescribed statins for at least six weeks. The patients’ medical records were obtained from their clinic files and the hospital’s database. The exclusion criteria are as follows: patients who have been (i) prescribed other lipid-lowering medications as concurrent treatment; (ii) diagnosed with malignancy, liver disease, thyroid disease, kidney disease, or familial hypercholesterolemia; (iii) receiving other medications known to interfere with statin’s efficacy.

The DNA from the whole blood (3 mL) was extracted according to the manufacturer’s protocols (GeneAll Biotechnology, Korea). The extracted DNA was then subjected to genotyping using amplification-refractory mutation system-polymerase chain reaction (ARMS-PCR). The temperatures during the ARMS-PCR cycles were as follows: initial denaturation at 95 °C for 5 minutes, 30 denaturation cycles at 94 °C for 30 seconds, annealing at 60.9 °C for 30 seconds, extension at 72 °C for 30 seconds, and an additional extension at 72 °C for 5 minutes. Table 1 lists the primer sets used in this study, which were designed using a website (http://primer1.soton.ac.uk/primer1.html); see Appendix 1 for a supplementary file highlighting the primer sequences within the gene sequences. The products of the ARMS-PCR were visualized on the 2% agarose gel for 60 minutes.

The data analysis was carried out using the SPSS software version 26.0 (IBM, United States). The data were reported as frequency (%) and mean ± standard error of the mean (SEM) for categorical and continuous data, respectively. The normality of the continuous data was checked using a test of normality, the Kolmogorov–Smirnov test. For the categorical data, the Chi-square test was used in a contingency table to obtain the odds ratio (OR); the test required at least 80% of the cells to have an expected count greater than 5. Student’s t-test was used to assess differences between two continuous data; in the case of non-parametric data, the Mann–Whitney test was used instead. Repeated measures ANOVA was used to compare means across one or more variables that are based on related subjects; in the case of non-parametric data, the Friedman test was used. When conducting multiple statistical tests, the Bonferroni correction was used to adjust the probability (P) values to avoid a Type I error. The observed genotype frequency was checked for the Hardy–Weinberg equilibrium (HWE) using https://gene-calc.pl/hardy-weinberg-page. Since the homozygous mutant GG genotypes are rare, a genetic dominant model (AA vs. GA + GG genotypes) was applied to assess the SNP association. Multiple binary logistic regression was used to determine which independent factors were associated with the likelihood of achieving the LDL target of < 2.6 mmol/L. Throughout the analysis, a P-value of 0.05 or lower was considered statistically significant.

Figure 1

Table 1

Primer sequences	Product size	Interpretation of the genotypes
FO: 5'-TGGTGAAAAGGACACCTTCC-3'	Outer; 500 bp	Heterozygous; 500 bp, 420 bp and 129 bp
FI (G allele): 5'-TATTTGGGAGCAGTGTCATGGACGTG-3'	G allele; 420 bp	Homozygous recessive; 500 bp and 420 bp
RI (A allele): 5'-GGCTGATAAGCCTGTCCCTCTGACT -3'	A allele; 129 bp	Wild type; 500 bp and 129 bp
RO: 5'-CTTTTCTGCCGTGGATTCTC- 3'

Abbreviations: Bold indicates mismatches that were introduced in the sequence, while underlined sequences indicate the SNP location. FI: Forward inner primer; RI: Reverse inner primer; FO: Forward outer primer; RO: Reverse outer primer; bp: base pair

Table 2

Characteristics	n = 122
Age (years + SEM)	53.62 ± 6.67
Gender, n (%)
Female	71 (58.2)
Male	51 (41.8)
Race, n (%)
Malay	117 (95.9)
Chinese	3 (2.5)
Indian	1 (0.8)
Others	1 (0.8)
Genotype of rs646776, n (%)
aPresent study
AA	102 (83.6)
GA	20 (16.3)
GG	0
Reference population
AA	461 (91.4)
GA	40 (7.9)
GG	3 (0.5)
Type and dosage of statin, n (%)
Atorvastatin
10 mg/day	33 (27.0)
20 mg/day	32 (26.2)
30 mg /day	4 (3.3)
40 mg /day	8 (6.6)
Simvastatin
10 mg /day	4 (3.3)
20 mg /day	23 (18.9)
40 mg /day	3 (2.5)
Pravastatin
10 mg /day	2 (1.6)
20 mg /day	7 (5.7)
Lovastatin
20 mg /day	6 (4.9)
Comorbidities, n (%)
Hypertension	52 (42.6)
Diabetic + hypertension	44 (36.0)
Diabetic	12 (9.8)
Others	8 (6.6)
None	6 (4.9)
Concomitant drugs, n (%)
Anti-diabetic drugs class
Biguanides only	1 (0.8)
Two> anti-diuretic drugs or more	11 (9.0)
Anti-hypertensive drugs class
Calcium channel blockers only	17 (13.9)
Angiotensin-converting enzyme (ACE) inhibitors only	5 (4.1)
Alpha-blockers only	1 (0.8)
Beta blockers only	1 (0.8)
Two > anti-hypertensive drugs or more	28 (23.0)
Anti-diabetic + Anti-hypertensive	44 (36.1)
Others	8 (6.6)
None	6 (4.9)
Baseline level (mmol/L ± SD)
TC (normal range <5.2 mmol/L)	5.67 ± 1.56
HDL-C (normal range >1.0 in males and >1.3 in women)	1.33 ± 0.63
LDL-C (normal range <2.6 mmol/L)	3.61 ± 1.32
TG (normal range <1.7 mmol/L)	1.60 ± 0.89

Abbreviations: ^aMAF of the SNP was 0.08 and the genotype frequencies were in accordance with Hardy-Weinberg Equilibrium (P value = 0.6149) calculated via https://gene-calc.pl/hardy-weinberg-page, ^bThe frequencies of rs646776 in East Asian population obtained from the Ensemble genome browser. SEM: standard error; ACE: angiotensin-converting enzyme; SD: standard deviation; TC: total cholesterol; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; TG: triglycerides.

Table 3

Lipid profile (mmol/L)	AA (n = 102)	GA (n = 20)	P-value
TC (mean ± SD)
Baseline level	5.66 ± 1.64	5.69 ± 1.03	0.936
0-6 months treatment	4.61 ± 1.75	5.25 ± 0.91	0.111
7-12 months treatment	5.02 ± 1.58	5.11 ± 1.55	0.816
aBaseline vs 0-6 months	<.001	0.451
aBaseline vs 7-12 months	0.004	0.344
HDL-C (mean ± SD)
Baseline level	1.32 ± 0.66	1.43 ± 0.37	0.450
0-6 months treatment	1.17 ± 0.41	1.42 ± 0.34	0.013
7-12 months treatment	1.27 ± 0.35	1.45 ± 0.38	0.043
aBaseline vs 0-6 months	0.158	1.000
aBaseline vs 7-12 months	1.000	1.000
LDL-C (mean ± SD)
Baseline level	3.68 ± 1.33	3.28 ± 1.21	0.205
0-6 months treatment	2.73 ± 1.27	3.11 ± 0.85	0.197
7-12 months treatment	3.04 ± 1.14	3.21 ± 1.02	0.539
aBaseline vs 0-6 months	<.001	1.000
aBaseline vs 7-12 months	<.001	1.000
TG (mean ± SD)
Baseline level	1.58 ± 0.89	1.64 ± 0.91	0.783
0-6 months treatment	1.48 ± 1.08	1.59 ± 0.62	0.663
7-12 months treatment	1.60 ± 1.09	1.38 ± 0.69	0.378
aBaseline vs 0-6 months	0.905	1.000
aBaseline vs 7-12 months	1.000	1.000

Abbreviations: ^aThe comparison within two groups obtained using one-way ANOVA repeated measures with Bonferroni adjustments. TC: total cholesterol; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; TG: triglycerides.

Table 4

Independent variables	P-value	Odds ratio Exp (B)	95% CI
SORT1/CELSR2/PSRC1 (rs646776)	0.709	1.219	0.431 - 3.449
Age	0.235	1.037	0.977 - 1.102
aGender	0.135	0.560	0.262 - 1.199
Type of Statin
Atorvastatin, mg/day
10	0.725	0.000	0.000
20	0.934	1.045	0.365 - 2.998
30	0.112	6.900	0.637 - 74.690
40	0.695	1.380	0.275 - 6.921
Simvastatin, mg/day
10	0.827	0.767	0.071 - 8.299
S20	0.731	0.812	0.247 - 2.670
S40	0.913	1.150	0.093 - 14.188
Pravastatin, mg/day
10	0.999	0.000	0.000
P20	0.522	1.725	0.324 - 9.172
Lovastatin, mg/day
20	0.106	4.600	0.721 - 29.332
Baseline lipid (mmol/L)
Total cholesterol	0.524	0.925	0.729 - 1.175
High-density lipoprotein	0.519	0.744	0.303 - 1.826
Low-density lipoprotein	0.688	0.941	0.699 - 1.267
Triglycerides	0.389	0.820	0.522 - 1.289

Abbreviations: ^aMale category was used as the reference; ^bAtorvastatin 10 mg was used as the reference category. CI: confidence interval; LDL-C: low-density lipoprotein cholesterol.

A total of 122 subjects with hyperlipidemia, mostly of self-reported Malay ethnicity (95.9%), were recruited into the study. The demographic profiles and clinical characteristics of the subjects are presented in Table 2. The mean age of the subjects is 53 ± 6.67 years; most subjects are female (58.2%) and treated with atorvastatin (63.1%). Many subjects had comorbidities such as diabetes and hypertension, while only 4.9% had no pre-existing diseases. Specifically, 42.6% had hypertension, 36.0% had diabetes and hypertension, 9.8% had diabetes, and the remaining 6.6% had other diseases. Because of their pre-existing illnesses, 88.5% of the subjects were treated with other drugs in addition to statins. Before starting statin therapy, the baseline lipid levels for total cholesterol (TC, 5.67 ± 1.56 mmol/L) and LDL-C (3.61 ± 1.32 mmol/L) were higher than the recommended levels for a healthy individual.

The lipid levels of the statin users (Table 3) were evaluated over two periods: 0–6 months and 7–12 months. Compared to common AA genotypes, GA genotypes had substantially higher HDL-C levels after 0–6 months (P = 0.013) and 7–12 months (P = 0.043) of statin treatment. Despite not being observed in the GA genotypes, statin treatment resulted in lower TC and LDL-C levels in the AA genotypes after 0–6 months and 7–12 months, both at P < 0.05.

Additionally, we stratified the HDL-C levels based on the patient’s genotypes into males and females to examine the impact of patient gender on the effects of statins on the HDL-C profile (Figure 1). Females with minor allele G carriers (GA) demonstrated significantly higher HDL-C levels after 0–6 months (P = 0.010) and 7–12 months (P=0.013) of statin treatment. Next, multiple binary logistic regression analysis was carried out to determine the association between the independent variables and patient attainment of the target LDL-C level of < 2.6 mmol/L. Table 4 illustrates that none of the variables tested predicted the subject’s attainment of the target LDL-C level of < 2.6 mmol/L.