Compatibility between an overnight fasting and random cholesterol tests in Asians

Ian Kwong Yun Phoon, Yi Ling Eileen Koh, Xiaoxuan Guo, Usha Sankari, Ngiap Chuan TanScientific 2021 · DOI

TL;DR

This study investigated whether non-fasting cholesterol tests are a reliable substitute for fasting tests in Asian patients with type 2 diabetes and dyslipidemia. The findings suggest that for most lipid markers, non-fasting tests provide comparable results, potentially simplifying clinical management and improving patient convenience.

What Problem Does This Paper Address?

Traditionally, assessing lipid profiles for cardiovascular risk management, particularly for cholesterol levels like LDL-C, has required patients to fast for 8-12 hours. This fasting requirement presents several practical challenges: it complicates scheduling phlebotomy services, can lead to patient discomfort, and may even cause adverse effects like hypoglycemia, especially for individuals with diabetes. Recent Western guidelines have begun to question the necessity of fasting for routine lipid monitoring. However, the applicability of these recommendations to Asian populations, who may have different cardiovascular risk profiles and metabolic responses, remains unclear. This paper aims to bridge this gap by evaluating the compatibility of fasting versus non-fasting lipid profiles specifically within a multi-ethnic Asian population in Singapore, focusing on patients with type 2 diabetes mellitus (T2DM) who are already on statin therapy.

Methods

The researchers conducted a prospective, intraindividual comparison study. They collected blood samples for lipid profiles from 470 adult patients diagnosed with T2DM and dyslipidemia who were on stable statin therapy. Each patient provided two samples: one after an overnight fast and another non-fasting sample collected within 6 hours of their last meal and within 14 days of the fasting sample.

The primary analytical tool used to compare the fasting and non-fasting lipid profiles was the intraclass correlation coefficient (ICC). An ICC value greater than 0.75 was considered indicative of good correlation. They also used paired t-tests to assess the mean differences between the fasting and non-fasting measurements for total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), and non-HDL-C. Patient adherence to statin treatment was assessed using the Medication Adherence Report Scale (MARS-5).

Datasets

The study utilized data from 470 multi-ethnic Asian adult patients with T2DM and dyslipidemia from two primary care clinics in Singapore. The data collected included demographic information (age, gender, race), medical history (T2DM duration, statin use, insulin use, HbA1c), and lipid profiles (TC, HDL-C, TG, LDL-C, non-HDL-C) from both fasting and non-fasting blood samples. Information on food and drink consumption prior to the non-fasting test was also collected. No new benchmarks were introduced; the study focused on evaluating existing lipid measurement protocols.

Key Insights and Ideas

  • Non-fasting tests are largely compatible with fasting tests in this Asian cohort: The study's core finding is that for most lipid markers, non-fasting tests provide results that are highly correlated with fasting tests, suggesting they can be reliably used for monitoring.
  • Triglycerides are the exception: While most lipid markers showed good agreement, triglycerides (TG) exhibited only moderate correlation. This is an important distinction, as elevated TG can impact LDL-C calculations.
  • Clinical implications for convenience and efficiency: If non-fasting tests are adopted, it could significantly reduce patient burden by eliminating the need for fasting, and streamline laboratory operations by allowing for more flexible scheduling.
  • Statin adherence is high: The study found very high adherence to statin treatment among participants, which is crucial because it minimizes the likelihood that changes in lipid levels are due to variations in medication rather than the fasting status.
  • Asian-specific data is crucial: The paper emphasizes that evidence from Western populations may not directly translate to Asians, highlighting the need for studies like this to inform guidelines for diverse ethnic groups.

Key Results

  • The mean difference between non-fasting and fasting levels for TC, HDL-C, TG, LDL-C, and non-HDL-C were +0.04, -0.001, +0.48, -0.15, and -0.05 mmol/l, respectively.
  • ICC values:
    • TC: 0.820 (Good)
    • HDL-C: 0.873 (Good)
    • TG: 0.579 (Moderate)
    • LDL-C: 0.764 (Good)
    • Non-HDL-C: 0.825 (Good)
  • All lipid components, except HDL-C, showed statistically significant differences between fasting and non-fasting measurements when analyzed by paired t-test. However, the absolute differences for most markers (excluding TG) were small.
  • 64.9% of patients achieved a perfect score on the MARS-5, indicating high adherence to statin therapy.
  • A small percentage of patients (1.28%) had non-fasting TG levels ≥ 4.5 mmol/l, which prevented LDL-C calculation using the Friedewald formula.

Limitations and Future Work

  • LDL-C calculation: The study relied on the Friedewald formula for LDL-C calculation, not direct measurement. While this reflects real-world laboratory practice, direct LDL-C measurement would offer a more definitive comparison.
  • TG impact on LDL-C: The elevated TG in non-fasting samples in some individuals prevented LDL-C calculation, highlighting a limitation for those specific cases.
  • Food intake details: The data on food and drink intake was based on patient recollection and was categorized broadly, limiting detailed analysis of specific food impacts.
  • Ethnicity stratification: The study population was not large enough to perform meaningful sub-group analyses by ethnicity to see if different ethnic groups responded differently.
  • Other lipid particles: The study did not investigate other lipid particles like VLDL or chylomicrons, which are not routinely measured.

Future Work: The authors suggest that further studies are needed to evaluate the implementation and cost-effectiveness of introducing non-fasting lipid panels routinely in primary care settings. Additionally, larger studies with more diverse ethnic representation could clarify any ethnic-specific differences.

Summary updated

Abstract

Recent Western guidelines recommend non-fasting lipid profiles to manage dyslipidaemia. We explored its applicability to an Asian population. We determined the differences between an overnight fasting and non-fasting cholesterol profiles of patients with type-2 diabetes mellitus (T2DM) in Singapore. We studied 470 multi-ethnic Asian adult patients with T2DM and dyslipidaemia from 2 primary care clinics in Singapore. Non-fasting blood specimens were collected within 6 h after their last meal and within 14 days of a fasting specimen. The intraclass correlation coefficient (ICC) was used to compare the intraindividual lipid profiles. An ICC value > 0.75 implies good correlation. The mean age and T2DM duration of the study population were 62.5 years and 9.8 years respectively. Their mean non-fasting period was 2.46 h. The mean differences between non-fasting and fasting total cholesterol (TC), high-density lipoprotein (HDL-C), triglyceride (TG), low-density lipoprotein (LDL-C), and non HDL-C were + 0.04 mmol/l, - 0.001 mmol/l, + 0.48 mmol/l, - 0.15 mmol/l, and - 0.05 mmol/l respectively. The ICC (95% CI) for TC, HDL-C, TG, LDL-C and non-HDL-C were 0.820 (0.788-0.847), 0.873 (0.850-0.893), 0.579 (0.516-0.636), 0.764 (0.723-0.799), and 0.825 (0.794-0.852) respectively. The fasting and non-fasting lipid profiles were similar in our local Asian patients with T2DM taking statin. Their non-fasting lipid profile can be used to assess their cholesterol treatment status.
BibTeX
@inproceedings{phoon2021compatibility,
  author = {Ian Kwong Yun Phoon and Yi Ling Eileen Koh and Xiaoxuan Guo and Usha Sankari and Ngiap Chuan Tan},
  title = {{Compatibility between an overnight fasting and random cholesterol tests in Asians}},
  booktitle = {Scientific},
  year = {2021},
  doi = {10.1038/s41598-021-85914-y},
}
Figure