Few Lab Tests Actually Need 8 Hours of Fasting, Study Suggests

Fasting less than 8 hours before routine blood tests generally did not change results, with few exceptions, according to an observational study.

In paired within-person analyses of over 100,000 patients undergoing multiple blood collections, the vast majority of 372 test items did not have established analytical performance specifications exceeded with less than 8 hours of fasting, reported Sollip Kim, MD, PhD, of the University of Ulsan College of Medicine in Seoul, Korea, and colleagues in JAMA Internal Medicine.

However, within-person differences were somewhat higher for certain tests comparing <8 hours fasting versus 8-12 hours and versus >12 hours:

• Glucose: +13.0 mg/dL and +10.8 mg/dL, respectively
• Triglycerides: +17.1 mg/dL and +20.1 mg/dL
• Gamma-glutamyl transferase: +21.7 U/L and +14.8 U/L
• Lipase: comparison not available versus 8-12 hours and +36.4 U/L versus >12 hours

“This cross-sectional study found that fasting duration was associated with differences exceeding analytical performance specifications for only a few analytes,” Kim and colleagues wrote. “These findings support analyte-specific rather than uniform fasting requirements.”

Fasting, or the intentional allowance of time between the patient’s last meal and blood collection, is a longstanding practice thought to standardize preanalytical variation of laboratory results.

Indeed, in this study, those who fasted less than 8 hours before a test showed a linear trend by which glucose and lactate were significantly lower per additional hour of fasting. In contrast, lipid results were not any better with more hours of fasting, the authors noted.

“Clinicians should be aware that results from very short fasting may require careful interpretation for a limited set of postprandial-sensitive analytes,” they wrote.

Current practice in managing and diagnosing diabetes is already reliant on HbA1c, a standard measure of blood sugar over the previous 2-3 months that is less sensitive to day-to-day variability. However, fasting plasma glucose is still an important test in diabetes, especially in cases where HbA1c is not reliable.

Regarding cholesterol checks, U.S. guidelines already strongly endorse either a nonfasting or fasting lipid profile for adults and children in general; a fasting lipid profile is favored, however, for LDL cholesterol assessment in those with a family history of dyslipidemia or premature heart disease, a personally known or suspected disorder in triglyceride metabolism, or nonfasting triglyceride levels exceeding 400 mg/dL.

In an accompanying editorial, Børge Grønne Nordestgaard, MD, DMSc, and colleagues of Copenhagen University Hospital in Denmark noted that the study “provides strong support for accumulating evidence that fasting blood sampling is not necessary.”

“Evidence showing that fasting adds little if any benefit to the interpretation of laboratory test results for most blood samples is an important message for laboratories, clinicians, and patients,” they wrote. “Eliminating the need for fasting should reduce the number of patients who present to the laboratory unprepared for sampling, decrease unnecessary administrative tasks related to fasting blood samples, and decompress morning schedules, freeing time and resources that could go into more important tasks.”

“For clinicians, same-day nonfasting testing can simplify patient instructions and lead to faster diagnosis and treatment,” they added. “For patients, lessening the burden of preparation before blood sampling and repeated visits should improve adherence with testing, follow-up visits, and prescribed therapies, and may lead to better satisfaction with the healthcare system.”

This retrospective single-center study from Korea included adult outpatients with documented fasting duration. Blood samples were drawn by the institution’s central phlebotomy unit.

The final dataset comprised 9,755,547 test results from 101,148 patients across 372 test items. Median age was 61 years, and 43% were women.

Kim and team used the European Federation of Clinical Chemistry and Laboratory Medicine Biological Variation Database as a reference for total allowable error, ultimately defining the largest difference in a laboratory result that is still acceptable without materially affecting clinical interpretation. They minimized between-patient confounding by opting for paired within-person analyses.

Prolonged fasting beyond 12 hours conferred no measurable advantage for any analyte, they reported.

The study was limited by the restriction to a single center’s outpatient phlebotomy unit and the lack of adjustment for macronutrient composition of the food consumed prior to testing. Its applicability to younger people is uncertain as well.

Nordestgaard and colleagues also pointed to the limited applicability of the results to ethnicities other than East Asian. “That said, we believe that the results in this study are generally relevant, including for all ethnicities and countries,” they wrote.