The in vitro reflection of nutrition in the clinic is the changes in the biochemical parameters in blood samples as a result of metabolic changes in the body. As a result of digesting food, serum glucose, amino acid, and triacylglycerol levels increase.1As a response to ingestion of food, the pancreas reduces glucagon secretion by increasing insulin secretion and tries to keep the blood glucose level in the reference range. This change in the insulin/glucagon levels triggers the anabolic stage in tissues (especially the liver, muscle, and adipose tissues). The liver forms glycogen and lipids from the substrates that are supplied. While the glycogen that forms is stored in the liver, lipids are transferred to the blood via very-low-density lipoproteins (VLDL). Adipose tissue removes free fatty acids from lipoproteins, synthesizes triacylglycerols, and stores them in the form of insoluble droplets. When the metabolism is in the state of non-fasting, the brain and the heart use glucose as their energy source. Only after long-term fasting, they can use only ketone bodies ( 1 , 2 ). While transitioning from non-fasting to fasting, the body secretes the glucagon hormone from the α cells of the pancreas. The main purpose of glucagon is to try to keep the blood glucose level around normoglycemic levels in fasting ( 3 ). By regulating the metabolism, the insulin/ glucagon hormones regulate lipolysis/lipogenesis and glycogenolysis or glycogen synthesis in a balanced way. Therefore, all bodily enzymes/hormones function in a way to keep the metabolism in balance despite fasting or non-fasting ( 4 ).

While the metabolism is conducting all these processes, a difference in blood biochemistry occurs between the states of fasting and non-fasting. However, this difference is on a negligible level for many tests ( 5 – 7 ). Importance is paid to the states of fasting and non-fasting in Turkey in terms of biochemical tests. For this reason, in our study, by taking the biochemical parameters of samples arriving at our laboratory, it was aimed to determine whether or not there were differences between the fasting and non-fasting blood samples in the statistical and clinical sense, and if any, how this was reflected on the clinic.

The study included 18 different biochemical tests used most frequently by clinics. As we considered that the probability of blood samples taken at early hours of the morning to be fasting blood samples would be higher than the same probability for those taken towards the afternoon and in the afternoon, the patients’ blood tests were categorized as fasting for 07:00-10:00 and non-fasting for 10.01-17:00. The blood samples that were transferred to the laboratory under appropriate conditions were kept for 20 min, centrifuged for 10 minutes at 4°C and 4000 rpm, separated into their serums, and analyzed with an ADVIA 1800 Clinical Chemistry System autoanalyzer.

Our study covered a total of 70,352 individuals including 42,696 (60%) women and 27,656 (40%) men. Among these individuals, 46,731 (66%) were assessed as fasting, and 23,621 (34%) were assessed as non-fasting. As seen in Table 1, while there was no significant difference in Amylase (AMY), Iron, Lipase (LIP), Phosphor, Total cholesterol, High-density lipoprotein, Triglycerides, and Alkaline Phosphatase (ALP), some differences were observed in Albumin (ALB), Alanin Aminotransferase (ALT), Aspartate Aminotransferase (AST), Creatinine (CRE), Calcium (Ca), Gamma-Glutamyl Transferase (GGT), Glucose (GLU), Blood Urea Nitrogen (BUN), Total Protein (TP), and Creatinine kinase (CK) at p<0.05. As shown in Table 2, between the fasting and non-fasting values of the women, there was no significant difference in AMY, AST, Phosphor, Total cholesterol, CRE, iron, ALP, TP, Triglycerides, High-density lipoprotein and CK, while there were significant differences in ALB, ALT, Ca, GGT, GLU, BUN, and LIP at p<0.05.

Table 3 has demonstrated that, between the fasting and non-fasting values of the men, while there was no significant difference in amylase, Total cholesterol, CRE, Iron, GGT, BUN, LIP, Triglycerides, ALP, and High-densitylipoprotein, there were significant differences in ALB, ALT, AST, Phosphor, Ca, GLU, TP, and CK at p<0.05. However, according to the effect size values that were calculated, even the parameters that were found to be statistically significant were not significant (effect size min-max=0.006- 0.104) (Table 2-3).

Additionally, while there was no significant difference among the women in AST, IRON, TP, ALP, High-density lipoprotein, and CK (Table 2), there was no significant difference among the men in IRON, GGT, BUN, ALP, and High-density lipoprotein (Table 3). As it may be seen here, although some parameters such as CK and AST were not statistically significantly different among the women between the fasting and non-fasting values (p>0.05), the differences in the men were significant (p<0.05) (Tables 2-3). Again, while parameters such as GGT and BUN were not significantly different among men, they were significantly different among the women. According to the effect size values that we measured to determine whether or not the parameters that turned out to be statistically significantly different at p<0.05 between the fasting and non-fasting groups were clinically significant, no clinically significant difference was observed in any of the parameters. While the smallest effect size value was 0.016 for High-density lipoprotein, the largest one was 0.104 for CK, and there was no clinical significance according to Cohen’s d (Table 1).

In the literature, while it had been accepted until the 2010s that the appropriate metabolic state is the fasting state since the length of stay is desired to be shortened and since the laboratory is asked for serving all applicant patients 24 hours a day, blood is taken from patients at any time of the day, and it is not possible to be sure of the fasting state of these patients ( 11 , 12 ). Due to the crowdedness of blood collection units in the early morning, patients who arrive at the hospital to provide blood samples at 8 am may have to wait until noon. The fact that this waiting process is difficult for diabetic, pregnant, elderly, or pediatric patients, in particular, or the fact that these people cannot wait on an empty stomach, may cause distress for phlebotomists and patients.

In fact, outside the hours of the day, one is required to fast, the body metabolism is in a state of non-fasting most of the day. It might not be the right thing to still assess individuals with fasting metabolism tests ( 6 , 13 ). While research on the difference between fasting and non-fasting values in humans is mostly on lipids, a limited number of studies covered other biochemistry parameters including albumin, bilirubin, and uric acid ( 13 – 15 ).

Langsted et al. assessed lipids, lipoproteins, apolipoproteins, and albumin in diabetic and non-diabetic individuals at different periods following their latest meal. As a result, they reported that the plasma Triglycerides amounts in a non-fasting state increased by only 0.2 mmol/L in comparison to fasting in diabetic and non-diabetic individuals, non-High-density lipoprotein and Apolipoprotein-B stayed constant before and after meals, LDL-CHOL and albumin amounts decreased in diabetic and non-diabetic individuals, and this was probably caused by hemodilution as a result of fluid intake. They also emphasized that the use of non-fasting blood in lipid profile measurements is more useful than fasting blood ( 16 ). Mora et al. assessed lipid profiles and reported changes in Triglycerides by 0.2 mmol/L, Total cholesterol by 0.1 mmol/L, and LDL-CHOL by 0.2 mmol/L, while there was no significant change in High-density lipoprotein ( 17 ). Similarly, in our study, no significant difference was seen between the fasting and non-fasting values for Total cholesterol, Triglycerides, and High-density lipoprotein (p=0.366, p=0.307, p=0.527, respectively) (Table-1). In the review by Nordestgaard et al. which was supported by the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM), according to the results of studies assessing fasting and non-fasting lipid profiles, it was proposed that non-fasting should be the standard, several countries (the United Kingdom, Canada, Scandinavian countries, etc.) have changed their clinical guidelines, and fasting is not required in lipid profile measurements in either diabetic or non-diabetic individuals, those with or without cardiovascular diseases and children or adults ( 5 , 18 ). They argued that fasting samples could be repeated for only patients with a doubt in routine results, and fasting samples could be collected based on their own guidelines in patients with non-fasting triglyceride levels of >5 mmol/L (440 mg/dL) or chronic hyper triglyceridemia ( 5 ).

Plumelle et al. assessed the fasting and non-fasting results of 77 tests (37 biochemistry, 16 hematology, 3 coagulation, 21 endocrine parameters). They stated that, among the 37 biochemistry parameters, 29 were not affected by fasting or non-fasting (ALT, ALB, ALP, AMY, Apo A1, AST, B ₂ M, BIC, BUN, Ca, Total cholesterol, CK, GGT, HbA1c, High-density lipoprotein, Iron, K, LDH), the remaining were affected by food ingestion at p<0.05 (UA, TBIL, BNP, CREA, GLU), but as they did not affect the total change limit (TCL), they did not have clinical significance (TCL=√(2.77CVa)2+ (0.5CVb)2).14 In our study, similar to the findings of the aforementioned studies, there were no differences in the ALT, ALP, AST, ALB, BUN, Ca, GGT, IRON, CK, TP, and Total cholesterol parameters.

The Bispebjerg study conducted by Sennels et al. prospectively examined the effects of fasting and non-fasting states on the circulating concentrations of 14 frequently used clinical biochemical parameters. Consequently, among the parameters common with ours in the 14 examined parameters, while there was no difference between fasting and non-fasting values in creatinine, there were differences in creatine kinase, ALT, and AST ( 19 ). Additionally, Plumelle et al. stated that there was no significant difference between fasting and non-fasting in ALT and AST values, and therefore, regardless of the hour of the day, measurements on the metabolism did not differ based on non-fasting or fasting states ( 14 ). In our study, the p-value between the fasting and non-fasting values was 0.001 for creatinine, 0.007 for ALT, and 0.001 for AST (Table 1).

While studies on the effects of fasting and non-fasting on only the lipid profile among the 18 parameters that were examined in our study could be encountered in the literature, there is a limited number of studies regarding the other parameters we included. For this reason, we were not able to compare our results on some parameters to values reported in the literature.

Considering the mean values of the ALB, ALT, AST, Ca, CRE, GGT, Glu, TP, CK, and BUN parameters, it was found that they were close to each other between the statuses of fasting and non-fasting, but there were differences. However, when effect size analysis that is used in studies with large samples was applied, no clinical significance was found. In addition to this, no significant difference was found between the fasting and non-fasting values in the parameters of High-density lipoprotein, Triglycerides, Phosphor, Total cholesterol, AMY, Lipase, and ALP. Moreover, while there were differences in some parameters when this situation was assessed based on sex, again, clinically insignificant results were obtained. In conclusion, biochemical values obtained by collecting blood at any time of the day may be used to make diagnoses of diseases without considering fasting/non-fasting status.