Article Text
PDF
Abstract
Aim To investigate whether patients follow advice regarding the duration of a fast, prior to blood tests which require an overnight fast for their correct interpretation.
Method Patients referred for a glucose-tolerance test gave informed consent. They were asked to complete a questionnaire regarding the duration of their fast, and an additional blood sample was collected for fasting total bile-acid measurement.
Results 184 patients were recruited. The median duration of fast self reported by patients was 14 h (range 12–26 h). Total bile-acid concentrations were within the fasting reference range except for six individuals. The median total bile-acid concentration found was 2.1 μmol/l (range 0.1–19.5 μmol/l).
Conclusion All patients who self-reported a duration of fast reported a fast recognised as being adequate for fasting blood tests. This was confirmed in almost all by low normal total bile-acid measurements, which were well within the fasting range.
- Fasting status
- glucose-tolerance test
- total bile acids
- impaired fasting glycaemia
- impaired glucose tolerance
- chemical pathology
Statistics from Altmetric.com
- Fasting status
- glucose-tolerance test
- total bile acids
- impaired fasting glycaemia
- impaired glucose tolerance
- chemical pathology
Introduction
Individuals undergoing biochemical investigation may be asked to fast overnight prior to the collection of a blood sample in order to control known variations in analyte concentration as a result of diet and fasting status. There is significant variation in the advice found regarding the required duration of a fast prior to a blood test. The local laboratory advises a 14 h fast. Some clinicians may advise patients to fast from midnight the night before a fasting blood test, regardless of the time of the patients' appointments for blood collection. Internet sites such as lab-test online sites advise a fast of 8–10 h, while other NHS sites may advise a 12–16 h fast.1 2
Besides there being variable advice for duration of fasting, it is also possible that individuals may not completely understand or be able to follow the instructions they have been given prior to a blood test.
The interpretation of the results of a glucose-tolerance test is dependent on an adequate duration of fast. An elevated fasting glucose level can result in individuals being categorised as having either diabetes mellitus or impaired fasting glycaemia. On occasion, there may be doubt about the significance of elevated baseline blood glucose levels if the 2 h value is not also significantly raised.
Total bile-acid concentrations are also altered, by feeding or fasting. Total bile-acid concentrations are known not to be altered in individuals with diabetes mellitus independently of fasting in the absence of cholestasis.3 Therefore, the measurement of total bile-acid concentrations in serum at the same time as the baseline blood-glucose concentration could give an assessment of the effectiveness of fasting. For these reasons, a study was conducted to compare fasting total bile-acid concentrations in individuals undergoing a glucose-tolerance test with the outcome of the glucose-tolerance test. The fasting total bile-acid concentrations were also compared with the subjects' anonymously self-reported durations of fast for the test.
Methods
Patients were recruited sequentially through the standard glucose-tolerance test clinic appointment process. Prior to the date of appointment, patients were sent an information sheet describing the study. On the day of their appointment, patients were interviewed by MS, one of the investigators, and informed consent was obtained. Subsequently, patients completed a questionnaire regarding their intake of food and fluids during the 24 h period prior to their glucose-tolerance test. Completed questionnaires were collected anonymously using a ‘ballot box’. All the glucose-tolerance tests were conducted using the standard protocol used in the hospital. When a fasting sample was collected for the fasting blood glucose into a fluoride oxalate tube, an additional baseline sample was also collected into a plain tube for total bile-acid concentration determination. An equivalent of 75 g of anhydrous glucose was administered as 113 ml Polycal diluted to 250–300 ml with water.
Two hours after the polycal had been consumed, a further blood sample was taken for glucose measurement.
The outcomes of the glucose-tolerance tests were interpreted according to WHO criteria.4
Plasma-glucose concentrations in the samples were measured using the Olympus supplied and CE-marked hexokinase method installed on an Olympus 2700 and an Olympus 5400 (Beckman Coulter UK, High Wycombe, UK). The method was conducted using suppliers' settings throughout.
The linear range of the method was 0.6–45.0 mmol/l. The internal coefficient of variation for the duration of the study ranged from 2.10 to 2.22% at a concentration of 3.4 mmol/l to 1.82–1.91% at a concentration of 19.2 mmol/l.
Total bile-acid concentrations in serum were measured using a 3-hydroxy steroid dehydrogenase enzymatic method with reagents supplied by Bio-stat (Bio-stat Diagnostic Systems, Stockport, Cheshire UK). The linear range of the total bile-acid method was 1–180 μmol/l. The detection limit was 1 μmol/l. The internal CV for the period of the study was 7.83% at a concentration of 19.5 μmol/l. The local laboratory reference range was 1–14 μmol/l, based on local data for intrahepatic cholestasis of pregnancy. The Bio-stat supplied fasting reference range was 1–10 μmol/l.
Instrument parameters for the Olympus AU2700 and AU5400 (Beckman Coulter UK, High Wycombe, UK) were set according to the manufacturers' instructions.
All results were statistically analysed using the Stats Direct commercial Windows-based statistical package version 2.66, supplied by Stats Direct Limited. Non-parametric statistical methods were used for comparisons of the median and range and distribution of values.
Ethical permission was obtained from the local Cheshire NHS Ethics committee.
Results
A total of 195 patients were recruited, but 11 did not have a total bile-acid concentration measurement, leaving 184 with total bile-acid concentrations. In this group of 184, there were 98 (53%) men and 86 (47%) women. The median age for the men was 62 years (range 28–93), while for the women it was 67 years (range 31–89). The overall median age was 65 years, range of 28 to 93 years. The median duration of fast reported by 168 patients who provided a completed questionnaire was 14 h (range 12 to 26 h). The median total bile-acid concentration for all individuals studied was 2.1 μmol/l (range <1.0–19.5 μmol/l). There were no statistically significant differences in median total bile-acid concentrations for men, 2.4 μmol/l (<1.0–10.1) and women 1.9 μmol/l (<1.0–19.5). The spread of total bile-acid concentrations observed did not follow a normal distribution (figure 1). There was a prolonged shoulder running into higher values but with only six values (3.2%) above the fasting bio-stat reference range of 1–10 μmol/l. There were only two values above the local reference for an obstetric cholestasis range of 1–14 μmol/l.
Distribution of total bile-acid concentrations in fasting samples.
With regard to the outcome of the glucose-tolerance test, 27 patients had a normal response. A diabetic response was found in 68. Of these, 21 had elevations of both the fasting glucose concentration and the 2 h glucose concentration consistent with diabetes mellitus. Thirty-seven had a non-diabetic fasting glucose level but a 2 h glucose level consistent with diabetes mellitus. Only 10 individuals were classified as having diabetes mellitus solely on the basis of their fasting glucose concentration. Impaired glucose tolerance was found in 66 individuals. The remaining 22 individuals were found to have impaired fasting glycaemia.
The median fasting total bile-acid concentrations and ranges found associated with the different responses to the glucose-tolerance test are shown in table 1.
Median (range) total bile-acid concentrations in the baseline blood samples collected at the start of a glucose-tolerance test compared with the outcome of the glucose-tolerance test
There were no statistically significant differences between the median total bile-acid concentrations found in the fasting samples, with respect to the outcome of the glucose-tolerance test for all groups except for individuals with diabetes mellitus identified on the fasting sample. When these individuals were compared with individuals with a normal glucose-tolerance test, the median total bile-acid concentration was higher. For those with a normal glucose-tolerance test response, the median total bile-acid concentration was 1.8 μmol/l (1.0–9.0) compared with 3.2 μmol/l (1.2–4.3) for individuals with diabetes on the fasting sample, p<0.05. The median fasting bile-acid concentrations in both groups were within the lower part of the reference range and all bile-acid results in both groups were within the fasting range for the test. Individuals with impaired fasting glycaemia did not have a statistically different median fasting bile-acid concentration from those with a normal glucose-tolerance test.
The overall median total bile-acid concentration of 2.1 μmol/l was well within the lower part of the fasting reference range, of 1–10 μmol/l. Almost all subjects' total bile-acid concentrations were within the lower part of the reference range. Only 15 individuals had total bile-acid concentrations greater than or equal to 7 μmol/l, and only six had concentrations greater than the fasting Bio-stat reference range of 1–10 μmol/l.
When the total bile-acid concentrations were divided into two groups using a cut-off of less than 7 μmol/l or 7 μmol/l or more, there was no statistical difference in the outcome of the glucose-tolerance test from the χ2 test. Questionnaires regarding the length of fast were completed by 168 individuals. No relationship between declared length of fast and total serum bile-acid concentration in serum was found for these 168 subjects (figure 2).
Stated duration of fast and total bile-acid concentration.
Discussion
Only 15% of individuals studied had a normal glucose-tolerance test response. The most frequent glucose-tolerance test interpretation made was that the result was consistent with diabetes mellitus. This was found in 37% of patients. Impaired glucose tolerance was found in 36%. The high proportion of abnormal responses to the glucose-tolerance test studied suggested that patients were being appropriately referred for this investigation.
Diabetes mellitus and impaired fasting gycaemia as well as impaired glucose tolerance are all associated with raised liver function tests in up to 15–27% and also with a finding of fatty liver.3 5 6 It has previously been shown that elevations of conventional liver function tests in individuals with diabetes mellitus are not associated with raised total bile-acid concentrations unless there was cholestasis.3 The existence of slight abnormalies of routine liver-function tests in individuals referred for a glucose-tolerance test is therefore not a contraindication to the measurement of total bile-acid concentrations to assess the level of fasting. Individuals who have clinical or laboratory evidence of cholestasis should not undergo a glucose-tolerance test.
Individuals who were found to have diabetes mellitus based solely on the fasting blood glucose value had a statistically significantly higher median total bile-acid concentration when compared with individuals with a normal glucose-tolerance test. Despite this finding all total bile-acid concentrations in the group with diabetes mellitus from the fasting blood glucose sample were within the fasting total bile-acid reference range, suggesting that this finding was not due to a poor duration of fasting or to undiagnosed liver disease.
In this study, the median self-reported duration of fast was 14 h (range 10–26 h). Most individuals reported a fast of between 13 and 15 h with only a few fasting less than 12 h and more than 16 h. The local laboratory-recommended duration of fast was 14 h. These results indicate that patients in general had been given advice regarding fasting and had largely followed it.
For most subjects (91%) the fasting total bile-acid concentrations were below 7 μmol/l suggesting that most individuals had fasted well indeed. All individuals with a bile-acid concentration of more than 7 μmol/l had also stated that they had fasted more than 12 h. Six (3.2%) individuals studied had a total bile-acid concentration greater than 10 μmol/l. Only two had values greater than 14 μmol/l. Of these six individuals, five self reported a fast duration of between 13 and 16 h. One did not return a questionnaire. This individual had the highest total bile-acid concentration of 19.5 μmol/l and had impaired fasting glycaemia with a fasting glucose of 6.7 mmol/l and a 2 h glucose of 6.7 mmol/l. They also had normal routine ‘liver function tests.’ Together, these findings might suggest that this individual had not fasted as required.
Overall, self-reporting of duration of fast by patients therefore appears to be a valid way of accepting or excluding patients for fasting tests.
The observation that 96.8% of fasting bile-acid concentrations were below 10 μmol/l, which was consistent with the self-reported duration of fast, suggests that if there is any doubt about the fasting status of an individual, measurement of total bile-acid concentration in clotted blood sample collected at the start of the test will confirm the adequacy of the fast.
In conclusion, although patients might have had access to variable advice regarding the duration of fast, almost all patients fasted correctly and, according to local guidance, prior to their glucose-tolerance test. The addition of a total bile-acid concentration measurement to a glucose-tolerance test might rarely aid in the interpretation of fasting glucose-tolerance test results.
Take-home messages
Varying advice regarding the correct length of fast for fasting blood tests exists.
Nonetheless, patients do fast correctly for glucose-tolerance tests.
Fasting total bile-acid concentration measurement may rarely help in deciding if patients have fasted correctly.
Interactive multiple choice questions
This JCP article has an accompanying set of multiple choice questions (MCQs). To access the questions, click on BMJ Learning: take this module on BMJ Learning from the content box at the top right and bottom left of the online article. For more information please go to: http://jcp.bmj.com/education. Please note: the MCQs are hosted on BMJ Learning the best available learning website for medical professionals from the BMJ Group. If prompted, subscribers must sign into JCP with their journal's username and password. All users must also complete a onetime registration on BMJ Learning and subsequently log in (with a BMJ Learning username and password) on every visit.
Footnotes
Competing interests None.
Ethics approval Ethics approval was provided by the Cheshire REC Committee.
Provenance and peer review Not commissioned; externally peer reviewed.