Article Text
PDF
Abstract
Introduction Identify non-glycemic factors affecting the relationship between fasting plasma glucose (FPG) and glycated hemoglobin (HbA1c), in order to refine diabetes diagnostic criteria.
Research design and methods Relationship between FPG–HbA1c was assessed in 12 531 individuals from 2001 to 2018 US National Health and Nutrition Examination Survey. Using a recently described method, FPG and HbA1c were used to calculate apparent glycation ratio (AGR) of red blood cells for different subgroups based on age, race, and gender.
Results At an FPG of 7 mmol/L, black individuals had a higher HbA1c (p<0.001, mean: 50.2 mmol/mol, 95% CI (49.8 to 50.4)) compared with white individuals (47.4 mmol/mol (47.2 to 47.5)). This corresponds to NGSP (National Glycohemoglobin Standardization Program) units of 6.7% and 6.5% for black versus white individuals, respectively. Similarly, individuals under 21 years had lower HbA1c (p<0.001, 47.9 mmol/mol (47.7 to 48.1), 6.5%) compared with those over 50 years (48.3 mmol/mol (48.2 to 48.5), 6.6%). Differences were also observed between women (p<0.001, 49.2 mmol/mol (49.1 to 49.3), 6.7%) and men (47.0 mmol/mol (46.8 to 47.1), 6.5%). Of note, the difference in HbA1c at FPG of 7 mmol/L in black females over 50 and white males under 21 years was 5 mmol/mol (0.46%). AGR differences according to race (p<0.001), age (p<0.001), and gender (p<0.001) explained altered glucose–HbA1c relationship in the analyzed groups.
Conclusions FPG–HbA1c relationship is affected by non-glycemic factors leading to incorrect diagnosis of diabetes in some individuals and ethnic groups. Assessment of AGR helps understand individual-specific relationship between glucose levels and HbA1c, which has the potential to more accurately diagnose and manage diabetes.
- Diabetes Mellitus, Experimental
- Diagnosis
Data availability statement
Data are available in a public, open access repository. Data were compiled from the National Health and Nutrition Examination Survey (NHANES) cohorts, between the period of 2001 and 2018 which are publicly available on the US Centers for Disease and Control website.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Glycated hemoglobin (HbA1c) can vary based on factors other than glycemia, including conditions that affect red blood cell (RBC) lifespan.
WHAT THIS STUDY ADDS
The current study employs a mechanistic model based on RBC lifespan and glucose uptake to both provide a systematic explanation for this discrepancy and practical input on its impact in diabetes diagnosis on an individual basis.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE, OR POLICY
Data indicate that HbA1c measurement alone is not sufficient to accurately make the diagnosis of diabetes; for example, careful attention should be given to older adults of black origin who are at higher risk of incorrect diagnosis of diabetes with the sole use of HbA1c.
Introduction
Glycated hemoglobin (HbA1c) remains the benchmark biomarker for glycemic management.1 However, HbA1c can vary based on factors other than glycemia, including conditions that affect red blood cell (RBC) lifespan.1
A previous study assessed the agreement between HbA1c and fasting plasma glucose (FPG) as means of diagnosing diabetes in 6890 adults who self-identified as normoglycemic from the 1999–2006 cohorts of the National Health and Nutrition Examination Survey (NHANES).2 The study highlighted non-glycemic differences associated with racial background and age that in turn affected the relationship between FPG and HbA1c. Participants with HbA1c ≥48 mmol/mol (6.5%) but FPG <7 mmol/L (126 mg/dL) were significantly older and more likely to be black than participants with HbA1c <48 mmol/mol (6.5%) but FPG >7 mmol/L (126 mg/dL). This is consistent with literature suggesting that black individuals can have higher HbA1c than white individuals despite similar glucose levels.3–8 These ethnic differences in HbA1c are clinically meaningful with one study demonstrating that HbA1c overestimates average glucose control in black individuals compared with white individuals by 4.4 mmol/mol (0.4%).9 Others have shown that half of the normoglycemic black women have HbA1c values greater than 39 mmol/mol (5.7%).10 It has also been shown that black individuals are more likely than other populations to be diagnosed with diabetes based on HbA1c, while diabetes diagnosis based on oral glucose tolerance test (OGTT) is similar to other populations.3 4 11
In addition to ethnic background, age has also been shown to impact the relationship between glucose and HbA1c with each decade resulting in an increase of 0.57 mmol/mol (0.074%) in HbA1c for a given average glucose.12 These findings offer explanations for the inferior accuracy of HbA1c alone for diabetes diagnosis compared with the combination of this marker with FPG or OGTT.13 Underdiagnosis or overdiagnosis of diabetes has important clinical implications at the individual and population level, an area that remains incompletely studied. Therefore, there is a need to increase accuracy of diabetes diagnosis through a full understanding of the non-glycemic factors that modulate HbA1c, including RBC properties.
Recently, a glucose–HbA1c kinetic model has been described that accounts for individual RBC glucose uptake and lifespan.14 This kinetic model relies on an individual’s apparent glycation ratio (AGR) to characterize the personalized relationship between glucose and HbA1c. Incorporating a personalized AGR value has allowed the kinetic model to more accurately predict HbA1c compared with established markers, such as glucose management indicators.14–16
The main aim of the present study is to examine the role of personal non-glycemic factors on the relationship between fasting glucose and HbA1c. More specifically, we analyzed the impact of race, age, and gender on HbA1c at FPG Diabetes Diagnosis Criteria (DDC) as defined by the American Diabetes Association’s Standards of Medical Care in Diabetes17 (126 mg/dL, 6.99 mmol/L), employing a modified version of the kinetic model. We complemented the work by providing a mechanistic explanation for the discrepancy between HbA1c and FPG in specific subgroups of individuals through analysis of personalized AGR. While previous studies have documented the effects of race and age on the glucose–HbA1c discrepancy, the current study employs the mechanistic model based on RBC lifespan and glucose uptake to both provide a systematic explanation for this discrepancy and practical input on its impact in diabetes diagnosis on an individual basis.
Subjects, materials, and methods
Data were compiled from the National Health and Nutrition Examination Survey (NHANES) cohorts, between the period of 2001 and 2018 which are publicly available on the US Centers for Disease and Control website.18 Annual NHANES cohorts include nationally representative cross-sectional samples of about 5000 non-institutionalized civilians and include FPG and HbA1c for those over the age of 12.18 The National Centers for Health Statistics Ethics Review Board approves each NHANES data collection protocol (Protocol #98-12, 2005-06, 2011-17, 2018-01) and participants provide written informed consent for inclusion in these studies. Additional details regarding ethical approval are publicly available.19 NHANES data are anonymized, deidentified, and publicly available.18 The current analyses used data from the 2001–2018 cohorts and were limited to non-pregnant individuals who fasted for ≥8 hours at the time of their blood collection, had valid plasma glucose and HbA1c values, and did not have a self-reported history of diabetes. A total of 12 530 individuals were included in the analysis and were divided into groups based on their self-identified race, age, and gender. The individual’s age provided in years was binned into three groups <21 years, 21–50 years, and >50 years. These bin ranges were chosen because the legal age for adulthood in all US states is achieved by the age of 21. Moreover, the transition from pediatric to adult medical care generally occurs prior to age 21.20 The upper limit of 50 years was chosen because the risk for various diseases and comorbidities including diabetes increases after this age.21 The classification for race was optimized to ensure data continuity from 2001 to 2018. The full list of variables used is provided in online supplemental table 1.
Supplemental material
Determination of individual AGR and HbA1c at FPG DDC
HbA1c is affected not only by average glucose but also glucose uptake by RBCs and cellular lifespan. AGR takes into account RBC glucose uptake and lifespan, calculated from steady-state glucose levels and HbA1c, as previously described.14 22 23 In the present work, AGR was calculated by substituting steady-state glucose with FPG, consequently allowing the prediction of HbA1c at FPG DDC.14–16
Statistical analysis
A one-way analysis of variance (ANOVA) was used to identify significant factors that affect HbA1c at FPG DDC by individual characteristics of race, age, and gender. Additional effect tests were conducted by calculating the F-ratio of one-way ANOVA with individual factors compared with two factors (eg, race and age) and three factors (race, age, and gender) at a time. Therefore, a total of 24 subgroups were constructed and analyzed that comprised combinations of males/females, three distinct age groups (<21 years, 21–50 years, and >50 years), and race/ethnicities of white, black, Hispanic, and other/multiracial. Pairwise comparisons by t-tests were used between groups for HbA1c at FPG DDC. A post-hoc Tukey-Kramer test was then used to identify significantly different pairwise comparisons.
Results
Population
Baseline characteristics of individuals studied based on race, age, and gender are shown in table 1. There was equal gender distribution with white individuals representing the largest group (n=5352), followed by Hispanic individuals (n=3428), black individuals (n=2590), and individuals of multiracial origin (n=1160). The majority were between the ages of 21 and 50 years (n=5203), followed by those aged >50 years (n=4255) with those younger than 21 years representing the smallest group (n=3072). Groups were further subdivided according to gender, the four racial groups and three age groups, resulting in 24 subgroups. Main characteristics of the subgroups analyzed are detailed in table 2. A summary of the effects of race, gender, and age on HbA1c values is displayed in figure 1.
The effects of race, age, and gender on HbA1c at FPG DDC. Circles represent the mean value while bars represent the 95% CI. Populations in each subpanel that share a common color are not significantly different. FPG DDC, fasting plasma glucose Diabetes Diagnosis Criteria; HbA1c, glycated hemoglobin.
Baseline characteristics for subjects included in the analysis
Baseline characteristics for each subgroup included in the analysis
Race
Race affected the relationship between FPG and HbA1c in the NHANES dataset (p<0.001). Black individuals had the highest HbA1c at FPG DDC (mean: 50.2 mmol/mol, 95% CI (49.8 to 50.4), IFCC (The International Federation of Clinical Chemistry and Laboratory Medicine), 6.7% NGSP, figure 1) and their AGR was 72.67 mL/g (72.4 to 72.94). In contrast, white individuals had the lowest HbA1c at FPG DDC, 47.4 mmol/mol ((47.2 to 48.0), 6.5%) with AGR of 69.78 mL/g (69.61 to 69.95). The Hispanic population had a similar HbA1c at FPG DDC compared with white individuals at 47.4 mmol/mol ((47.3 to 47.7), 6.5%, p=0.22) and their mean AGR was 69.87 mL/g (69.67 to 70.07). However, these values were different when compared with the black population (p<0.001). Multiracial participants had an HbA1c at FPG DCC of 48.5 mmol/mol ((48.2 to 48.9), 6.6%) with AGR of 70.95 mL/g (70.59 to 71.32). These values were significantly different from all other races (p<0.001).
Gender
There were significant gender-based differences in glucose–HbA1c relationship (p<0.001). Women had a higher HbA1c at FPG DDC 49.2 mmol/mol ((49.1 to 49.3), 8%) and AGR of 71.66 mL/g (71.5 to 71.81) compared with men with HbA1c at FPG DDC was 47.0 mmol/mol ((46.8 to 47.1), 7.7%; p<0.001) and AGR of 69.34 mL/g ((69.18 to 69.51); p<0.001).
Age
There were significant age-based differences in the glucose–HbA1c relationship when subjects were binned into three groups: <21 years, 21–50 years, and >50 years (p<0.001). There were no significant differences between the two younger groups (p=0.28). However, subjects >50 years of age were different compared with <21 years (p<0.001) and the 21–50 years group (p=0.002). Subjects >50 years had a higher HbA1c at FPG DDC (48.3 mmol/mol (48.2 to 48.5), 6.6%), AGR (70.8 mL/g (70.6 to 71)). In comparison, subjects <21 years and 21–50 years had similar HbA1c at FPG DDC of 47.9 mmol/mol ((47.7 to 48.1), 6.5%) and 48.0 mmol/mol ((47.8 to 48.2), 6.5%), respectively. They had similar AGR of 70.29 mL/g (70.08 to 70.5) and 70.41 mL/g (70.23 to 70.59), respectively, with both being different when compared with those>50 years (p<0.001 for both).
Race, gender, and age
HbA1c at FPG DDC for each of the 24 subpopulations based on race, gender, and age was analyzed and results are shown in figure 2. Hispanic males <21 years had the lowest HbA1c, while black females had the highest HbA1c with a difference of 5 mmol/mol (0.46%) comparing these two subgroups.
HbA1c at FPG DDC for each subgroup based on race, age, and gender. Circles represent the mean value while bars represent the 95% CI. Subgroups that share a common color are not significantly different from each other. FPG DDC, fasting plasma glucose Diabetes Diagnosis Criteria; HbA1c, glycated hemoglobin.
While figure 2 illustrates the overlap between some subgroups, it is clear that within each age and gender category, black race had the highest HbA1c followed by multiracial individuals, while whites and Hispanics had similar HbA1c. Also, women and older individuals had higher HbA1c regardless of race.
AGR mirrored the changes in HbA1c with lowest values in white individuals and Hispanic individuals while black individuals showed the highest value with those of multiracial background having an intermediate value (table 3). AGR was higher in women and tended to increase with age (table 3). Therefore, the lowest AGR was observed in white/Hispanic young men, while black old women displayed the highest AGR (table 2).
HbA1c at FPG DDC and AGR based on race, age, and gender
Although there are distinct variations among the subgroups, within-group variations showed even larger differences. For example, comparing the two most extreme groups: Hispanic males <21 years and black females >50 years, the SDs (range) of HbA1c at FPG DDC were 5.5 mmol/mol (0.5%, 27.0 to 76.0), whereas interindividual SD within the black female group was higher at 7.1 mmol/mol (0.7%, 32.5 to 77.4) with a similarly high SD comparing interindividual variability with the Hispanic male group
Interactions between race, age, and gender
Significant interactions were identified between gender and race (p=0.006) as well as race and age (p=0.018). Men of all races, except for black men, had lower HbA1c at FPG DCC than women. Age also impacted subjects from different races (table 2). For example, there are no significant differences in HbA1c at FPG DCC comparing white males <21 and >50 years old, while a difference was detected comparing Hispanic males <21 and >50 years old. No significant interactions were found between gender and age (p=0.086) or race, age, and gender (p=0.827).
Within-group and between-group variation in AGR
Although there are differences between the subgroup means, there was higher variation in AGR values within each group compared with differences between groups. The SD in group mean AGR for all 24 subgroups was 1.71 mL/g. For comparison, black females older than 50 years of age had a 7.48 mL/g SD in AGR while white males less than 21 years of age had an SD of 5.07 mL/g in AGR. Thus, the variation in AGR within a group was distinctly larger than the variation between groups.
Our data indicate that AGR values are higher in black versus white individuals and in old versus younger individuals. To visualize this, we generated a scatter plot of HbA1c versus FPG for black females older than 50 years of age and white males less than 21 years of age (figure 3). Individuals with the group mean AGR for either black females older than 50 years of age (blue circles) or white males less than 21 years of age (red circles) are indicated by the blue and red solid lines, respectively.
Scatter plot of HbA1c versus fasting plasma glucose for black females greater than 50 years of age (red circles) and white males less than 21 years of age (blue circles). Each circle represents data from one subject. The solid red and blue lines for the correspondingly colored circles represent where individuals with the group mean AGR would be found. The four labeled quadrants (1–4) arise from an intersection of the lines corresponding to a fasting plasma glucose of ~5.6 mmol/L (100 mg/dL) and an HbA1c of 39 mmol/mol (~5.7% units), cut-off used for normal glucose metabolism. AGR ranges of the encompassed individuals are indicated for each quadrant. Individuals in quadrant 1 exhibit values for HbA1c and FPG that are in agreement with a diagnosis of normal glucose metabolism, while those in quadrant 3 are consistent with a diagnosis of diabetes. Quadrant 2 encompasses individuals potentially with an HbA1c value that incorrectly points to a diagnosis of abnormal glucose metabolism (pre-diabetes/ diabetes). Quadrant 4 encompasses individuals with an HbA1c value that potentially leads to a missed diagnosis of abnormal glucose metabolism. AGR, apparent glycation ratio; FPG, fasting plasma glucose; HbA1c, glycated hemoglobin.
We then dissected the scatter plot into four quadrants using two lines corresponding to thresholds for individuals with impaired glucose tolerance (pre-diabetes) or diabetes, namely lines for an HbA1c of 39 mmol/mol (~5.7%) and an FPG of ~5.6 mmol/L (100 mg/dL). Examination of the data indicated that each quadrant encompassed individuals with distinct AGR values. Quadrants 1 and 3 consist of individuals whose values for HbA1c and FPG are in agreement with a diagnosis of either normal glucose metabolism or a diagnosis of diabetes, respectively.
The individuals in quadrant 2, with an AGR >73 mL/g, potentially exhibit an HbA1c value that incorrectly points to a diagnosis of diabetes. From the scatter plot, these individuals appear to predominantly consist of the older black females. In contrast, the individuals in quadrant 4, with an AGR <71 mL/g, possess an HbA1c value that potentially leads to a missed diabetes diagnosis. This group tends to have more of the younger white male individuals.
Discussion
The objective of this study was to use a novel kinetic model to understand the relationship between FPG and HbA1c in subpopulations within the 2001–2018 NHANES cohort, representing individuals without a known diagnosis of diabetes. Our data show that race, age, and gender all impact AGR and HbA1c at FPG DDC, which may lead to incorrect diagnosis of diabetes if only HbA1c is tested. The results indicate up to 5 mmol/mol (0.46%) difference in average HbA1c at FPG DDC, comparing subpopulations of individuals. While this work has been conducted in individuals without a known diagnosis of diabetes, findings are consistent with studies in those with diabetes demonstrating an effect of race, age, and gender on HbA1c.4 6 9 12 24 Therefore, a single HbA1c alone is insufficient to give an accurate reflection of average glycemia across diverse populations resulting in some cohorts being underdiagnosed while others are being incorrectly labeled as having diabetes and consequently started on unnecessary therapies.4 25 26
Race had the greatest impact on HbA1c at FPG DDC and this was further compounded by gender and age. Black race resulted in higher HbA1c at FPG DDC, while Hispanics displayed similar HbA1c to the white population. Age had an equally important role in determining the relationship between fasting glucose and HbA1c. Older individuals had higher HbA1c at FPG DDC, which was also the case for women. Taken together, older black women had the highest HbA1c while young white men displayed the lowest HbA1c at FPG DDC. It should be noted that variations were apparent not only between groups but interindividual variations were evident within groups, arguing for personalized glycemic targets that take into account non-glycemic factors when assessing for the presence of diabetes. Incorrect diagnosis of diabetes can create unnecessary anxiety for an individual, increase treatment burden, and put additional pressure on health resources, while also having insurance implications. Unfortunately, NHANES began explicitly collecting data on Asian Americans only in 2011. Hence, this diverse population with high diabetes prevalence could not be included in the current analysis but warrants further study.27
In addition to HbA1c at FPG DDC, the current analysis used AGR as an indicator of individual and subpopulation-based differences in the relationship between HbA1c and glucose. AGR is the product of RBC lifespan and glucose uptake into RBCs considering apparent hemoglobin glycation rate constant.14 Average RBC lifespan affects individual cell exposure to circulating glucose levels. While the average lifespan of RBCs is thought to be approximately 105 days,28 experimental studies have recorded a 20% variation in mean RBC lifespan between healthy individuals with no pathological hemoglobin abnormalities.29 This explains the AGR variation observed in the NHANES population, which can be further impacted by cellular glucose uptake.
Among black females >50 years and Hispanic males <21 years old, a difference of 5.45 units in AGR corresponded to a 0.47% (5.1 mmol/mol) difference in HbA1c at FPG DDC. In total, 35% of subjects who self-identified as not having diabetes had an FPG DDC that was >0.5% (5.4 mmol/mol) HbA1c points different from 47.5 mmol/mol (6.5%). While 7.49% had ≥1% (11 mmol/mol) discrepancy from 47.5 mmol/mol HbA1c at mean FPG DDC. This indicates that up to a third of individuals can be misdiagnosed as having impaired glucose tolerance or diabetes (when their glucose levels are normal) while others incorrectly labeled as having normal glucose metabolism (when they have abnormal levels). Therefore, characterizing the effects of non-glycemic factors such as race, gender, and age on the relationship between plasma glucose and HbA1c allows clinicians to account for these factors when diagnosing diabetes.
We also compared black females >50 years and white males <21 years in a scatter plot of HbA1c versus FPG. Data were divided into four quadrants based on an HbA1c of 39 mmol/mol (5.7%) and an FPG of 5.6 mmol/L (pre-diabetes or diabetes). Among the individuals with an HbA1c that incorrectly leads to a diagnosis of deranged glucose metabolism, these were primarily the older black women. In contrast, individuals with an HbA1c leading to a missed diagnosis of diabetes tended to be the younger white males. These observations correlated with AGR values, consistent with the concept that higher AGR increases hemoglobin glycation for a given plasma glucose level, whereas the opposite would hold true for lower AGR values. Thus, the ability to calculate an individual’s AGR could potentially refine the HbA1c–glucose relationship at the personalized level.
One criticism of the current study is the use of fasting, rather than mean glucose, to calculate AGR and evaluate the relationship with HbA1c. It can be argued that our findings of elevated HbA1c in the black population, particularly women, are due to higher postprandial glucose in this population. However, this is unlikely given previous work showing lower postprandial glucose in black females compared with whites matched for age and body mass index.30 Moreover, Bergenstal and colleagues9 previously used continuous glucose monitoring (CGM) to measure mean glucose in 104 black and 104 white individuals with type 1 diabetes and correlated mean glucose to HbA1c in these cohorts. Their findings showed a 0.4% (4.4 mmol/mol) increase in HbA1c of black individuals for a given mean glucose compared with those of white origin. Despite the use of FPG and a population without a previous diabetes diagnosis, our data of 5 mmol/mol (0.46%) increase in the HbA1c at FPG DDC of black females >50 years compared with Hispanic (or white) males <21 years old, closely mirror these results. Moreover, these findings are consistent with reports of females requiring more aggressive hypoglycemic therapies to bring their HbA1c to similar levels of male counterparts and may also explains their higher hypoglycemic exposure.31 32
The current findings suggest that non-glycemic factors affect the relationship between glucose and HbA1c regardless of whether mean or fasting glucose is used for the analysis (eg, mean glucose via BGM/CGM or FPG). However, we acknowledge that the population studied will have limited postprandial glucose fluctuations compared with an established diabetes group. Therefore, it remains unclear whether relying on FPG to calculate AGR is valid in a diabetes population on hypoglycemic therapies and future work in this area is required. A large population-based study with high definition HbA1c and regular/consistent glucose readings is needed to deconstruct AGR in subgroups of individuals into its components in order to determine whether observed differences are due to altered glucose uptake and/or variations in RBC lifespan. Moreover, it would be valuable to study how variability in the glucose–A1c relationship impacts long-term diabetes outcomes.
Our findings have implications for clinical practice as data indicate that HbA1c measurement alone is not sufficient to accurately make the diagnosis of diabetes. Careful attention should be given to older adults of black origin who are at higher risk of incorrect diagnosis of diabetes with the sole use of HbA1c. While combining HbA1c with FPG will improve diagnostic accuracy, there will be cases of discrepancy between these two measures, which is currently resolved by a glucose tolerance test. A potential alternative is CGM that gives a more global assessment of glycemia while also facilitating the calculation of personalized AGR, using both fasting and mean glucose levels, thus further improving diagnostic accuracy. In addition to refining the diagnosis of diabetes, establishing AGR will help with the long-term management of diabetes and impaired glucose metabolism.
Taken together, future work is required to develop additional criteria for the accurate diagnosis of diabetes, particularly in some ethnic groups and older individuals. Moreover, studies are required to explore the role of CGM in diabetes diagnosis, given the wealth of glycemic data provided by these systems and the opportunity to calculate person-specific AGR, which can increase diagnostic accuracy and subsequent management of those confirmed to have diabetes.
Supplemental material
Data availability statement
Data are available in a public, open access repository. Data were compiled from the National Health and Nutrition Examination Survey (NHANES) cohorts, between the period of 2001 and 2018 which are publicly available on the US Centers for Disease and Control website.
Ethics statements
Patient consent for publication
Ethics approval
The National Centers for Health Statistics Ethics Review Board approves each NHANES data collection protocol (Protocol #98-12, 2005-06, 2011-17, 2018-01) and participants provide written informed consent for inclusion in these studies. Additional details regarding ethical approval are publicly available. Centers for Disease Control and Prevention (CDC), National Center for Health Statistics (NCHS), NCHS Ethics Review Board (ERB) Approval. Hyattsville, Maryland: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, NHANES 199-2018 (https://www.cdc.gov/nchs/nhanes/irba98.htm). Participants gave informed consent to participate in the study before taking part.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors YX developed the kinetic model and equations used to analyze the data. YR performed calculations and data analysis. RAA and TD managed and supervised the project. All authors were involved in the design of the research, analysis, interpretation of the data, and article writing. All authors worked collaboratively to review and prepare the final article. RAA is the guarantor of the work.
Funding This work was funded by Abbott Diabetes Care.
Competing interests YR, YX, AC, and TD are employees of Abbott Diabetes Care. RAA received other research support and Honoraria from Abbott Diabetes Care.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.