Abstract
Background and Aims:
Many traditional assessments of risk for coronary heart disease (CHD) and diabetes require laboratory studies performed after an 8-hour fast. We assessed whether metabolic-syndrome (MetS) severity would remain linked to future CHD and diabetes even when assessed from non-fasting samples.
Methods and Results:
Participants in the Atherosclerosis Risk in Communities study were assessed at 4 visits and followed for 20-years of adjudicated CHD outcomes. We used Cox proportional-hazard models (for 20-year CHD outcomes) and logistic regression (for 9-year diabetes outcomes) to compare incident disease risk associated with a race/ethnicity-specific MetS-severity Z-score (MetS-Z) calculated in participants who were fasting (≥8 hours) or non-fasting. All analyses were adjusted for sex, race, education, income and smoking. MetS Z-scores were overall similar between participants who were always fasting vs. those non-fasting at Visits 1–3 (all values −0.1 – 0.4), while MetS-Z for participants who were non-fasting at Visit-4 were higher at each visit. Baseline MetS-Z was linked to future CHD when calculated from both fasting and non-fasting measurements, with hazard ratio (HR) for fasting MetS-Z 1.53 (95% confidence interval [CI] 1.42,1.66) and for non-fasting 1.28 (CI 1.08,1.51). MetS-Z at Visit-1 also remained linked to future diabetes when measured from non-fasting samples, with odds ratio for fasting MetS-Z 3.10 (CI 2.88,3.35) and for non-fasting 1.92 (CI 1.05,3.51).
Conclusions:
MetS-Z remained linked to future CHD and diabetes when assessed from non-fasting samples. A score such as this may allow for identification of at-risk individuals and serve as a motivation toward interventions to reduce risk.
Keywords: cardiovascular disease, diabetes, metabolic syndrome, risk, fasting
INTRODUCTON
Many traditional assessments of risk for coronary heart disease (CHD) and diabetes require laboratory studies performed after at least an 8-hour fast, including lipid panels to assess for elevations in triglyceride and LDL levels[1] and glucose measurements to reveal elevated fasting glucose (5.5 – 6.9 mmol/L; 100 – 125 mg/dL) or diabetes (≥ 7.0 mmol/L, 126 mg/dL).[2] In particular, triglycerides rise by approximately 20% following a meal [3] (an absolute change of 0.29 mmol/L, 26 mg/dL [4]) while non-fasting glucose levels that are below the threshold for diagnosing diabetes (≥ 11.1 mmol/L, 200 mg) lack additional cut-off values that are helpful for interpreting results from fasting samples.[2] However, obtaining fasting samples cannot be done during the majority of clinic appointments and requires significant planning and effort on the part of the patient—such that these assessments and the opportunity to identify at-risk individuals are often missed. In part because of these issues, recent statements by the National Society of Clinical Biochemistry in Denmark[5] and the American Diabetes Association (ADA)[2] advise non-fasting assessments as screening tools, such as total cholesterol, non-HDL cholesterol and HbA1c. These non-fasting assessments provide levels of triglycerides and glucose that may vary depending on time since meal, in contrast to standardized assessments such as an oral fat tolerance test or oral glucose tolerance test, in which a patient ingests a known quantity of fat or glucose, with tests at specific intervals to measure variation.[6, 7] Proponents of non-fasting samples point out that non-fasting measures can provide a high degree of accuracy in risk identification and should become standard for routine screening.[1, 5, 8]
One risk identifier that relies on fasting laboratory information is characterization of the metabolic syndrome (MetS), a cluster of cardiovascular disease (CVD) and diabetes risk factors that occur together more often than would be expected by chance.[9] MetS is traditionally assessed using sets of criteria such as those of the Adult Treatment Panel-III[9] that provide cut-off values to identify abnormalities in the individual components of elevated waist circumference, high blood pressure, low HDL cholesterol, high triglycerides and elevated fasting glucose—with three out of five components requiring fasting measurements. As an alternative to sets of dichotomous criteria to diagnose MetS, we derived a continuous MetS severity Z-score (MetS-Z) based on how the MetS components correlated together on a sex- and race/ethnicity-specific basis.[10] In contrast to dichotomous criteria, which are not associated with risk beyond what is seen for the individual MetS components (causing some to say that the value of MetS, “is not worth the sum of its parts,” this score is associated CVD and diabetes, even in models that include the individual MetS components.[11, 12] The association of this score with features of CVD,[11] diabetes,[12] chronic kidney disease[13] and non-alcoholic fatty liver disease[14] make it a potential tool for assessing risk in clinical settings, particularly with incorporation into electronic health record systems.[15]
We previously assessed for whether MetS-Z would be accurate at predicting CHD risk in the setting of diabetes, which included individuals with elevations in fasting glucose > 6.9 mmol/L (125 mg) and found that both the 5-component MetS-Z score and a version of the score derived without glucose were both associated with risk of future CHD.[16] This assessment validated use of the score in the setting of high levels of fasting glucose; however, the association of the score with future disease in the context of non-fasting glucose is unclear. Similarly, while artificial elevations in triglycerides due to non-fasting status would be expected to increase the score, these levels would only pertain to the triglyceride portion of the score, and may not greatly alter associations of the complete score with future disease risk.
We hypothesized that MetS-Z when measured using non-fasting blood samples (compared to fasting samples) would be similarly associated with risk for CHD and diabetes. We set out to evaluate this using fasting and non-fasting measures from the Atherosclerosis Risk in Communities study.
MATERIAL AND METHODS
Study Population
The ARIC study is a community-based epidemiological cohort study across 4 field centers in the US, with the following data collection cycles: Visit 1 (1987–1989), Visit 2 (1990–1992), Visit 3 (1993–1995) and Visit 4 (1996–1998), and on-going follow-up for adjudicated CVD outcomes thereafter. This study and/or these analyses analysis were approved by the Institutional Review Boards of the University of Florida and the ARIC study sites. The study conformed to the ethical guidelines of the Declaration of Helsinki. Further details of the study design and objectives have been previously described[17]. A total of 15,792 participants aged 45–64 years provided informed consent to be enrolled in the study.
Measurement of metabolic syndrome components
Details regarding procedures for blood collection and analysis for lipids[18] and glucose[19] have been reported previously. Briefly, participants were instructed to fast overnight for 12 hours before the examination; at each study visit, participants reported whether they had performed this fast. Phlebotomy was performed and serum and plasma samples were sent to a central laboratory for analysis. Triglycerides were measured using enzymatic methods, and HDL was measured after dextran-magnesium precipitation. Serum glucose was measured via the hexokinase-6-phosphate dehydrogenase method[20]. Trained clinical staff measured waist circumference at the umbilical level to the nearest cm. BP was examined in sitting position with three measurements performed and the average of the last two utilized for analysis. Smoking was classified at each visit by participant self-report of current smoking status.
Study Outcomes
Incident CHD
Incident CHD was determined from adjudicated outcomes using standard ARIC protocols that identified fatal or nonfatal hospitalized myocardial infarction, fatal CHD, silent myocardial infarction identified by electrocardiography, or coronary revascularization, and hospitalized and fatal stroke.[19, 21] Adjudicated CHD outcomes used in the current analysis were through Dec 31, 2011, providing up to 24 years of follow-up information.
T2DM
Incident diabetes was determined for those without diabetes at Visit 1 if for Visits 2–4 participants reported that a physician had told them they had diabetes, had a fasting glucose ≥ 7.0 (126 mg/dL) or a non-fasting glucose ≥ 11.1 (200 mg/dL), or if they reported use of insulin or oral hypoglycemic medications[22]. Visits were approximately 3 years apart, providing 9 years of follow-up.
Predictor: MetS Severity Score
We calculated MetS-severity Z-scores for study participants (both fasting and non-fasting) based on equations derived previously and described elsewhere[10]. MetS-severity score equations were derived from the five traditional MetS components (WC, fasting triglycerides, HDL-cholesterol, systolic BP, fasting glucose) using a factor analysis approach. Because of differences in traditional MetS criteria by race/ethnicity[23–25], a confirmatory factor analysis was performed as previously described[10] to determine the weighted contribution of each component to a latent MetS factor on a sex- and race/ethnicity-specific basis, using the National Health and Nutrition Examination Survey (NHANES) data (including fasting triglycerides and glucose) for adults aged 20–64 years. For each of the subgroup defined by sex and race/ethnicity, factor loadings from the five MetS components were determined and used to generate equations for computing a standardized MetS severity score (https://metscalc.org/). The MetS severity score was shown to correlate with other MetS risk markers, such as insulin[26] and adiponectin[26], and is predictive of long-term risk of T2D[12, 26, 27] and CVD[11, 26, 28]—and these associations are above and beyond the individual MetS components alone.[11, 12] In addition, MetS-Z is able to track changes in risk over time.[29, 30]
Statistics
From the ARIC sample described above, we excluded participants with consent-related exclusions (n = 395), those other than blacks or whites (n=46) and those with a history of CHD (n = 758), stroke (n = 278), or diabetes (n = 1,835) at baseline (Visit 1), multiple non-fasting visits (n=207), or missing MetS components (n = 4,881). Participants could have been excluded on the basis of ≥ 1 of these criteria. From this sample, we were primarily interested in those individuals with MetS severity scores at each of the first four visits (n = 8,555). This sample was used to calculate mean MetS severity scores (and components of MetS) across the first four visits, and was used to estimate risk of CHD given the lengthy follow-up. Given the relatively short period of follow-up in determining incident diabetes, we included all individuals with just a MetS severity score at Visit 1 and diabetes status across the four visits, which expanded the sample for this analysis (n = 11,197).
From the sample of 8,555, we focused on individuals who either had fasting labs across all four visits (n = 8,072), or who had just one non-fasting lab at one of the visits (n = 483). Among the 483, we classified individuals by the visit in which they had a non-fasting lab: Visit 1 (n = 119), Visit 2 (n = 93), Visit 3 (n = 112), and Visit 4 (n = 159). Mean MetS-Z scores, and the components, were calculated at each visit across these five groups. Participant demographics and other characteristics were compared between those with fasting labs throughout and those with one non-fasting lab via t-tests or chi-square tests, as appropriate. In examining risk of CHD, we calculated time between the CHD event (or last follow-up) and Visit 1 (for those with fasting labs throughout), or from the visit when the non-fasting lab occurred (for those with a non-fasting lab). Cox proportional hazards regression was used to estimate hazard ratios (HR’s) of incident CHD associated with MetS-Z at the above visit, including an interaction term with fasting-status. This allowed for HR’s to be estimated for MetS-Z by fasting status, and to test whether the two HR’s were statistically different (i.e., a statistically significant interaction between fasting status and MetS-Z). Characteristics found to be different between the fasting categories were included in the model to account for potential confounding.
For the diabetes risk analysis, we focused on those individuals with a fasting lab at visit 1 (n = 11,004) and those with a non-fasting lab at visit 1 (n = 193). Given the nature of diabetes classification, we utilized logistic regression to model the odds of incident diabetes over the approximately 10-year period, using a similar approach described above (i.e., examination of the MetS-Z by fasting status interaction) to estimate fasting-status-specific odds ratios (OR’s), adjusting for identified confounders. All analyses were performed using SAS Version 9.4; statistical significance was defined as α = 0.05.
RESULTS
Participant characteristics
Table 1 displays baseline characteristics overall and by fasting status for the 8,555 participants who met inclusion/exclusion criteria for use in the CHD analysis. Relative to participants always fasting, those who were not fasting at one of the visits were more likely to be male, black, current smokers, and slightly younger. Participants who were non-fasting at either Visit 3 or 4 were also more likely to have higher incident CHD. Participants who were non-fasting at Visit 4 were also more likely to have income < $50,000. There were no differences by fasting vs. non-fasting status in BMI or in any of the individual MetS components.
Table 1.
ARIC descriptive statisticsa
| Overall | Fasting status | p-valueb | |||||
|---|---|---|---|---|---|---|---|
| Fasting all visits | Non-fasting: Visit 1 only | Non-fasting: Visit 2 only | Non-fasting: Visit 3 only | Non-fasting: Visit 4 only | |||
| N** | 8,555 | 8072 (94.4) | 119 (1.4) | 93 (1.1) | 112 (1.3) | 159 (1.9) | |
| Sex: (N, % Male) | 3762 (44.0) | 3503 (43.4) | 61 (51.3) | 62 (66.7) | 55 (49.1) | 81 (50.9) | < 0.0001 |
| Race: (N, % Black) | 1521 (17.8) | 1348 (16.7) | 35 (29.4) | 36 (38.7) | 34 (30.4) | 68 (42.8) | < 0.0001 |
| Visit 1 Statistics: | |||||||
| Education | 0.0029 | ||||||
| < High School | 1412 (16.5) | 1307 (16.2) | 22 (18.6) | 22 (23.9) | 28 (25.2) | 33 (20.8) | |
| High School | 2911 (34.1) | 2768 (34.3) | 37 (31.4) | 30 (32.6) | 32 (28.8) | 44 (27.7) | |
| > High School | 4223 (49.4) | 3991 (49.5) | 59 (50.0) | 40 (43.5) | 51 (46.0) | 82 (51.6) | |
| Income | 0.0044 | ||||||
| < $50,000 | 5685 (70.1) | 5342 (69.7) | 79 (70.5) | 63 (71.6) | 77 (76.2) | 124 (82.7) | |
| ≥ $50,000 | 2427 (29.9) | 2319 (30.3) | 33 (29.5) | 25 (28.4) | 24 (23.8) | 26 (17.3) | |
| Current Smoker (N, %) | 1822 (21.3) | 1648 (20.4) | 50 (42.0) | 39 (41.9) | 42 (37.5) | 43 (27.0) | < 0.0001 |
| Age | 53.8 (5.6) | 53.8 (5.6) | 52.7 (5.4) | 52.7 (5.6) | 53.3 (5.5) | 53.5 (6.0) | 0.0066 |
| BMI | 27.0 (4.8) | 27.0 (4.8) | 26.3 (4.7) | 26.5 (4.4) | 26.8 (4.1) | 28.6 (5.50) | 0.2864 |
| Waist Circumference | 95.2 (13.0) | 95.1 (13.0) | 93.4 (12.5) | 93.0 (12.2) | 95.3 (12.1) | 98.6 (13.8) | 0.5882 |
| HDL | 53.0 (16.8) | 53.0 (16.8) | 53.9 (17.4) | 53.8 (18.2) | 50.7 (17.7) | 50.6 (15.5) | 0.2165 |
| SBP | 120.2 (18.1) | 120.1 (18.1) | 118.4 (16.8) | 122.0 (21.7) | 120.3 (17.0) | 124.4 (17.6) | 0.0924 |
| Triglycerides | 122.3 (73.4) | 122.2 (73.4) | 120.1 (71.5) | 122.2 (66.6) | 124.0 (73.7) | 127.3 (78.6) | 0.6445 |
| Glucose | 98.3 (9.1) | 98.2 (9.0) | 97.4 (8.5) | 99.0 (9.3) | 99.1 (10.0) | 100.3 (10.3) | 0.0530 |
| ATP-III MetS (N, %) | 2702 (31.6) | 2531 (31.4) | 29 (24.4) | 31 (33.3) | 46 (41.1) | 65 (40.9) | 0.0630 |
| MetS Severity Score | 0.04 (0.75) | 0.04 (0.75) | −0.09 (0.76) | −0.07 (0.84) | 0.11 (0.77) | 0.22 (0.81) | 0.5369 |
| Incident Disease: | |||||||
| CVD (N, %) | 1344 (15.7) | 1252 (15.5) | 21 (17.7) | 13 (14.0) | 27 (24.1) | 31 (19.5) | 0.0380 |
| Diabetes(N, %) | 1896 (22.2) | 1775 (22.0) | 18 (15.1) | 18 (19.4) | 27 (24.1) | 58 (36.5) | 0.1155 |
Unless noted, mean ± SD are provided. All statistics are relative to the group defined by the column (except for overall N)
Chi-square test (categorical variables) or t-test (continuous variables) comparing those fasting at all visits (n=8072) to those with one non-fasting lab at one visit (n = 483)
MetS-Z by fasting status
Figure 1 displays mean levels of MetS-Z score by visit and fasting/non-fasting status for the analytic cohort. Across all time points, MetS-Z levels were overall similar when calculated using non-fasting vs. fasting samples at individual visits. The exception to this was among participants who were non-fasting at Visit 4 only, who had higher MetS severity already at Visit 1 compared to the always-fasting group and continued with higher levels than all other groups thereafter—even at Visits 2 and 3, when they were fasting. In evaluating the individual components at each given visit, these were not different between groups, with the exception of those not fasting at Visit 4, who had higher waist circumference, SBP and glucose throughout the visits (Supplementary Figure 1).
Figure 1.
Mean MetS-Z Severity Score by Visit and Fasting Status.
CHD Hazard Ratio by Fasting Status
Figure 2 displays HR’s for incident CHD by fasting status. The association between MetS Z-score and for future CHD by the standard MetS Z-score was overall not different when calculated from fasting measurements at Visit 1 and when calculated from non-fasting measures at any subsequent visit (HR for fasting 1.53 [95% CI 1.42–1.66]; HR for non-fasting 1.28 [95% CI 1.08–1.51]). The same was true when using a version of the MetS severity score derived and calculated without glucose and another version of the MetS severity score (“capped”) that utilized ceiling levels of triglycerides (4.5 mmol/L, 400 mg/dL) and glucose (13.9 mmol/L, 250 mg/dL), where any values above these ceiling levels were entered at the ceiling level.
Figure 2. Hazard Ratios of Incident CVD: Fasting (at Visit 1) vs. First Non-Fasting MetS-Z.
Data shown are from Cox Proportional Hazards Regression, comparing fasting vs. non-fasting “first” values (Visit 1 for those n=8072 who fasted all four visits, and the non-fasting value for those n=483 with a non-fasting MetS-Z). The model included sex, education, income, race, age and smoking status (at time of relevant visit) as covariates. The “Capped” score entered all values of triglycerides ≥400 mg/dL and glucose ≥250 mg/dL at these cut-off values. Interaction p-values between fasting status and MetS-Z: Standard: p = 0.0500, No-Glucose: p = 0.3255, “Capped”: p = 0.2937.
Diabetes Odds Ratio by Fasting Status
Figure 3 displays OR’s for future diabetes by MetS-Z as calculated using fasting vs. non-fasting measures at Visit 1 (to maximize follow-up time for incident diabetes). MetS-Z remained associated with future diabetes when calculated from non-fasting samples. There were no significant differences between OR’s from fasting samples and non-fasting samples for either the standard 5-component MetS-Z or the “capped” MetS severity score. The no-glucose score was not associated with future diabetes.
Figure 3. Odds Ratios of Incident Diabetes for MetS-Z by Fasting Status at Visit 1.
Data shown are from logistic regression, comparing fasting vs. non-fasting Visit 1 values. The model included sex, education, income, race, age and smoking status (at time of relevant visit) as covariates. The “Capped” score entered all values of triglycerides ≥400 mg/dL at these cut-off values; glucose values did not require capping in this analysis because no participants had diabetes at the time of MetS-Z assessment. There were no significant interactions between fasting status and all three forms of MetS-Z (standard p = 0.1212, no glucose p = 0.0745, trimmed p = 0.1259).
DISCUSSION
This study addressed the potential for a decrease in accuracy of a MetS severity score when calculated from non-fasting blood samples. We noted that MetS-Z levels calculated from non-fasting individuals in 3 out of 4 study visits were overall similar to levels from fasting individuals, despite the potential that high levels of non-fasting triglycerides and glucose would contribute to an inaccurate elevation in score. These non-fasting MetS-Z levels continued to have long-term links to future CVD and diabetes, with risk for future disease that was not significantly different from MetS-Z calculated from fasting samples (though with fairly wide confidence intervals). While prior data had established risk based on individual non-fasting risk factors,[3, 5] the validity of MetS as a disease predictor when measured from non-fasting samples had not been established. Given the inherent inconvenience in obtaining fasting samples,[1, 5, 8] these data may have implications for clinical use of a score such as this—which could ultimately be calculated via an electronic health record[15] to identify risk for future disease from measures performed during a clinical evaluation, irrespective of fasting status—improving the ability to assess MetS-related risk on a more convenient basis. Ultimately, risk scores like this could be helpful for triggering treatments and motivating patients toward lifestyle changes.[31]
Because ARIC study visits occurred every 3 years, we were unable to determine if MetS-Z assessed from non-fasting samples were higher than if they had been calculated at a similar time from fasting samples in the same individual. The components most commonly associated with elevations following food ingestion—triglycerides and glucose—were not significantly different from samples drawn among fasting participants at the same visit, likely emphasizing that food ingestion is just one influence of many in the wide range of values across the population for each of these components. It is important to note, though, that we found significant differences in characteristics of individuals who were non-fasting, introducing caution for any direct comparison of MetS-Z levels between these groups. In addition, these data reflect the predictive ability within groups, whereas there remains the theoretical potential that particularly high elevations in MetS-Z—for example due to high non-fasting levels of triglycerides—may over-estimate disease risk in a given individual. We addressed this possibility by assessing for changes in the strength of association of MetS severity with future disease when using a “capped” scoring system that assigned a value of triglyceride of 400 mg/dL and glucose levels at 250 mg/dL for all individuals with measures above these thresholds. While we did not note a difference in the predictive ability when using such caps, it remains possible that a capped system like this would be helpful in minimizing inaccuracies of risk prediction among non-fasting samples with triglyceride or glucose values at the high end of the spectrum.
While risk estimates for both CHD and diabetes were not significantly different when assessed between MetS-Z calculated from non-fasting and fasting samples, the point estimates for both CHD and diabetes tended to be lower when assessed among non-fasting samples. This may have been due in part to the underlying risk profile of non-fasting participants. For example, when compared to individuals who were fasting at all four visits, non-fasting individuals were more likely to be male, black and in the lower income category—necessitating inclusion of these variables in the prediction models. Participants who were non-fasting at Visits 3 and 4 had a higher incidence of CHD, a finding not seen in another prospective study comparing CHD risk among fasting and non-fasting participants.[32] The potential exists that additional unmeasured confounders further contributed to risk in the non-fasting individuals, resulting in the lower point estimates for MetS-Z-related risk.
Our findings that risk for future CHD correlated with non-fasting samples of MetS-Z were similar to prior studies of other estimates of CVD risk from non-fasting samples. Nordestgaard et al. reported that triglyceride levels from non-fasting samples were nevertheless highly correlated with risk of myocardial infarction, ischemic heart disease and total death over a mean 26-year period, with OR’s of 1.16 for men and 1.41 for women, for every 1 mmol/L increase in non-fasting triglyceride level.[33] Similarly, Langsted et al reported links between non-fasting LDL and triglycerides with cardiovascular events during 14 years of follow-up,[1] and Doran et al found that LDL cholesterol from non-fasting samples had an OR of 4.0 for cardiovascular mortality among individuals in the third vs. first tertile of LDL.[8] Each of these support that either non-fasting triglycerides or LDL levels estimated in part based on non-fasting triglycerides remain associated with future CVD.
Because of the possibility for diabetes risk prediction to be driven predominantly by glucose elevations—potentially revealing the presence of glucose intolerance, for example—we additionally assessed a four-component MetS-Z score derived and utilized without glucose.[16] We found that in this non-fasting setting the no-glucose MetS score was not predictive of future diabetes—due at least in part to wide confidence intervals. This suggests that elevations in glucose values as used in calculating MetS severity appear to have contributed to the significant prediction of future diabetes.
While this analysis benefitted from assessment of a large, prospectively-studied cohort, we recognize several limitations. We lacked assessment of MetS-Z measures in close proximity in the fasting and non-fasting state in the same individual, which would have allowed a more direct comparison of future risk by fasting status. Non-fasting occurred in a limited number of study participants, limiting power to assess for differences between the fasting and non-fasting groups. Non-fasting was only known by participant report of non-adherence to the study protocol and did not occur by design of the original study, likely resulting in important differences in demographics and lifestyle between fasting and non-fasting participants; however, we accounted for these differences by including multiple potential confounders in the models. Finally, although CHD outcomes were assessed via adjudicated review over the course of > 20 years, diabetes incidence is notoriously more difficult to document, leaving us to rely on assessments only at study visits.
CONCLUSIONS
We found that a MetS severity score as assessed from non-fasting samples remained associated with future risk for CHD and diabetes. This supports the validity of MetS when measured from non-fasting samples and suggests potential for a score such as this to be evaluated during clinical encounters among non-fasting blood individuals—improving the ability to assess risk on a more convenient basis. Future application of a score such as this would benefit from cut-off values designating particularly elevated risk for disease. Further research is needed to determine whether such cut-offs would be need to be determined separately for scores calculated from fasting vs. non-fasting samples.
Supplementary Material
Supplementary Figure 1: Mean MetS Components by Visit and Fasting Status. Data shown reflect levels of MetS components at each of the four visits evaluated by timing of a participant’s non-fasting status.
Acknowledgements
Funding: This work was supported by NIH grants 1R01HL120960 (MJG and MDD). The Atherosclerosis Risk in Communities study has been funded in whole or in part with Federal funds from the National Heart, Lung, and Blood Institute, National Institutes of Health, Department of Health and Human Services, under Contract numbers HHSN268201700001I, HHSN2682017000021, HHSN268201700003I, HHSN268201700004I, HHSN268201700005I. The authors thank the staff and participants of the ARIC study for their important contributions.
Footnotes
Completing Interests: None
BIBLIOGRAPHY
- 1.Langsted A, Freiberg JJ, Nordestgaard BG: Fasting and nonfasting lipid levels: influence of normal food intake on lipids, lipoproteins, apolipoproteins, and cardiovascular risk prediction. Circulation 2008, 118(20):2047–2056. [DOI] [PubMed] [Google Scholar]
- 2.American_Diabetes_Association: 2. Classification and Diagnosis of Diabetes:. Diabetes Care 2018, 41(Suppl 1):S13–S27. [DOI] [PubMed] [Google Scholar]
- 3.Sidhu D, Naugler C: Fasting time and lipid levels in a community-based population: a cross-sectional study. Arch Intern Med 2012, 172(22):1707–1710. [DOI] [PubMed] [Google Scholar]
- 4.Nordestgaard BG, Langsted A, Mora S, Kolovou G, Baum H, Bruckert E, Watts GF, Sypniewska G, Wiklund O, Borén J et al. : Fasting Is Not Routinely Required for Determination of a Lipid Profile: Clinical and Laboratory Implications Including Flagging at Desirable Concentration Cutpoints-A Joint Consensus Statement from the European Atherosclerosis Society and European Federation of Clinical Chemistry and Laboratory Medicine. Clin Chem 2016, 62(7):930–946. [DOI] [PubMed] [Google Scholar]
- 5.Langsted A, Nordestgaard BG: Nonfasting Lipid Profiles: The Way of the Future. Clin Chem 2015, 61(9):1123–1125. [DOI] [PubMed] [Google Scholar]
- 6.Kolovou GD, Watts GF, Mikhailidis DP, Pérez-Martínez, Mora S, Bilianou H, Panotopoulos G, Katsiki N, Ooi TC, Lopez-Miranda J et al. : Postprandial Hypertriglyceridaemia Revisited in the Era of Non-fasting Lipid Profiles: Executive Summary of a 2019 Expert Panel Statement. Curr Vasc Pharmacol 2019. [DOI] [PubMed] [Google Scholar]
- 7.DeBoer MD, Gurka MJ: Low sensitivity of the metabolic syndrome to identify adolescents with impaired glucose tolerance: an analysis of NHANES 1999–2010. Cardiovasc Diabetol 2014, 13:83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Doran B, Guo Y, Xu J, Weintraub H, Mora S, Maron DJ, Bangalore S: Prognostic value of fasting versus nonfasting low-density lipoprotein cholesterol levels on long-term mortality: insight from the National Health and Nutrition Examination Survey III (NHANES-III). Circulation 2014, 130(7):546–553. [DOI] [PubMed] [Google Scholar]
- 9.Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, Gordon DJ, Krauss RM, Savage PJ, Smith SC et al. : Diagnosis and management of the metabolic syndrome - An American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation 2005, 112(17):2735–2752. [DOI] [PubMed] [Google Scholar]
- 10.Gurka MJ, Lilly CL, Norman OM, DeBoer MD: An Examination of Sex and Racial/Ethnic Differences in the Metabolic Syndrome among Adults: A Confirmatory Factor Analysis and a Resulting Continuous Severity Score. Metabolism 2014, 63(2):218–225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.DeBoer MD, Gurka MJ, Hill Golden S, Musani SK, Sims M, Vishnu A, Guo Y, Pearson TA: Independent Associations between Metabolic Syndrome Severity & Future Coronary Heart Disease by Sex and Race. J Am Coll Card 2017, 69:1204–1205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Gurka MJ, Golden SH, Musani SK, Sims M, Vishnu A, Guo Y, Cardel M, Pearson TA, DeBoer MD: Independent associations between a metabolic syndrome severity score and future diabetes by sex and race: the Atherosclerosis Risk In Communities Study and Jackson Heart Study. Diabetologia 2017, 60(7):1261–1270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.DeBoer MD, Filipp SL, Musani SK, Sims M, Okusa MD, Gurka M: Metabolic Syndrome Severity and Risk of CKD and Worsened GFR: The Jackson Heart Study. Kidney Blood Press Res 2018, 43(2):555–567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Fermin CR, Lee AM, Filipp SL, Gurka MJ, DeBoer MD: Serum Alanine Aminotransferase Trends and Their Relationship with Obesity and Metabolic Syndrome in United States Adolescents, 1999–2014. Metab Syndr Relat Disord 2017, 15(6):276–282. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Scheitel MR, Kessler ME, Shellum JL, Peters SG, Milliner DS, Liu H, Komandur Elayavilli R, Poterack KA, Miksch TA, Boysen J et al. : Effect of a Novel Clinical Decision Support Tool on the Efficiency and Accuracy of Treatment Recommendations for Cholesterol Management. Appl Clin Inform 2017, 8(1):124–136. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Gurka MJ, Guo Y, Filipp SL, DeBoer MD: Metabolic syndrome severity is significantly associated with future coronary heart disease in Type 2 diabetes. Cardiovasc Diabetol 2018, 17(1):17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC investigators. Am J Epidemiol 1989, 129(4):687–702. [PubMed] [Google Scholar]
- 18.Sharrett AR, Ballantyne CM, Coady SA, Heiss G, Sorlie PD, Catellier D, Patsch W, Atherosclerosis Risk in Communities Study G: Coronary heart disease prediction from lipoprotein cholesterol levels, triglycerides, lipoprotein(a), apolipoproteins A-I and B, and HDL density subfractions: The Atherosclerosis Risk in Communities (ARIC) Study. Circulation 2001, 104(10):1108–1113. [DOI] [PubMed] [Google Scholar]
- 19.Folsom AR, Szklo M, Stevens J, Liao F, Smith R, Eckfeldt JH: A prospective study of coronary heart disease in relation to fasting insulin, glucose, and diabetes. The Atherosclerosis Risk in Communities (ARIC) Study. Diabetes Care 1997, 20(6):935–942. [DOI] [PubMed] [Google Scholar]
- 20.McNeill AM, Schmidt MI, Rosamond WD, East HE, Girman CJ, Ballantyne CM, Golden SH, Heiss G: The metabolic syndrome and 11-year risk of incident cardiovascular disease in the atherosclerosis risk in communities study. Diabetes Care 2005, 28(2):385–390. [DOI] [PubMed] [Google Scholar]
- 21.Folsom AR, Rasmussen ML, Chambless LE, Howard G, Cooper LS, Schmidt MI, Heiss G: Prospective associations of fasting insulin, body fat distribution, and diabetes with risk of ischemic stroke. The Atherosclerosis Risk in Communities (ARIC) Study Investigators. Diabetes Care 1999, 22(7):1077–1083. [DOI] [PubMed] [Google Scholar]
- 22.Schmidt MI, Duncan BB, Bang H, Pankow JS, Ballantyne CM, Golden SH, Folsom AR, Chambless LE, Investigators ARiC: Identifying individuals at high risk for diabetes: The Atherosclerosis Risk in Communities study. Diabetes Care 2005, 28(8):2013–2018. [DOI] [PubMed] [Google Scholar]
- 23.Sumner AE, Cowie CC: Ethnic differences in the ability of triglyceride levels to identify insulin resistance. Atherosclerosis 2008, 196(2):696–703. [DOI] [PubMed] [Google Scholar]
- 24.DeBoer MD, Dong L, Gurka MJ: Racial/Ethnic and Sex Differences in the Ability of Metabolic Syndrome Criteria to Predict Elevations in Fasting Insulin Levels in Adolescents. Journal of Pediatrics 2011, 159(6):975–981. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Walker SE, Gurka MJ, Oliver MN, Johns DW, DeBoer MD: Racial/ethnic discrepancies in the metabolic syndrome begin in childhood and persist after adjustment for environmental factors. Nutrition Metabolism and Cardiovascular Diseases 2012, 22(2):141–148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.DeBoer MD, Gurka MJ, Morrison JA, Woo JG: Inter-relationships between the severity of metabolic syndrome, insulin and adiponectin and their relationship to future type 2 diabetes and cardiovascular disease. Int J Obes (Lond) 2016, 40(9):1353–1359. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.DeBoer MD, Gurka MJ, Woo JG, Morrison JA: Severity of the metabolic syndrome as a predictor of type 2 diabetes between childhood and adulthood: the Princeton Lipid Research Cohort Study. Diabetologia 2015, 58(12):2745–2752. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.DeBoer MD, Gurka MJ, Woo JG, Morrison JA: Severity of Metabolic Syndrome as a Predictor of Cardiovascular Disease Between Childhood and Adulthood: The Princeton Lipid Research Cohort Study. J Amer Coll Card 2015, 66(6):755–757. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.DeBoer MD, Filipp SL, Gurka MJ: Use of a Metabolic Syndrome Severity Z Score to Track Risk During Treatment of Prediabetes: An Analysis of the Diabetes Prevention Program. Diabetes Care 2018, 41(11):2421–2430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Gurka MJ, Filipp SL, Pearson TA, DeBoer MD: Assessing Baseline and Temporal Changes in Cardiometabolic Risk Using Metabolic Syndrome Severity and Common Risk Scores. J Am Heart Assoc 2018, 7(16):e009754. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Chamnan P, Simmons RK, Sharp SJ, Griffin SJ, Wareham NJ: Cardiovascular risk assessment scores for people with diabetes: a systematic review. Diabetologia 2009, 52(10):2001–2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Mora S, Chang CL, Moorthy MV, Sever PS: Association of Nonfasting vs Fasting Lipid Levels With Risk of Major Coronary Events in the Anglo-Scandinavian Cardiac Outcomes Trial-Lipid Lowering Arm. JAMA Intern Med 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Nordestgaard BG, Benn M, Schnohr P, Tybjaerg-Hansen A: Nonfasting triglycerides and risk of myocardial infarction, ischemic heart disease, and death in men and women. Jama-Journal of the American Medical Association 2007, 298(3):299–308. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Figure 1: Mean MetS Components by Visit and Fasting Status. Data shown reflect levels of MetS components at each of the four visits evaluated by timing of a participant’s non-fasting status.