This Article Abstract Full Text (PDF) All Versions of this Article: 92/10/3780    most recent Author Manuscript (PDF) Submit a related Letter to the Editor Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Christen, A. Articles by Godsland, I. F. Search for Related Content PubMed PubMed Citation Articles by Christen, A. Articles by Godsland, I. F. Related Collections Lipid Cardiovascular Endocrinology Diabetes and Insulin Metabolism

Rate of Change and Instability in Body Mass Index, Insulin Resistance, and Lipid Metabolism as Predictors of Atherosclerotic Vascular Disease

Endocrinology and Metabolic Medicine, Division of Medicine, Faculty of Medicine, Imperial College School of Science, Technology and Medicine, London W2 1NY, United Kingdom

Address all correspondence and requests for reprints to: Ian F. Godsland, Ph.D., Endocrinology and Metabolic Medicine, Imperial College Faculty of Medicine, St. Mary’s Hospital Mint Wing, Praed Street, London W2 1NY, United Kingdom. E-mail i.godsland{at}ic.ac.uk.

	Abstract

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Context: By definition, levels of metabolic risk factors predict atherosclerotic vascular disease, but the effects of long-term adverse change and instability remain underresearched.

Objective: Our objective was to quantify long-term rates of change and instability in risk factors and relate these measures to clinical atherosclerotic vascular disease outcomes.

Design and Setting: We conducted a prospective cohort study with unmatched and age- and follow-up-matched control analyses at a teaching hospital day ward.

Participants: Participants included 465 predominantly healthy white males in an occupational cohort who had undergone repeated metabolic risk factor measurements (mean observation period 11.6 yr, range 2–28 yr), 62 of whom developed clinical atherosclerotic vascular disease.

Main Outcome Measures: Rate of change and instability in metabolic risk factor levels were quantified in each individual by linear regression with time and evaluated as predictors of atherosclerotic vascular disease and coronary and cerebrovascular disease separately.

Results: As expected, baseline and/or mean follow-up measures of established risk factors relating to blood pressure, lipid metabolism, and subclinical inflammation were significant predictors. Predictors independent of baseline and mean follow-up levels, confirmed in matched and unmatched analyses, were 1) for atherosclerotic vascular disease, instability in weight (cases vs. controls: 2.9 vs. +2.5%); 2) for coronary heart disease, instability in body mass index (3.0 vs. +2.3%), a decline (–0.041 vs. –0.011 per decade) and instability (19.1 vs. 14.6%) in the high-density lipoprotein/non-high-density lipoprotein cholesterol ratio, declining erythrocyte sedimentation rate, and increasing uric acid; and 3) for cerebrovascular disease, a decline in insulin sensitivity (–0.394 vs. 0.324 per decade).

Conclusions: Within an individual, long-term change in metabolic risk factors, as well as their absolute levels, can be important in atherosclerotic vascular disease.

	Introduction

Top Abstract Introduction Subjects and Methods Results Discussion References

 
BY DEFINITION, LEVELS of metabolic risk factors predict atherosclerotic vascular disease. However, the rate at which they change could also be important, with, for example, faster deterioration in individuals who will develop clinical atherosclerotic vascular disease. Instability independent of overall long-term change could also be important, possibly resulting in repeating cycles of vascular damage that exceed the capacity of existing repair mechanisms.

To evaluate relationships between long-term rate of change and instability in risk factor levels and clinical onset of atherosclerotic vascular disease, successive measurements are necessary before diagnosis, and there appear to have been few such studies. Published information relates mainly to weight cycling (1, 2, 3, 4, 5, 6, 7, 8, 9, 10) and blood pressure variation (11, 12), and results are inconsistent.

In the Heart Disease and Diabetes Risk Indicators in a Screened Cohort study, risk factor measurements relating to carbohydrate and lipid metabolism and subclinical inflammation were repeatedly recorded. To explore the hypothesis that long-term rate of change or instability in risk factors relates to the development of atherosclerotic vascular disease, we have evaluated individual baseline and mean follow-up values, overall long-term rate of change, and instability in risk factor measures during follow-up in 403 controls and in 62 men, free of atherosclerotic vascular disease and diabetes at baseline, who were subsequently diagnosed with atherosclerotic vascular disease. Subanalyses restricted to coronary heart disease and cerebrovascular disease were also undertaken.

	Subjects and Methods

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Design

The Heart Disease and Diabetes Risk Indicators in a Screened Cohort study is an open, prospective cohort study of risk factors for cardiovascular disease and diabetes. It derives from a health screening program for senior staff of a group of companies and was designed to detect and evaluate metabolic risk factors for cardiovascular disease and diabetes. Interventions were limited to dietary and lifestyle advice. The screening program lasted for 29 yr, during which 1278 individuals (including 1190 white males) received a range of measurements at their first and subsequent visits, generally spaced at 2- to 3-yr intervals. The study received local ethics committee approval, and each participant gave their written informed consent.

Procedures

Risk factor evaluations were carried out in a metabolic day ward, as previously described (13). In brief, subjects attended after an overnight fast. Systolic and diastolic blood pressures (SBP and DBP) were measured with a mercury sphygmomanometer, and blood was taken for measurement of fasting plasma glucose and insulin concentrations, serum total cholesterol and triglyceride concentrations, and liver function and hematological measures, including erythrocyte sedimentation rate and white blood cell count. From 1978 onward, serum high-density lipoprotein (HDL) cholesterol concentration was measured. An additional blood sample for glucose and insulin concentrations was taken. The participant then underwent an oral glucose tolerance test glucose load 1 g/kg body weight) with sampling for measurement of plasma glucose and insulin concentrations at 30, 60, 90, 120, 150, and 180 min. From 1986 onward, the oral glucose tolerance test was no longer used, but fasting glucose and insulin measurements continued.

Laboratory determinations

Routine hematological and biochemical parameters were measured centrally by standard laboratory methods. Glucose, insulin, lipids, and lipoproteins were measured in our own laboratory. Plasma glucose was determined on fresh samples using glucose oxidase-based procedures (14, 15). Plasma insulin was measured by RIA (16) on samples stored at –20 C. Serum total cholesterol and triglycerides were measured by chemical (17, 18) or enzymatic (19, 20) assays. Concentrations of HDL were measured after separation by preparative ultracentrifugation up to 1980 and then by sequential precipitation with heparin/manganese ions thereafter (21).

Quality control was monitored with pooled, frozen plasma samples and lyophilized sera and by participation in national schemes. Particular attention was given to maintaining long-term continuity of measurement with replicate assay of previously analyzed frozen samples and extensive comparisons when there was any change in assay methodology (e.g. Ref. 22).

Ascertainment of atherosclerotic vascular disease

Deaths from atherosclerotic vascular disease were identified up to January 2007 through the United Kingdom Office of National Statistics National Health Service Central Register. Information on atherosclerotic vascular disease morbidity derived from 1) ongoing clinical follow-up of those still participating in the health program, and 2) a questionnaire-based clinical follow-up, undertaken 17 yr into the project. Cases of atherosclerotic vascular disease and the date of its first recorded appearance were confirmed by agreement between a physician and a cardiologist. Atherosclerotic vascular disease was considered to have been present if myocardial infarction, coronary artery disease, or coronary artery thrombosis, cerebrovascular incident, stroke, peripheral vascular disease, or aortic aneurysm was recorded as a cause of death. Nonfatal coronary artery disease was identified on the basis of recorded nonfatal myocardial infarction (characteristic symptomatology, enzyme or electrocardiogram changes), coronary artery bypass graft, angioplasty, positive exercise electrocardiogram, sustained consistent history of angina with use of anti-anginal medication, or heart failure of probable atherosclerotic origin. Cerebral, peripheral, and other nonfatal atherosclerotic vascular disease was determined from the patients’ records or follow-up questionnaire responses.

Analytical design

Cases were assigned an index date corresponding to the first occasion on which clinical atherosclerotic vascular disease was detected. For controls, the index date was the most recent time at which the subject was known to be free of atherosclerotic vascular disease. To evaluate variability in risk factors before onset of clinical atherosclerotic vascular disease, individuals were selected who had no evidence of atherosclerotic vascular disease or diabetes at their first visit and at least three visits for risk factor evaluation, with their first visit at least 5 yr before their index date. The number of individuals qualifying for inclusion was 465 (with 2616 visit records) of which 62 (385 visits) were cases who developed clinical atherosclerotic vascular disease.

Because this was an open cohort study, it was not practical to rigidly control entry and exit of subjects from the cohort or the timing of visits. There were also, as mentioned, periods when some measurements were not available. These irregularities could lead to spurious intergroup differences or obscure genuine effects. Three differently structured datasets were, therefore, selected from the data available, each one designed to address an aspect of these problems.

Analysis A matched controls to cases according to age and number of follow-up visits. Matching by age was to within 5 yr of age at first and last visits, and there was an exact match in the number of follow-up visits. For controls matched to cases whose first clinical evidence of atherosclerotic vascular disease was morbidity rather than mortality, matching was also by age at index date to ensure positive confirmation that the matched controls were free of atherosclerotic vascular disease at comparable ages. This criterion was relaxed for controls matched to cases whose first evidence of atherosclerotic vascular disease was death (n = 18) because, in these cases, the generally long duration between final visit and death limited the number of controls available for matching to age at index date. A systematic approach to selection of controls was used, which established a maximum of three matched controls per case.

Analysis B employed all the data available for qualifying cases and controls. Although providing for strict comparability in age and frequency of follow-up visits, or maximizing use of the data, analyses A and B did not take into account periods when some measurements were not available. In analysis C, the first three recorded observations for each variable were selected for analysis. This ensured that the number of measurements contributing to the derivation of change and instability measures was the same in each subject. After selection of data for inclusion, analyses A, B, and C were each analyzed identically, as described below.

Data analysis

The variables considered were body mass index (BMI); SBP and DBP; fasting serum total, HDL, and low-density lipoprotein (calculated by the Friedewald method) cholesterol and triglycerides; the HDL/non-HDL cholesterol ratio; fasting plasma glucose and insulin (both calculated as the mean of the two fasting measurements); oral glucose tolerance test plasma glucose and insulin responses [quantified as area under the curve, as previously described (13)]; insulin resistance [calculated from fasting glucose and insulin concentrations using homeostasis model assessment insulin resistance (23)]; insulin sensitivity [calculated from fasting and oral glucose tolerance test glucose and insulin concentrations by the method of Matsuda and De Fronzo (24)]; WBC; erythrocyte sedimentation rate; and uric acid. Cigarette smoking was categorized as never smoked or ex-smoker (code 0), fewer than five cigarettes/d (code 1), five to 14 cigarettes/d (code 2), 15–24 cigarettes/d (code 3), or more than 24 cigarettes/d (code 4). Alcohol intake was expressed as never (0), light irregular (1), less than 28 U/wk (2), 28–56 U/wk (3), and more than 56 U/wk (4) (a unit of alcohol approximates to 10 ml or 8 g of pure ethanol). Exercise habit was expressed as none (0), moderate (1), or aerobic (2). As mentioned, measurements were incomplete for some variables. These were fasting plasma insulin (missing for 19% of visits for analysis A and 26% of visits for analyses B and C), oral glucose tolerance test glucose (31 and 37%), oral glucose tolerance test insulin (36 and 42%), and HDL cholesterol (12 and 14%).

Statistical analyses were undertaken using STATA 8 (Stata Corp., College Station, TX). Triglyceride and insulin measures were logarithmically transformed to normalize their distributions. To minimize the effects of any undetected drift in laboratory methodology over the 29-yr period of data collection, variation by 3-yr time period in each variable, independent of age, BMI, smoking, alcohol, exercise, and 3-yr recruitment cohort was quantified in the full Heart Disease and Diabetes Risk Indicators in a Screened Cohort dataset using a random-effects model with subject identification number as grouping variable and date of measurement as time variable. The coefficients for each 3-yr time period (taken as representative of method variation) were used to construct a time/method variation-corrected dataset. This dataset was then used in the analyses described below.

For each individual, four different derived measures of risk factor status were considered.

Level at first observation. Evaluation of first observation rather than first visit data ensured maximum representation of participants, given that there was missing data for some participants at first visit.

Mean follow-up level. The mean of all follow-up measurements included in the analysis for a given variable for a given individual provided a measure of overall risk factor exposure.

Rate of change. For each subject, measurements for a given variable were regressed against time before index date to give a regression line slope that quantified overall long-term rate of change per decade in the variable in that subject (we did not attempt to distinguish different patterns of change, such as continuous change or sustained change after a step in risk factor levels, which would have required more intensive and regular follow-up).

Instability. The regression analysis for the rate of change also provided a measure of variability about the regression line. This was the root mean square error, which quantified individual instability and was expressed as the percentage of the individual’s overall mean risk factor level.

First observation, mean, and change measures were considered for individuals with a minimum of two observations recorded for a given variable, this being the minimum necessary to derive a mean and a slope. Instability was considered for individuals with a minimum of three observations.

A survival analysis was then undertaken using the Cox proportional hazards model to evaluate each of the four derived measures for each variable as a predictor of atherosclerotic vascular disease. Survival time was taken as time from first visit to first evidence of atherosclerotic vascular disease for the cases and, for the controls, to the most recent occasion on which subjects were positively known to have been free of a diagnosis of atherosclerotic vascular disease. One exception to this was in the matched control analysis for controls matched to those whose first evidence of atherosclerotic vascular disease was death certificate-recorded mortality from atherosclerotic vascular disease. Their survival time was taken to the date at which mortality was most recently reviewed, namely January 1, 2007.

Where two or three of the analyses A, B, and C demonstrated a variable to be a significant (P < 0.05) predictor of atherosclerotic vascular disease, a probable significant effect was inferred. Where only one analysis returned a significant effect, this was disregarded. Independence of significant predictors was assessed using multivariate Cox proportional hazards modeling. Two repeat subanalyses were also undertaken, each restricted to those who developed coronary heart disease or cerebrovascular disease.

	Results

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Of the 62 atherosclerotic vascular disease cases, 32 developed coronary heart disease, 25 cerebrovascular disease, and three symptomatic peripheral vascular disease, and two died of an aortic aneurysm. On occasions during their follow-up, 16% of cases and 8% of controls were taking antihypertensive therapy. Equivalent figures for lipid-lowering agents were 3 and 2%, respectively, and for uric acid-lowering agents, 6 and 4%. Five or more cigarettes/d were smoked by 23% of cases and 14% of controls, more than 28 U of alcohol consumed/wk by 44% of cases and 39% of controls, and regular aerobic exercise taken by 24 of cases and 16% of controls. Age and duration and frequency of follow-up characteristics for cases and matched and unmatched controls are shown in Table 1.

	Discussion

Top Abstract Introduction Subjects and Methods Results Discussion References

Of these variables, there appears to be substantial published information only on variability in weight, generally with regard to repeated gain and loss in weight. In accord with our findings, a positive association between such weight cycling and coronary heart disease events has been reported (1, 2, 3, 4, 5, 7), and it has been suggested that lower HDL cholesterol concentrations might account for this (5). We found the association between coronary heart disease and instability in BMI was affected by HDL cholesterol but not eliminated. It is noteworthy that controls rather than cases tended to show the most rapid increase in weight during follow-up and, in multivariate analysis, rate of change in weight was a negative predictor of both atherosclerotic vascular disease and cerebrovascular disease. There is evidence from one other prospective study for an association between weight loss and increased total mortality (10). This could be explained by comorbidity, although our cohort comprised predominantly healthy, employed men. Moreover, a separate analysis of reported diagnoses (not shown) gave little indication of differences in comorbidity, apart from the expected risk factor differences. One exception was a 15% prevalence of varicose veins in the cases before atherosclerotic vascular disease compared with 5% in the controls (² significance P = 0.02), which suggests some underlying vascular pathology before atherosclerotic vascular disease. Cigarette smoking in the cases could be another factor contributing to their decline in weight, but we found no association between weight change and smoking characteristics (not shown).

Cigarette smoking, an adverse lipid risk factor profile and inflammatory indices predicted coronary heart disease, and the importance of lipid metabolism in coronary heart disease appeared to extend to rates of change and instability in HDL metabolism. The long-term decline we observed in erythrocyte sedimentation rate before clinical onset of coronary heart disease appears paradoxical, but low erythrocyte sedimentation rate correlates with hemodynamic abnormalities in heart failure (25), so our observation is consistent with a progressing vascular pathology in the cases before clinical onset of coronary heart disease. Also consistent is the long-term increase in uric acid, given uric acid’s association with inflammation (26). In those who would develop cerebrovascular disease, the independence of declining weight and insulin sensitivity as predictors of cerebrovascular disease and the increase in their significance in multivariate analysis suggests the existence of subgroups showing a long-term decline in weight or an increase in insulin resistance.

Repeated acquisition of a broad range of risk factor measurements in a group sufficiently large for prospective analysis represents a research challenge, and our study has limitations. Repeated measurements of indices of atherosclerosis (e.g. carotid intima media thickness) were not carried out in the Heart Disease and Diabetes Risk Indicators in a Screened Cohort. Consequently, the variation in risk factors we detected in the cases might have been a consequence of developing subclinical atherosclerosis, as suggested by several of our observations (see above). Moreover, our focus on onset of clinically apparent atherosclerotic vascular disease means that atherosclerotic plaque instability is likely to have been a factor in defining our end points, and risk factor changes more proximal to the initial symptoms or event than those we measured could have played a role in this respect. Another limitation was that controls were matched for duration, timing, and frequency of follow-up rather than range of calendar dates covered. Ascertainment of cases was not 100% because, without an updated clinical follow-up, only cases already identified or deaths ascertained from the Central Register were selected for analysis. Therefore, some cases may not have been identified, and some of the controls might have developed atherosclerosis after their index data. These factors would have diminished the power of the study to detect significant effects, but we were, nevertheless, able to detect as significant a broad range of well-established risk factors for atherosclerosis, and the ability of our study to detect these effects lends confidence to our novel findings. It is also noteworthy that relatively few individuals commenced drug therapies during follow-up. It is unlikely that a comparable study could be undertaken now without a significant impact of lipid- and blood pressure-lowering treatment.

Whether identification of novel risk factors will contribute further to the assessment of cardiovascular risk has been questioned (27), but our findings suggest that novel characteristics of established risk factors may be worth exploring and of value in elucidating the true importance, or otherwise, of novel risk factors. In this respect, the issue of regression dilution bias is likely to be significant, according to which relative risks of disease predicted by a putative true measure of the risk factor level will be underestimated if a single risk factor measurement subject to measurement imprecision and intra-individual variation is entered in the analysis. Rather than quantifying regression dilution bias, our intention was to evaluate whether features of long-term variation in risk factor levels per se were associated with disease risk. Our findings, nevertheless, exemplify how a long-term, mean risk factor level may relate more closely to risk than a single baseline measure (mean but not first observation levels of three risk factors predicted coronary heart disease) and highlight the considerable differences in variability, and hence regression dilution, that exist between different risk factors. Given the marked differences in long-term rate of change we detected, our findings also raise questions about what might constitute a true risk factor level.

In summary, we have identified instability in weight and a long-term decline in erythrocyte sedimentation rate as significant predictors of clinical atherosclerotic vascular disease. These were also important in those who developed coronary heart disease, as were a long-term decline and instability in the HDL/non-HDL cholesterol ratio and a long-term increase in uric acid. In those who developed cerebrovascular disease, a long-term increase in insulin resistance independent of changes in weight was clearly apparent. These findings were made in a selected group with relatively high BMI and alcohol intake and need to be confirmed in other groups. Nevertheless, they establish the principle that variation in risk factor levels, as well as risk factor levels per se, may be important in atherosclerosis.

	Acknowledgments

 
The late Professor Victor Wynn initiated and established the Heart Disease and Diabetes Risk Indicators in a Screened Cohort study. Since it began, the study has been sustained by many clinical, scientific, technical, nursing, and administrative staff, to each of whom we extend our thanks.

	Footnotes

 
The study was funded by the Atherosclerosis Research Trust, the Heart Disease and Diabetes Research Trust, and the Cecil Rosen Foundation. I.F.G. is supported by the Heart Disease and Diabetes Research Trust.

Disclosure Statement: The authors have no interests to declare.

First Published Online July 31, 2007

Abbreviations: BMI, Body mass index; DBP, diastolic blood pressure; HDL, high-density lipoprotein; SBP, systolic blood pressure.

Received October 17, 2006.

Accepted July 24, 2007.

	References

Top Abstract Introduction Subjects and Methods Results Discussion References

 

1. Hamm P, Shekelle RB, Stamler J 1989 Large fluctuations in body weight during young adulthood and 25-year risk of coronary death in men. Am J Epidemiol 129:312–318[Abstract/Free Full Text]
2. Lissner L, Odell PM, D’Agostino RB, Stokes J, Kreger BE, Belanger AJ, Brownell KD 1991 Variability of body weight and health outcomes in the Framingham population. N Engl J Med 324:1839–1844[Abstract]
3. Blair SN, Shaten J, Brownell K, Collins G, Lissner L 1993 Body weight change, all-cause mortality and cause-specific mortality in the Multiple Risk Factor Intervention Trial. Ann Intern Med 119:749–757[Abstract/Free Full Text]
4. Folsom AR, French SA, Zheng W, Baxter JE, Jeffrey RW 1996 Weight variability and mortality: the Iowa Women’s Health Study. Int J Obes Relat Metab Disord 20:704–709[Medline]
5. Olsen MB, Kelsey SF, Bittner V, Reis SE, Reichek N, Handberg EM, Bairey Merz CN 2000 Weight cycling and high-density lipoprotein cholesterol in women: evidence of an adverse effect. J Am Coll Cardiol 36:1565–1571[Abstract/Free Full Text]
6. Yarnell JW, Patterson CC, Thomas HF, Sweetnam PM 2000 Comparison of weight in middle age, weight at 18 years, and weight change between, in prediction subsequent 14 year mortality and coronary events. J Epidemiol Community Health 54:344–348[Abstract/Free Full Text]
7. Cho E, Manson JE, Stampfer MJ, Solomon CG, Colditz GA, Speizer FE, Willett WC, Hu FB 2002 A prospective study of obesity and risk of coronary heart disease among diabetic women. Diabetes Care 25:1142–1148[Abstract/Free Full Text]
8. Wannamethee SG, Shaper AG, Walker M 2002 Weight change, weight fluctuation and mortality. Arch Intern Med 162:2575–2580[Abstract/Free Full Text]
9. Wannamethee SG, Shaper AG, Walker M 2005 Overweight and obesity and weight change in middle aged men: impact on cardiovascular disease and diabetes. J Epidemiol Community Health 59:134–139[Abstract/Free Full Text]
10. Soerensen TIA, Rissanen A, Korkeila M, Kaprio J, Iacono JM 2005 Intention to lose weight, weight changes, and 18-y mortality in overweight individuals without co-morbidities. PLOS Med 6:510–520
11. Tervahauta M, Pekkanen J, Enlund H, Nisinen A 1994 Change in blood pressure and 5-year risk of coronary heart disease among elderly men: the Finnish cohorts of the Seven Countries Study. J Hypertens 12:1183–1189[Medline]
12. Menotti A, Lanti M, Kafatos A, Nissinen A, Dontas A, Nedeljkovic S, Kromhout D 2004 The role of a baseline casual blood pressure measurement and of blood pressure changes in middle age in prediction of cardiovascular and all-cause mortality occurring late in life: a cross-cultural comparison among the European cohorts of the Seven Countries Study. J Hypertens 22:1683–1690[CrossRef][Medline]
13. Godsland IF, Leyva F, Worthington M, Walton C, Stevenson JC 1998 Associations of smoking, alcohol and physical activity with risk factors for coronary heart disease and diabetes in the first follow-up cohort of the HDDRISC Study (HDDRISC-1). J Intern Med 244:33–41[CrossRef][Medline]
14. Cramp DG 1967 New automated method for measuring glucose by glucose oxidase. J Clin Pathol 20:910–912[Abstract/Free Full Text]
15. Trinder P 1969 Determination of blood glucose using an oxidase-peroxidase system with non-carcinogenic chromogen. J Clin Pathol 22:158–161[Abstract/Free Full Text]
16. Albano JDM, Ekins RP, Maritz G, Turner RC 1972 A sensitive, precise radioimmunoassay of serum insulin relying on charcoal separation of bound and free hormone moieties. Acta Endocrinol 70:487–509[Medline]
17. Robertson G, Cramp DG 1970 An evaluation of cholesterol determinations in serum and serum lipoprotein fractions by a semi-automated fluorimetric method. J Clin Pathol 23:243–245[Abstract/Free Full Text]
18. Cramp DG, Robertson G 1968 The fluorimetric assay of triglyceride by a semiautomated method. Anal Biochem 25:246–251[CrossRef][Medline]
19. Seidel J, Hagele E, Ziagenhorn J, Wahlefield A 1983 Reagent for the enzymatic determination of serum total cholesterol with improved lipolytic efficiency. Clin Chem 29:1075–1080[Abstract/Free Full Text]
20. Bucolo G, David H 1973 Quantitative determination of serum triglycerides by the use of enzymes. Clin Chem 19:476–482[Abstract]
21. Warnick GR, Albers JJ 1978 A comprehensive evaluation of the heparin-manganese precipitation procedure for estimating high density lipoprotein cholesterol. J Lipid Res 19:65–76[Abstract]
22. Khan SR, Elkeles RS, Niththyananthan R, Seed M, Wynn V 1982 Analysis of high density lipoprotein subclasses by a precipitation procedure: comparison with preparative ultracentrifuge. Clin Chim Acta 120:49–55[CrossRef][Medline]
23. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC 1985 Homeostasis model assessment: insulin resistance and ß-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28:412–419[CrossRef][Medline]
24. Matsuda M, DeFronzo RA 1999 Insulin sensitivity indices obtained from oral glucose tolerance testing. Diabetes Care 22:1462–1470[Abstract/Free Full Text]
25. Haber H, Leavy J, Kessler P, Kukin M, Gottlieb S, Packer M 1991 The erythrocyte sedimentation rate in congestive heart failure. N Engl J Med 324:353–358[Abstract]
26. Leyva F, Anker SD, Godsland IF, Teixeira M, Hellewell PG, Kox WJ, Poole-Wilson PA, Coats AJ 1998 Uric acid in chronic heart failure: a marker of chronic inflammation. Eur Heart J 19:1814–1822[Abstract/Free Full Text]
27. Folsom AR, Chambless LE, Ballantyne CM, Coresh J, Heiss G, Wu KK, Boerwinkle E, Mosley JR TH, Sorlie P, Diao G, Sharrett AR 2006 An assessment of incremental coronary risk prediction using C-reactive protein and other novel risk markers: the atherosclerosis risk in communities study. Arch Intern Med 166:1368–1373[Abstract/Free Full Text]

This article has been cited by other articles:

				R. Amin, L. Anthony, V. Somers, M. Fenchel, K. McConnell, J. Jefferies, P. Willging, M. Kalra, and S. Daniels Growth Velocity Predicts Recurrence of Sleep-disordered Breathing 1 Year after Adenotonsillectomy Am. J. Respir. Crit. Care Med., March 15, 2008; 177(6): 654 - 659. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 92/10/3780    most recent Author Manuscript (PDF) Submit a related Letter to the Editor Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Christen, A. Articles by Godsland, I. F. Search for Related Content PubMed PubMed Citation Articles by Christen, A. Articles by Godsland, I. F. Related Collections Lipid Cardiovascular Endocrinology Diabetes and Insulin Metabolism