{"slug":"triglyceride-hdl-ratio","name":"Triglyceride-to-HDL ratio","aliases":["TG/HDL-C ratio","TG:HDL","Triglyceride/HDL cholesterol ratio","Atherogenic index of plasma (log TG/HDL-C)"],"kind":"biomarker","system":"metabolomics","unit":"ratio (mg/dL to mg/dL)","unitNote":"Fasting triglycerides (mg/dL) divided by HDL cholesterol (mg/dL). Because triglycerides convert at about 88.5 mg/dL per mmol/L and HDL cholesterol at about 38.7, the same blood gives a ratio about 2.3 times lower in mmol/L: a ratio of 3.0 in mg/dL is about 1.3 in mmol/L. Cutoffs from papers that used mmol/L must be converted before they are compared, and every cutoff on this page states the unit its source used.","answer":"The triglyceride-to-HDL ratio is a free calculation from any lipid panel that tracks insulin resistance and small, dense LDL in people of European, Hispanic and some Asian ancestry, where cutoffs of about 3.0 to 3.5 (mg/dL) identified insulin resistance with roughly two-thirds sensitivity and specificity; it does not work in Black Americans, whose triglycerides often stay normal while they are insulin resistant, and it adds little once ApoB and fasting insulin have been measured. No laboratory prints it and no guideline sets a cutoff.","summary":"Dividing fasting triglycerides by HDL cholesterol gives a single number for the lipid pattern of insulin resistance: VLDL overproduction raises the top and strips the bottom. Stanford studies against direct measures put the cutoff near 3.0 in overweight adults and 3.5 in a general sample (mg/dL), and the ratio predicted heart attacks, coronary extent and mortality in cohorts. It fails in people of African ancestry, moves with meals and assays, and is a prompt to measure ApoB and insulin rather than a verdict.","whatItIs":[{"text":"The ratio puts the two lipid signatures of insulin resistance into one number. In 18 non-diabetic, moderately overweight men sorted by a direct measure of insulin action (steady-state plasma glucose), the most insulin-resistant third had higher fasting insulin, higher fasting triglycerides and lower HDL cholesterol than the other two groups, with no difference in body mass index, waist-to-hip ratio or treadmill fitness. Triglycerides and HDL move in opposite directions for the same reason, so dividing one by the other amplifies the signal; it also doubles the noise, since both inputs vary from day to day.","cites":["laws1992","bookstein1990"],"grade":"C"},{"text":"The cutoffs most often quoted come from Stanford. In 258 non-diabetic overweight volunteers measured by the insulin suppression test, the triglyceride-to-HDL ratio was among the three most useful fasting markers of insulin resistance, with a cutoff of 3.0 in mg/dL units giving 64 percent sensitivity and 68 percent specificity, about the same as the Adult Treatment Panel III metabolic syndrome criteria (52 and 85 percent); the authors suggested it as the practical choice in the absence of a standardized insulin assay. In 449 apparently healthy adults, with LDL particle size measured in 1,135, a ratio of 3.5 or more was the best routine predictor of both insulin resistance and small, dense LDL, and it predicted the LDL pattern more accurately than it predicted insulin resistance.","cites":["mclaughlin2003","mclaughlin2005"],"grade":"C"},{"text":"Larger samples find a weaker, still real, correlation. In 3,014 Framingham Offspring participants, the ratio correlated 0.46 with HOMA-IR after adjustment for age and sex and identified the top quartile of HOMA-IR with an area under the curve of 0.745, slightly better than the total-to-HDL cholesterol ratio (0.707); the authors called it an imperfect surrogate for insulin resistance and its coronary risk. In 2,652 non-diabetic US adults from a national survey, the ratio predicted hyperinsulinemia with areas under the curve of 0.74 to 0.77 across three ancestry groups, and adults above a cutoff of 3.5 were 2.16 times as likely to have a high fasting insulin.","cites":["kannel2008","li2008"],"grade":"B"}],"origins":[{"text":"The ratio has no evolutionary history of its own; it is a reading on two ancient transport systems and on what insulin resistance does to the traffic between them. Apolipoprotein B, the protein that builds every VLDL particle, belongs to a superfamily of large lipid transfer proteins that also includes insect apolipophorin, the egg-yolk protein vitellogenin and the microsomal triglyceride transfer protein that loads lipid onto apoB inside the cell; shared motifs, identical residues and ancestral exon boundaries show the four descend from one ancestral lipid-transfer module, and comparative work places that family in the earliest animals. The high-density particle is the older design: insects move fat on a high-density lipophorin that is loaded, unloaded at the tissues and reloaded as a reusable shuttle, where mammalian LDL is swallowed whole by its receptor and destroyed. Triglyceride on the top of the ratio and HDL on the bottom are, at the level of the proteins, the two halves of one solution to moving fat through water.","cites":["babin1999","vanderhorst2009"],"grade":"C"},{"text":"What drives the two numbers apart is the liver. In insulin-resistant states the fundamental defect is overproduction of large, triglyceride-rich VLDL particles, which sets off a sequence of lipoprotein changes: more remnant particles, smaller LDL and lower HDL cholesterol, a pattern that precedes the diagnosis of type 2 diabetes by several years. With more triglyceride-rich particles in circulation, cholesteryl ester transfer protein exchanges triglyceride from VLDL for cholesteryl ester from HDL and LDL; the triglyceride-enriched HDL becomes a substrate for hepatic lipase, which lipolyzes it into smaller particles that are cleared faster, so HDL cholesterol and apoA-I fall. The same exchange and the same enzyme act on LDL that lingers in plasma for about 5 days, generating the small, dense LDL that is retained more readily by the artery wall. The ratio rises because one process pushes its numerator up and its denominator down at the same time.","cites":["adiels2008","rashid2003","packard2003","berneis2002"],"grade":"C"},{"text":"That shared mechanism is why a ratio of two cheap numbers stands in for particle measurements that laboratories rarely run. In 1,433 subjects from 35 cohorts spanning cord blood, children, healthy adults and patients with hypertension, diabetes, dyslipidemia and coronary disease, the logarithm of the ratio correlated 0.803 with the cholesterol esterification rate in apoB-depleted plasma, an indirect measure of lipoprotein particle size, and inversely with LDL size (minus 0.776) in 35 normal subjects; the authors named the log ratio the atherogenic index of plasma. In 1,350,908 US adults referred for lipoprotein ultracentrifugation, with a median ratio of 2.2, remnant lipoprotein cholesterol, non-HDL cholesterol and LDL density all rose steadily across increasing ratios, and about 47 percent of the ratio's variance was attributable to remnant cholesterol. On the shelf, the ratio sits downstream of fasting insulin and HOMA-IR and beside ALT, as a marker of a liver exporting more fat than the periphery can clear, and upstream of ApoB, which counts the particles directly.","cites":["dobiasova2001","quispe2015","adiels2008"],"grade":"C"}],"labRange":{"typical":"No laboratory prints this ratio; it is calculated from the triglyceride and HDL cholesterol on the same panel. The 2001 US cutoffs for those two numbers (triglycerides under 150 mg/dL, HDL at least 40) imply a ratio under 3.8, which is the only 'laboratory' line that can be drawn","claims":[{"text":"The 2001 National Cholesterol Education Program Adult Treatment Panel III classed triglycerides under 150 mg/dL as normal and 150 to 199 as borderline high, and defined low HDL cholesterol categorically as under 40 mg/dL, raised from under 35 in the previous panel; its metabolic syndrome criteria used triglycerides of 150 or more and HDL under 40 in men and under 50 in women. The panel set no goal for raising HDL because the evidence was insufficient, and it listed the causes of a low HDL as largely those of insulin resistance: elevated triglycerides, overweight and obesity, physical inactivity and type 2 diabetes. It never mentioned the ratio.","cites":["ncep2001"],"grade":"A"},{"text":"Dividing the panel's two thresholds gives 150 / 40, or 3.8, and one study tested that line directly: among 658 patients with LDL phenotype measured by nuclear magnetic resonance, a ratio of 3.8 put 79 percent of people with small, dense LDL (phenotype B) above it and 81 percent of people with large LDL (phenotype A) below it, in men and women alike. It is a derived line, not a laboratory's reference interval, and no laboratory reports it.","cites":["hanak2004","ncep2001"],"grade":"C"},{"text":"The published cutoffs for insulin resistance sit lower than 3.8 and move with sex and population: 3.0 in overweight US adults and 3.5 in a general US sample (both mg/dL); above 2.5 in women and above 3.5 in men among 1,102 women and 464 men of primarily European ancestry in Argentina, where those sex-specific lines identified insulin resistance with about 40 percent sensitivity and about 80 percent specificity; 2.8 in men and 1.9 in women for the metabolic syndrome among 1,932 non-diabetic Germans (area under the curve 0.817); and 2.27 in white obese youths. A review of 1,350,908 US lipid referrals listed the cutoffs then in circulation as 2.5 (men) and 2 (women), 3.75 (men) and 3 (women), and 3.5 for both. There is no agreed threshold.","cites":["mclaughlin2003","mclaughlin2005","salazar2012","vonbibra2017","giannini2011","quispe2015"],"grade":"C"}]},"evidence":{"headline":"Against direct measures of insulin action, a ratio of about 3.0 to 3.5 (mg/dL) separated insulin-resistant from insulin-sensitive adults with roughly two-thirds sensitivity and specificity in US samples of mostly European ancestry, and it did no better than chance in Black Americans. In cohorts, a high ratio carried about 1.4 times the cardiovascular events of a low one across 13 studies, predicted heart disease and all-cause mortality and new diabetes in 39,447 men, and was the strongest lipid correlate of coronary extent at angiography. Genetics say the risk travels with ApoB particle number, not with the ratio itself, and no trial has treated anyone to a ratio target.","scale":{"min":0,"max":8,"ticks":[0,1,2,3,4,5,6,7,8]},"bands":[{"label":"Under 3.8: both inputs inside the 2001 US cutoffs (triglycerides under 150, HDL 40 or more); no laboratory prints the ratio","low":0,"high":3.8,"tone":"lab","source":"NCEP ATP III 2001 thresholds, combined by Hanak 2004"},{"label":"Under 3.0 (mg/dL): below the cutoff for insulin resistance in overweight US adults; not valid in Black Americans","low":0,"high":3,"tone":"evidence","source":"McLaughlin 2003; Sumner 2005"},{"label":"Under 2.8: lowest quartile of Tehran men, metabolic syndrome in 3 percent","low":0,"high":2.8,"tone":"evidence","source":"Hadaegh 2009"},{"label":"Sex-specific cutoffs in adults of European ancestry: above 2.5 in women, above 3.5 in men","low":2.5,"high":3.5,"tone":"neutral","source":"Salazar 2012"},{"label":"3.5 and above: insulin resistance and small, dense LDL in a general US sample; higher heart and all-cause mortality and twice the diabetes incidence in men","low":3.5,"high":8,"tone":"caution","source":"McLaughlin 2005; Vega 2014"},{"label":"3.8 and above: 79 percent of people with small, dense LDL sat here","low":3.8,"high":8,"tone":"caution","source":"Hanak 2004"},{"label":"6.9 and above: top quartile of Tehran men, metabolic syndrome in 64 percent, 1.75 times the coronary risk","low":6.9,"high":8,"tone":"caution","source":"Hadaegh 2009"}],"claims":[{"text":"In a Boston case-control study of 340 adults under 76 with a first confirmed heart attack and 340 controls matched on age, sex and community, fasting triglycerides in the top quartile carried 6.8 times the risk of the bottom quartile before adjustment and 2.7 times after adjusting for HDL cholesterol; the ratio of triglycerides to HDL was the strongest lipid predictor, with 16.0 times the risk in the top quartile against the bottom (95 percent confidence interval 7.7 to 33.1). A case-control design with wide intervals, but the first signal that the two numbers read together say more than either alone.","cites":["gaziano1997"],"grade":"C"},{"text":"Prospective cohorts confirm the direction with smaller numbers. Among 1,824 Iranian men aged 40 and over followed a median of 6.5 years with 163 new coronary events, men in the top quartile of the ratio (6.9 or more) had 1.75 times the coronary risk of the lowest quartile (under 2.8) after adjustment for total cholesterol and the other risk factors, while triglycerides alone dropped out of the model; the metabolic syndrome was present in 63.6 percent of the top quartile and 3.0 percent of the bottom. Among 2,288 Spanish male factory workers (208 cases, 2,080 controls), the ratio was associated with a first coronary event in every category of body mass index (odds ratio 1.47), where LDL cholesterol was barely associated (1.01). Pooling 13 cohorts with 207,515 participants, the highest category of the ratio carried 1.43 times the cardiovascular events of the lowest (95 percent confidence interval 1.26 to 1.62), shrinking to 1.26 after correction for publication bias, with 1.08 times the risk per unit of the ratio; adjustment for other lipids did not change the result.","cites":["hadaegh2009","cordero2009","chen2022"],"grade":"A"},{"text":"In 39,447 men from the Cooper Center Longitudinal Study followed for 581,194 person-years, a ratio at or above 3.5 predicted coronary, cardiovascular and all-cause mortality after adjustment for established risk factors and non-HDL cholesterol, as well as or better than the metabolic syndrome did, and among 22,215 men the annual incidence of type 2 diabetes was 2 times higher above the cutoff than below it. The triglyceride-glucose index predicted diabetes equally well but mortality less well. A men-only cohort of mostly white, affluent Texans.","cites":["vega2014"],"grade":"B"},{"text":"In women, the ratio predicted death. Among 544 women referred for coronary angiography in the Women's Ischemia Syndrome Evaluation, 84 percent white, with ratios from 0.3 to 18.4 (median 2.2) and a median follow-up of 6 years, deaths and cardiovascular events rose across quartiles, and each unit increase in the log of the ratio carried a hazard ratio of 1.95 for all-cause mortality after adjustment for age, race, smoking, hypertension, diabetes and angiographic severity. For cardiovascular events the adjusted hazard ratio was 1.54 but fell short of significance once angiographic results were included. In 374 high-risk Brazilian patients at angiography, the ratio had the strongest association of any lipid with extensive coronary disease (odds ratio 2.01), and only the ratio and HDL were useful on receiver operating curves; total and LDL cholesterol were not. In an Argentine town, among 527 women and 269 men resurveyed after 9 years, a ratio above the sex-specific cutoffs predicted a first cardiovascular event as well as the metabolic syndrome did once age and sex were adjusted (hazard ratios 2.01 and 2.09).","cites":["bittner2009","daluz2008","salazar2013"],"grade":"B"},{"text":"The ratio predicts, but neither of its components is causal in the way the prediction implies. Genetic variants that raise HDL cholesterol do not lower heart-attack risk: a variant raising HDL by 0.14 mmol/L should have cut risk by 13 percent and cut it by none (odds ratio 0.99 in 20,913 cases and 95,407 controls), and a 14-variant HDL score gave 0.93 per standard deviation against 0.62 expected from observation. Genetically higher triglycerides do raise coronary risk, through the triglyceride-rich lipoproteins that carry them, in a model accounting for LDL and HDL cholesterol; in 17 studies with 62,199 participants the triglyceride score carried 1.62 times the risk per log unit, but that estimate fell to 1.01 once HDL, LDL and statin use were adjusted. The denominator of the ratio is a marker, not a cause, and the numerator's effect is carried by particle number.","cites":["voight2012","do2013","holmes2015"],"grade":"B"},{"text":"Once ApoB is measured, the ratio adds little to heart risk. In 654,783 people, triglyceride-lowering LPL variants and LDL-lowering LDLR variants lowered coronary risk identically per 10 mg/dL lower ApoB (odds ratios 0.771 and 0.773), and the association of triglycerides with coronary disease became null after adjusting for ApoB. Across 12 studies with 233,455 people and 22,950 events, ApoB was the most potent lipid marker of cardiovascular risk (standardized relative risk 1.43), non-HDL cholesterol intermediate (1.34) and LDL cholesterol least (1.25). The 2019 review of that evidence concludes that cholesterol enters the artery wall only inside ApoB particles, that smaller, cholesterol-depleted particles are trapped more readily than large ones, and that with narrow exceptions all ApoB particles are equally atherogenic, so the count is what matters. The ratio's value is as a free hint that the count is high and the particles small, which ApoB then confirms or refutes; as a measure of insulin resistance it competes with fasting insulin, not with ApoB.","cites":["ference2019","sniderman2011","sniderman2019"],"grade":"A"},{"text":"The ratio does not work in people of African ancestry. In 98 overweight African American adults measured by frequently sampled intravenous glucose tolerance testing, triglycerides and the ratio did not rise across tertiles of insulin sensitivity; the ratio's area under the curve was 0.56, meaning the true-positive rate nearly equalled the false-positive rate, and the Stanford lipid criteria found only 17 percent of the insulin-resistant. The reason is that triglycerides run lower in Black Americans at every level of insulin resistance: in 2,804 US adults from a national survey, mean triglycerides were 99 mg/dL in non-Hispanic Black, 140 in non-Hispanic white and 144 in Mexican American adults, and 64 percent of insulin-resistant Black adults had triglycerides under 130 against 31 and 36 percent of the other two groups. In 99 African American and 50 white healthy adults, the ratio identified a high HOMA-IR with an area under the curve of 0.770 in whites and 0.639 in African Americans; in 1,452 obese youths, a cutoff of 2.27 carried 6.0 times the odds of insulin resistance in 612 white children, while in 357 Hispanic and 483 African American children the curves were not significant and no cutoff could be set. One national analysis found the ratio tracked hyperinsulinemia in Black adults when the cutoff was lowered to 2.0 (area under the curve 0.75), so the disagreement is partly about where the line sits; against direct measures, the ratio has not been shown to work.","cites":["sumner2005","sumner2008","kimdorner2010","giannini2011","li2008"],"grade":"C"},{"text":"Where it does work, it works in lean people too, and in many populations. Among 118 normal-weight, euthyroid, non-diabetic adults in Lima, a ratio of 3 or more (mg/dL) was associated with 3.16 times the prevalence of insulin resistance by HOMA-IR and 2.36 times the prevalence of hyperinsulinemia after a glucose load. In three cohorts totalling 13,908 people from Canada and Iran, a ratio above the cohort's own optimal cutoff carried 6.00, 4.04 and 3.50 times the odds of the metabolic syndrome, least in the healthy population, identified insulin resistance by HOMA-IR, and tracked the severity of carotid atherosclerosis. These are cross-sectional snapshots with HOMA-IR, itself an estimate, as the reference.","cites":["pantoja2019","azarpazhooh2021"],"grade":"C"}]},"levers":[{"name":"Weight loss","effect":"lowers the ratio from both ends","claim":{"text":"Across 73 randomized trials with 32,496 adults with overweight or obesity, each kilogram lost through lifestyle change lowered triglycerides by 4.0 mg/dL and raised HDL cholesterol by 0.46 mg/dL at 12 months, so the ratio falls faster than either number alone. An older pooling of 70 diet studies adds a caution: while weight is still falling HDL dips by 0.007 mmol/L per kilogram and recovers to a gain of 0.009 only once weight has stabilized, so a ratio drawn mid-diet reads worse than it will.","cites":["hasan2020","dattilo1992"],"grade":"A"}},{"name":"Cutting free sugar, including fructose","effect":"lowers it, independent of weight","claim":{"text":"In 39 randomized trials, higher free-sugar intake raised triglycerides by 0.11 mmol/L (about 10 mg/dL) and HDL cholesterol by a negligible 0.02 mmol/L, with the clearest effect in trials that held calories and weight constant. The 2001 US panel listed high-carbohydrate diets (over 60 percent of energy) and excess alcohol among the causes of raised triglycerides and very high carbohydrate intake among the causes of low HDL.","cites":["temorenga2014","ncep2001"],"grade":"A"}},{"name":"Lower-carbohydrate eating","effect":"lowers it more than a low-fat diet, with an LDL trade-off","claim":{"text":"In 11 randomized trials of 6 months or longer with 1,369 participants, a low-carbohydrate diet lowered triglycerides 0.26 mmol/L (about 23 mg/dL) more than a low-fat diet and raised HDL cholesterol 0.14 mmol/L (about 5 mg/dL) more, moving both halves of the ratio the right way, but it also raised LDL cholesterol by 0.16 mmol/L (about 6 mg/dL). A ratio that improves while ApoB rises is not a better lipid profile, which is one more reason to read the ratio beside ApoB.","cites":["mansoor2016"],"grade":"A"}},{"name":"Aerobic exercise","effect":"lowers it modestly","claim":{"text":"Across 25 randomized trials, aerobic training raised HDL cholesterol by a mean of 2.53 mg/dL, with a minimum of about 120 minutes or 900 kcal a week needed to see any rise and each extra 10 minutes per session worth about 1.4 mg/dL. Pooling six randomized trials of aerobic exercise with a prudent diet in up to 559 adults, triglycerides fell 10.6 mg/dL, about 18 percent, against controls, while HDL did not change. The triglyceride side of the ratio responds more than the HDL side.","cites":["kodama2007","kelley2011"],"grade":"A"}},{"name":"Omega-3 fatty acids at prescription doses","effect":"lowers the triglyceride half","claim":{"text":"The American Heart Association's 2019 advisory puts the triglyceride reduction from 4 grams a day of EPA plus DHA at 30 percent or more in people with very high triglycerides, with a concurrent rise in LDL cholesterol in that group, and roughly comparable lowering for EPA-only and combined products in the 200 to 499 mg/dL range; in the largest trials non-HDL cholesterol and ApoB fell modestly. The advisory concerns pharmacological doses above 3 grams a day, not over-the-counter capsules.","cites":["skulasray2019"],"grade":"A"}},{"name":"Alcohol","effect":"lowers the ratio without lowering risk","claim":{"text":"In 44 interventional studies, alcohol raised HDL cholesterol by a pooled 0.094 mmol/L (about 3.6 mg/dL) with a dose-response trend and did not change fasting triglycerides, so moderate drinking makes the ratio look better. It does not make the person better: in 261,991 people of European descent, carriers of a variant that cuts drinking by 17.2 percent had lower coronary (odds ratio 0.90) and ischemic stroke risk (0.83) at every level of consumption. And alcohol with a fatty meal adds to the post-meal triglyceride peak by acutely inhibiting lipoprotein lipase, so a ratio drawn the morning after a dinner with wine reads worse than usual. A lever to understand, not to pull.","cites":["brien2011","holmes2014","vandewiel2012"],"grade":"A"}},{"name":"Dietary fiber","effect":"little direct effect on the ratio; large effect on the outcomes","claim":{"text":"In 23 randomized trials with 1,513 participants, adding fiber lowered total and LDL cholesterol but did not change triglycerides (mean difference 0.00 mmol/L) and very slightly lowered HDL (minus 0.03 mmol/L), so fiber on its own barely moves this ratio. The reason to eat it is elsewhere: across 185 prospective studies and 58 trials, the highest fiber consumers had 15 to 30 percent lower all-cause and cardiovascular mortality and incidence of coronary disease, stroke and type 2 diabetes, with the greatest risk reduction at 25 to 29 grams a day. A marker can stay still while the risk it stands for falls.","cites":["hartley2016","reynolds2019"],"grade":"A"}},{"name":"Sleep","effect":"short sleep worsens the physiology; the lipid data are thin","claim":{"text":"Six nights of 4 hours in bed lowered glucose tolerance and raised evening cortisol and sympathetic activity in 11 healthy young men, which is the insulin-resistant state the ratio tracks, but the study measured glucose, not lipids. A meta-analysis of 6 prospective studies with 30,033 participants found no association between short sleep and new dyslipidemia (relative risk 1.01), with heterogeneity and publication bias. Sleep belongs on this list for the mechanism, not for a demonstrated change in the ratio.","cites":["spiegel1999","kruisbrink2017"],"grade":"C"}}],"caveats":[{"text":"Ancestry comes first. In Black Americans the ratio did no better than chance against a direct measure of insulin action, because triglycerides stay low in that population even when insulin resistance is present; the cutoffs on this page were derived in samples of mostly European ancestry and transfer to Hispanic and some Asian samples, but not to people of African descent, in whom fasting insulin, waist circumference and body mass index discriminated well (areas under the curve 0.85, 0.72 and 0.71) where the ratio did not.","cites":["sumner2005","sumner2008","kimdorner2010"],"grade":"C"},{"text":"The ratio inherits the noise of both inputs. In 51 volunteers measured on three occasions, day-to-day variability was 20 percent for triglycerides and 10 percent for HDL cholesterol, with analytic variability contributing significantly to the HDL figure; a ratio of 3.0 can read 2.4 or 3.6 on another morning with nothing changed. Compare ratios from the same laboratory, in the same fasting state, and read the trend over several draws rather than one value against a cutoff.","cites":["bookstein1990"],"grade":"C"},{"text":"Fasting changes the top of the ratio and not the bottom. After ordinary meals the maximal mean rise in triglycerides is about 26 mg/dL (0.3 mmol/L), while HDL cholesterol, apolipoprotein A1, apolipoprotein B and lipoprotein(a) are not affected by fasting status, so a non-fasting ratio reads higher than a fasting one from the same person and the published cutoffs, all derived fasting, do not apply to it. The European consensus flags non-fasting triglycerides at 175 mg/dL rather than 150 for the same reason.","cites":["nordestgaard2016"],"grade":"A"},{"text":"A very high HDL makes the ratio look better than the person is. In 52,268 men and 64,240 women in Copenhagen, all-cause mortality was U-shaped against HDL cholesterol, lowest at 73 mg/dL in men and 93 in women and higher above 97 in men (hazard ratios 1.36 and 2.06 for 97 to 115 and 116 or more) and above 116 in women; a triglyceride of 100 over an HDL of 110 gives a reassuring ratio under 1 in a range where the cohorts found more deaths, not fewer. The denominator is a marker of metabolism and of drinking, not a credit.","cites":["madsen2017","holmes2014"],"grade":"B"},{"text":"Medications and conditions move the two numbers independently. The 2001 US panel listed corticosteroids, estrogens, retinoids and higher doses of beta-blockers among drugs that raise triglycerides, and beta-blockers, anabolic steroids and progestational agents among those that lower HDL, alongside type 2 diabetes, chronic kidney disease, nephrotic syndrome, smoking and excess alcohol. A ratio that changes after a new prescription may be the prescription.","cites":["ncep2001"],"grade":"A"},{"text":"The ratio is a proxy for small, dense LDL, not a measurement of it, and small is not the only dangerous size. The cutoffs that flagged small LDL vary with the method and the unit: a molar ratio above 1.33 (about 3.0 in mg/dL) in 90 percent of people with small LDL against 16.5 percent with larger LDL among 60 adults with type 2 diabetes; above 0.9 in mmol/L (2.0 in mg/dL) in 75 percent of 39 Japanese men with small LDL against 25 percent without; 3.8 in 658 US patients. Large LDL particles also carry increased coronary risk, particularly when triglycerides are normal or low, so a low ratio does not clear the LDL question. ApoB counts all the particles at once.","cites":["boizel2000","maruyama2003","hanak2004","berneis2002"],"grade":"C"}],"seeAPhysician":["Triglycerides of 500 mg/dL or more on any draw, whatever the HDL, because of pancreatitis risk; the ratio is irrelevant at that level.","A high ratio together with a fasting glucose of 100 mg/dL or more, a hemoglobin A1c in the prediabetes range, a rising waist, or an ALT above the laboratory's range, which together describe insulin resistance that deserves measurement rather than watching.","A high ratio with a high ApoB or LDL cholesterol, or any lipid pattern in a parent or sibling who had a heart attack, stroke or stent before 55 (men) or 65 (women).","An HDL cholesterol above about 100 in a man or 120 in a woman, which lowers the ratio while the cohorts place higher mortality there, especially with heavy alcohol use or abnormal liver tests.","Chest pressure, breathlessness on exertion or calf pain when walking, regardless of any lipid number."],"openQuestions":["No guideline defines a cutoff for the ratio, and the published ones range from 1.9 to 3.8 (mg/dL) depending on sex, population and the reference method; whether a single number can ever serve, or whether the ratio should carry sex- and ancestry-specific lines, is unresolved.","The mg/dL and mmol/L versions of the ratio differ by a factor of about 2.3, and cutoffs are routinely quoted without their unit; the field has not settled on one.","No randomized trial has lowered the ratio toward a target and counted outcomes, so every threshold here describes who the ratio flags or what happened to them in observation.","Whether the ratio adds anything to fasting insulin for detecting insulin resistance, or to ApoB for predicting heart disease, when both are available, has not been tested head to head in a cohort with hard outcomes; the Framingham analysis found insulin resistance predicted coronary disease even after adjusting for the ratio.","Why triglycerides stay low in insulin-resistant people of African ancestry, and which cheap marker should replace the ratio in them, remain open; the national-survey data suggest a lower cutoff might partly rescue it, the direct-measure data say it should not be relied on.","The triglyceride-glucose index, which pairs triglycerides with fasting glucose, predicted diabetes as well as the ratio and mortality less well in one cohort and discriminated the metabolic syndrome better than HOMA-IR in another; which fasting index, if any, should be routine is not decided."],"questions":[{"q":"What is a good triglyceride to HDL ratio?","a":"No guideline says. In the Stanford studies that introduced it, 3.0 (in mg/dL) in overweight adults and 3.5 in a general sample best separated insulin-resistant people from the rest, with about two-thirds sensitivity and specificity; in adults of European ancestry the sex-specific lines were 2.5 for women and 3.5 for men. Lower is better as a rough signal, and the median in 1.35 million US lipid referrals was 2.2. Those are the studies' numbers, derived fasting and in mg/dL, and they do not hold in people of African ancestry."},{"q":"How do I calculate it, and does the unit matter?","a":"Divide fasting triglycerides by HDL cholesterol, both in mg/dL, from the same draw: 120 over 50 is 2.4. The unit matters a great deal. Triglycerides convert at about 88.5 mg/dL per mmol/L and HDL at about 38.7, so the same blood gives a ratio about 2.3 times lower in mmol/L; a cutoff of 3.0 in mg/dL is about 1.3 in mmol/L. A cutoff quoted without its unit cannot be used."},{"q":"Is a high ratio the same as insulin resistance?","a":"No. It is a free hint that the liver is overproducing triglyceride-rich particles and that HDL is being stripped, which is what insulin resistance does to lipids, and in mostly white samples it identified about two-thirds of insulin-resistant adults. A third are missed, some are flagged wrongly, and in Black Americans it does no better than chance. A high ratio is a reason to measure fasting insulin and ApoB, not a diagnosis."},{"q":"If I already have an ApoB, do I need the ratio?","a":"For heart risk, ApoB carries the information: once particle number is known, triglycerides add no independent coronary risk in the genetic studies, and a ratio that improves on a low-carbohydrate diet while LDL rises is not a win. For insulin resistance, the ratio is a weak stand-in for a fasting insulin. Where both ApoB and insulin have been measured, the ratio is a cross-check, not a separate test."}],"citations":[{"id":"laws1992","authors":"Laws A, Reaven GM","title":"Evidence for an independent relationship between insulin resistance and fasting plasma HDL-cholesterol, triglyceride and insulin concentrations","journal":"Journal of Internal Medicine","year":1992,"pmid":"1732395","doi":"10.1111/j.1365-2796.1992.tb00494.x","type":"cross-sectional","finding":"In 18 non-diabetic overweight men sorted by steady-state plasma glucose, the most insulin-resistant group had higher insulin and triglycerides and lower HDL cholesterol with no difference in body mass index, waist-to-hip ratio or fitness.","url":"https://pubmed.ncbi.nlm.nih.gov/1732395/"},{"id":"mclaughlin2003","authors":"McLaughlin T, Abbasi F, Cheal K, Chu J, Lamendola C, Reaven G","title":"Use of metabolic markers to identify overweight individuals who are insulin resistant","journal":"Annals of Internal Medicine","year":2003,"pmid":"14623617","doi":"10.7326/0003-4819-139-10-200311180-00007","type":"cross-sectional","finding":"In 258 non-diabetic overweight adults measured by the insulin suppression test, a ratio cutoff of 3.0 (mg/dL) identified insulin resistance with 64 percent sensitivity and 68 percent specificity, similar to the ATP III metabolic syndrome criteria (52 and 85 percent).","url":"https://pubmed.ncbi.nlm.nih.gov/14623617/"},{"id":"mclaughlin2005","authors":"McLaughlin T, Reaven G, Abbasi F, et al.","title":"Is there a simple way to identify insulin-resistant individuals at increased risk of cardiovascular disease?","journal":"American Journal of Cardiology","year":2005,"pmid":"16054467","doi":"10.1016/j.amjcard.2005.03.085","type":"cross-sectional","finding":"In 449 apparently healthy adults (LDL size in 1,135), a ratio of 3.5 or more was the best routine predictor of insulin resistance and LDL particle diameter, more accurate for small, dense LDL than for insulin resistance.","url":"https://pubmed.ncbi.nlm.nih.gov/16054467/"},{"id":"kannel2008","authors":"Kannel WB, Vasan RS, Keyes MJ, Sullivan LM, Robins SJ","title":"Usefulness of the triglyceride-high-density lipoprotein versus the cholesterol-high-density lipoprotein ratio for predicting insulin resistance and cardiometabolic risk (from the Framingham Offspring Cohort)","journal":"American Journal of Cardiology","year":2008,"pmid":"18312765","doi":"10.1016/j.amjcard.2007.09.109","type":"cohort","finding":"In 3,014 Framingham Offspring participants the ratio correlated 0.46 with HOMA-IR and identified its top quartile with an area under the curve of 0.745; insulin resistance predicted coronary disease (hazard ratio 2.71) even after adjusting for lipid ratios, so the authors called the ratio an imperfect surrogate.","url":"https://pubmed.ncbi.nlm.nih.gov/18312765/"},{"id":"li2008","authors":"Li C, Ford ES, Meng YX, Mokdad AH, Reaven GM","title":"Does the association of the triglyceride to high-density lipoprotein cholesterol ratio with fasting serum insulin differ by race/ethnicity?","journal":"Cardiovascular Diabetology","year":2008,"pmid":"18307789","doi":"10.1186/1475-2840-7-4","type":"cross-sectional","finding":"In 2,652 non-diabetic US adults, the ratio predicted hyperinsulinemia with areas under the curve of 0.77 (non-Hispanic white), 0.75 (non-Hispanic Black) and 0.74 (Mexican American); a cutoff of 3.5 carried a prevalence ratio of 2.16, and ancestry-specific cutoffs of 3.0 and 2.0 (Black adults) carried 2.23.","url":"https://pubmed.ncbi.nlm.nih.gov/18307789/"},{"id":"sumner2005","authors":"Sumner AE, Finley KB, Genovese DJ, Criqui MH, Boston RC","title":"Fasting triglyceride and the triglyceride-HDL cholesterol ratio are not markers of insulin resistance in African Americans","journal":"Archives of Internal Medicine","year":2005,"pmid":"15983289","doi":"10.1001/archinte.165.12.1395","type":"cross-sectional","finding":"In 98 overweight African Americans measured by frequently sampled intravenous glucose tolerance testing, the ratio's area under the curve for insulin resistance was 0.56 and the lipid criteria had 17 percent sensitivity; fasting insulin, body mass index and waist had 0.85, 0.72 and 0.71.","url":"https://pubmed.ncbi.nlm.nih.gov/15983289/"},{"id":"sumner2008","authors":"Sumner AE, Cowie CC","title":"Ethnic differences in the ability of triglyceride levels to identify insulin resistance","journal":"Atherosclerosis","year":2008,"pmid":"17254586","doi":"10.1016/j.atherosclerosis.2006.12.018","type":"cross-sectional","finding":"In 2,804 US adults (NHANES 1999 to 2002), mean triglycerides were 99 mg/dL in non-Hispanic Black, 140 in non-Hispanic white and 144 in Mexican American adults; 64 percent of insulin-resistant Black adults had triglycerides under 130 against 31 and 36 percent.","url":"https://pubmed.ncbi.nlm.nih.gov/17254586/"},{"id":"kimdorner2010","authors":"Kim-Dorner SJ, Deuster PA, Zeno SA, Remaley AT, Poth M","title":"Should triglycerides and the triglycerides to high-density lipoprotein cholesterol ratio be used as surrogates for insulin resistance?","journal":"Metabolism","year":2010,"pmid":"19796777","doi":"10.1016/j.metabol.2009.07.027","type":"cross-sectional","finding":"In 99 African American and 50 white healthy adults, the ratio identified a high HOMA-IR with an area under the curve of 0.770 in whites and 0.639 in African Americans; mean ratios were 2.8 and 1.8.","url":"https://pubmed.ncbi.nlm.nih.gov/19796777/"},{"id":"giannini2011","authors":"Giannini C, Santoro N, Caprio S, et al.","title":"The triglyceride-to-HDL cholesterol ratio: association with insulin resistance in obese youths of different ethnic backgrounds","journal":"Diabetes Care","year":2011,"pmid":"21730284","doi":"10.2337/dc10-2234","type":"cross-sectional","finding":"In 1,452 obese youths, a cutoff of 2.27 carried 6.023 times the odds of insulin resistance in 612 white children; in 357 Hispanic and 483 African American children the curves were not significant and no cutoff could be set.","url":"https://pubmed.ncbi.nlm.nih.gov/21730284/"},{"id":"hadaegh2009","authors":"Hadaegh F, Khalili D, Ghasemi A, Tohidi M, Sheikholeslami F, Azizi F","title":"Triglyceride/HDL-cholesterol ratio is an independent predictor for coronary heart disease in a population of Iranian men","journal":"Nutrition, Metabolism and Cardiovascular Diseases","year":2009,"pmid":"19091534","doi":"10.1016/j.numecd.2008.09.003","type":"cohort","finding":"In 1,824 Iranian men over a median 6.5 years with 163 coronary events, the top quartile of the ratio (6.9 or more) had a hazard ratio of 1.75 against the lowest (under 2.8); the metabolic syndrome was present in 63.6 and 3.0 percent of those quartiles.","url":"https://pubmed.ncbi.nlm.nih.gov/19091534/"},{"id":"salazar2012","authors":"Salazar MR, Carbajal HA, Espeche WG, et al.","title":"Relation among the plasma triglyceride/high-density lipoprotein cholesterol concentration ratio, insulin resistance, and associated cardio-metabolic risk factors in men and women","journal":"American Journal of Cardiology","year":2012,"pmid":"22449634","doi":"10.1016/j.amjcard.2012.02.016","type":"cross-sectional","finding":"In 1,102 women and 464 men of primarily European ancestry, ratios above 2.5 in women and 3.5 in men identified insulin resistance with about 40 percent sensitivity and about 80 percent specificity.","url":"https://pubmed.ncbi.nlm.nih.gov/22449634/"},{"id":"salazar2013","authors":"Salazar MR, Carbajal HA, Espeche WG, et al.","title":"Identifying cardiovascular disease risk and outcome: use of the plasma triglyceride/high-density lipoprotein cholesterol concentration ratio versus metabolic syndrome criteria","journal":"Journal of Internal Medicine","year":2013,"pmid":"23331522","doi":"10.1111/joim.12036","type":"cohort","finding":"Among 527 women and 269 men resurveyed after 9 years in Rauch City, Argentina, a ratio above the sex-specific cutoffs predicted a first cardiovascular event as well as the metabolic syndrome (hazard ratios 2.01 and 2.09 after adjustment for sex and age).","url":"https://pubmed.ncbi.nlm.nih.gov/23331522/"},{"id":"gaziano1997","authors":"Gaziano JM, Hennekens CH, O'Donnell CJ, Breslow JL, Buring JE","title":"Fasting triglycerides, high-density lipoprotein, and risk of myocardial infarction","journal":"Circulation","year":1997,"pmid":"9355888","doi":"10.1161/01.cir.96.8.2520","type":"cross-sectional","finding":"In 340 first heart attacks and 340 matched controls, the top quartile of the ratio carried a relative risk of 16.0 (95 percent CI 7.7 to 33.1) against the bottom; triglycerides alone carried 6.8 crude and 2.7 after adjustment for HDL.","url":"https://pubmed.ncbi.nlm.nih.gov/9355888/"},{"id":"daluz2008","authors":"da Luz PL, Favarato D, Faria-Neto JR Jr, Lemos P, Chagas AC","title":"High ratio of triglycerides to HDL-cholesterol predicts extensive coronary disease","journal":"Clinics (Sao Paulo)","year":2008,"pmid":"18719750","doi":"10.1590/s1807-59322008000400003","type":"cross-sectional","finding":"In 374 high-risk patients at angiography, the ratio had the strongest association of any lipid with extensive coronary disease (odds ratio 2.01); total and LDL cholesterol were not significant, and only the ratio and HDL were useful on receiver operating curves.","url":"https://pubmed.ncbi.nlm.nih.gov/18719750/"},{"id":"bittner2009","authors":"Bittner V, Johnson BD, Zineh I, et al.","title":"The triglyceride/high-density lipoprotein cholesterol ratio predicts all-cause mortality in women with suspected myocardial ischemia: a report from the Women's Ischemia Syndrome Evaluation (WISE)","journal":"American Heart Journal","year":2009,"pmid":"19249427","doi":"10.1016/j.ahj.2008.11.014","type":"cohort","finding":"In 544 women referred for angiography, followed a median 6 years, the log ratio carried a hazard ratio of 1.95 for all-cause mortality after adjustment including angiographic severity; 1.54 for cardiovascular events before angiographic adjustment.","url":"https://pubmed.ncbi.nlm.nih.gov/19249427/"},{"id":"vega2014","authors":"Vega GL, Barlow CE, Grundy SM, Leonard D, DeFina LF","title":"Triglyceride-to-high-density-lipoprotein-cholesterol ratio is an index of heart disease mortality and of incidence of type 2 diabetes mellitus in men","journal":"Journal of Investigative Medicine","year":2014,"pmid":"24402298","doi":"10.2310/JIM.0000000000000044","type":"cohort","finding":"In 39,447 men over 581,194 person-years, a ratio at or above 3.5 predicted coronary, cardiovascular and all-cause mortality after adjustment including non-HDL cholesterol; diabetes incidence in 22,215 men was 2 times higher above the cutoff.","url":"https://pubmed.ncbi.nlm.nih.gov/24402298/"},{"id":"cordero2009","authors":"Cordero A, Andrés E, Ordoñez B, et al.","title":"Usefulness of triglycerides-to-high-density lipoprotein cholesterol ratio for predicting the first coronary event in men","journal":"American Journal of Cardiology","year":2009,"pmid":"19892056","doi":"10.1016/j.amjcard.2009.07.008","type":"cross-sectional","finding":"In 208 cases and 2,080 controls among Spanish male workers, the ratio was associated with a first coronary event in every body mass index category (odds ratio 1.47) where LDL cholesterol was barely associated (1.01).","url":"https://pubmed.ncbi.nlm.nih.gov/19892056/"},{"id":"chen2022","authors":"Chen Y, Chang Z, Liu Y, et al.","title":"Triglyceride to high-density lipoprotein cholesterol ratio and cardiovascular events in the general population: A systematic review and meta-analysis of cohort studies","journal":"Nutrition, Metabolism and Cardiovascular Diseases","year":2022,"pmid":"34953633","doi":"10.1016/j.numecd.2021.11.005","type":"meta-analysis","finding":"Across 13 cohorts with 207,515 participants, the highest category of the ratio carried a pooled hazard ratio of 1.43 (95 percent CI 1.26 to 1.62) for cardiovascular events, 1.26 after correction for publication bias, and 1.08 per unit of the ratio.","url":"https://pubmed.ncbi.nlm.nih.gov/34953633/"},{"id":"quispe2015","authors":"Quispe R, Manalac RJ, Faridi KF, et al.","title":"Relationship of the triglyceride to high-density lipoprotein cholesterol (TG/HDL-C) ratio to the remainder of the lipid profile: The Very Large Database of Lipids-4 (VLDL-4) study","journal":"Atherosclerosis","year":2015,"pmid":"26232164","doi":"10.1016/j.atherosclerosis.2015.06.057","type":"cross-sectional","finding":"In 1,350,908 US adults referred for ultracentrifugation, median ratio 2.2, remnant cholesterol, non-HDL cholesterol and LDL density rose steadily with the ratio; about 47 percent of the ratio's variance was attributable to remnant cholesterol.","url":"https://pubmed.ncbi.nlm.nih.gov/26232164/"},{"id":"boizel2000","authors":"Boizel R, Benhamou PY, Lardy B, Laporte F, Foulon T, Halimi S","title":"Ratio of triglycerides to HDL cholesterol is an indicator of LDL particle size in patients with type 2 diabetes and normal HDL cholesterol levels","journal":"Diabetes Care","year":2000,"pmid":"11092292","doi":"10.2337/diacare.23.11.1679","type":"cross-sectional","finding":"In 60 adults with type 2 diabetes and normal HDL, the molar ratio correlated inversely with LDL size (R squared 0.59) and exceeded 1.33 in 90 percent of those with small LDL against 16.5 percent with larger LDL.","url":"https://pubmed.ncbi.nlm.nih.gov/11092292/"},{"id":"maruyama2003","authors":"Maruyama C, Imamura K, Teramoto T","title":"Assessment of LDL particle size by triglyceride/HDL-cholesterol ratio in non-diabetic, healthy subjects without prominent hyperlipidemia","journal":"Journal of Atherosclerosis and Thrombosis","year":2003,"pmid":"14564088","doi":"10.5551/jat.10.186","type":"cross-sectional","finding":"In 39 Japanese men, 75 percent of those with small LDL had a ratio above 0.9 in mmol/L (2.0 in mg/dL) against 25 percent with normal-sized LDL.","url":"https://pubmed.ncbi.nlm.nih.gov/14564088/"},{"id":"hanak2004","authors":"Hanak V, Munoz J, Teague J, Stanley A Jr, Bittner V","title":"Accuracy of the triglyceride to high-density lipoprotein cholesterol ratio for prediction of the low-density lipoprotein phenotype B","journal":"American Journal of Cardiology","year":2004,"pmid":"15246907","doi":"10.1016/j.amjcard.2004.03.069","type":"cross-sectional","finding":"A ratio of 3.8, derived from the ATP III thresholds of triglycerides under 150 and HDL over 40 mg/dL, placed 79 percent of LDL phenotype B above and 81 percent of phenotype A below it in 658 patients measured by nuclear magnetic resonance.","url":"https://pubmed.ncbi.nlm.nih.gov/15246907/"},{"id":"dobiasova2001","authors":"Dobiásová M, Frohlich J","title":"The plasma parameter log (TG/HDL-C) as an atherogenic index: correlation with lipoprotein particle size and esterification rate in apoB-lipoprotein-depleted plasma (FER(HDL))","journal":"Clinical Biochemistry","year":2001,"pmid":"11738396","doi":"10.1016/s0009-9120(01)00263-6","type":"laboratory","finding":"In 1,433 subjects from 35 cohorts, log (TG/HDL-C) correlated 0.803 with the cholesterol esterification rate in apoB-depleted plasma and minus 0.776 with LDL size in 35 normal subjects; the authors named it the atherogenic index of plasma.","url":"https://pubmed.ncbi.nlm.nih.gov/11738396/"},{"id":"adiels2008","authors":"Adiels M, Olofsson SO, Taskinen MR, Borén J","title":"Overproduction of very low-density lipoproteins is the hallmark of the dyslipidemia in the metabolic syndrome","journal":"Arteriosclerosis, Thrombosis, and Vascular Biology","year":2008,"pmid":"18565848","doi":"10.1161/ATVBAHA.107.160192","type":"review","finding":"Overproduction of large VLDL particles initiates a sequence of higher remnant levels, smaller LDL and lower HDL cholesterol, abnormalities that precede the diagnosis of type 2 diabetes by several years.","url":"https://pubmed.ncbi.nlm.nih.gov/18565848/"},{"id":"rashid2003","authors":"Rashid S, Watanabe T, Sakaue T, Lewis GF","title":"Mechanisms of HDL lowering in insulin resistant, hypertriglyceridemic states: the combined effect of HDL triglyceride enrichment and elevated hepatic lipase activity","journal":"Clinical Biochemistry","year":2003,"pmid":"12951168","doi":"10.1016/s0009-9120(03)00078-x","type":"review","finding":"Triglyceride enrichment of HDL through enhanced CETP-mediated exchange with triglyceride-rich lipoproteins, combined with hepatic lipase activity, drives the fall in HDL cholesterol and apoA-I in insulin-resistant states.","url":"https://pubmed.ncbi.nlm.nih.gov/12951168/"},{"id":"packard2003","authors":"Packard CJ","title":"Triacylglycerol-rich lipoproteins and the generation of small, dense low-density lipoprotein","journal":"Biochemical Society Transactions","year":2003,"pmid":"14505481","doi":"10.1042/bst0311066","type":"review","finding":"Elevated large triglyceride-rich VLDL leads to slowly metabolized LDL (5-day residence) that exchanges cholesteryl ester for triglyceride and, under hepatic lipase, becomes small and dense; small, dense LDL is retained by the artery wall and readily oxidized.","url":"https://pubmed.ncbi.nlm.nih.gov/14505481/"},{"id":"berneis2002","authors":"Berneis KK, Krauss RM","title":"Metabolic origins and clinical significance of LDL heterogeneity","journal":"Journal of Lipid Research","year":2002,"pmid":"12235168","doi":"10.1194/jlr.r200004-jlr200","type":"review","finding":"The atherogenic lipoprotein phenotype (small, dense LDL, raised remnants, low HDL, insulin resistance) raises coronary risk; large LDL particles also carry increased risk, particularly with normal or low triglycerides.","url":"https://pubmed.ncbi.nlm.nih.gov/12235168/"},{"id":"voight2012","authors":"Voight BF, Peloso GM, Orho-Melander M, et al.","title":"Plasma HDL cholesterol and risk of myocardial infarction: a mendelian randomisation study","journal":"Lancet","year":2012,"pmid":"22607825","doi":"10.1016/S0140-6736(12)60312-2","type":"mendelian-randomization","finding":"An HDL-raising LIPG variant (0.14 mmol/L higher) expected to cut heart-attack risk 13 percent had an odds ratio of 0.99 in 20,913 cases and 95,407 controls; a 14-variant HDL score gave 0.93 per SD against 0.62 from observation, while the LDL score behaved as expected (2.13).","url":"https://pubmed.ncbi.nlm.nih.gov/22607825/"},{"id":"do2013","authors":"Do R, Willer CJ, Schmidt EM, et al.","title":"Common variants associated with plasma triglycerides and risk for coronary artery disease","journal":"Nature Genetics","year":2013,"pmid":"24097064","doi":"10.1038/ng.2795","type":"mendelian-randomization","finding":"Using 185 lipid variants, the strength of a variant's effect on triglycerides correlated with its effect on coronary disease in a model accounting for LDL and HDL cholesterol, suggesting triglyceride-rich lipoproteins causally influence risk.","url":"https://pubmed.ncbi.nlm.nih.gov/24097064/"},{"id":"holmes2015","authors":"Holmes MV, Asselbergs FW, Palmer TM, et al.","title":"Mendelian randomization of blood lipids for coronary heart disease","journal":"European Heart Journal","year":2015,"pmid":"24474739","doi":"10.1093/eurheartj/eht571","type":"mendelian-randomization","finding":"In 17 studies with 62,199 participants, a triglyceride allele score carried an odds ratio of 1.62 per log unit for coronary disease, falling to 1.01 after adjustment for HDL, LDL and statin use; the HDL score showed no robust association once restricted or adjusted.","url":"https://pubmed.ncbi.nlm.nih.gov/24474739/"},{"id":"ference2019","authors":"Ference BA, Kastelein JJP, Ray KK, et al.","title":"Association of triglyceride-lowering LPL variants and LDL-C-lowering LDLR variants with risk of coronary heart disease","journal":"JAMA","year":2019,"pmid":"30694319","doi":"10.1001/jama.2018.20045","type":"mendelian-randomization","finding":"In 654,783 people, LPL and LDLR scores lowered coronary risk identically per 10 mg/dL lower ApoB (odds ratios 0.771 and 0.773); the associations of triglycerides and LDL cholesterol with coronary disease became null after adjusting for ApoB.","url":"https://pubmed.ncbi.nlm.nih.gov/30694319/"},{"id":"sniderman2011","authors":"Sniderman AD, Williams K, Contois JH, et al.","title":"A meta-analysis of low-density lipoprotein cholesterol, non-high-density lipoprotein cholesterol, and apolipoprotein B as markers of cardiovascular risk","journal":"Circulation: Cardiovascular Quality and Outcomes","year":2011,"pmid":"21487090","doi":"10.1161/CIRCOUTCOMES.110.959247","type":"meta-analysis","finding":"Across 12 studies with 233,455 subjects and 22,950 events, ApoB was the most potent marker (standardized relative risk 1.43), non-HDL cholesterol intermediate (1.34) and LDL cholesterol least (1.25).","url":"https://pubmed.ncbi.nlm.nih.gov/21487090/"},{"id":"sniderman2019","authors":"Sniderman AD, Thanassoulis G, Glavinovic T, et al.","title":"Apolipoprotein B Particles and Cardiovascular Disease: A Narrative Review","journal":"JAMA Cardiology","year":2019,"pmid":"31642874","doi":"10.1001/jamacardio.2019.3780","type":"review","finding":"Cholesterol enters the artery wall only inside ApoB particles; smaller cholesterol-depleted particles are trapped more readily than larger ones; with the exceptions of type III remnants and lipoprotein(a), all ApoB particles are equally atherogenic.","url":"https://pubmed.ncbi.nlm.nih.gov/31642874/"},{"id":"nordestgaard2016","authors":"Nordestgaard BG, Langsted A, Mora S, et al.","title":"Fasting is not routinely required for determination of a lipid profile: clinical and laboratory implications including flagging at desirable concentration cut-points: a joint consensus statement from the European Atherosclerosis Society and European Federation of Clinical Chemistry and Laboratory Medicine","journal":"European Heart Journal","year":2016,"pmid":"27122601","doi":"10.1093/eurheartj/ehw152","type":"consensus","finding":"Maximal mean post-meal rise in triglycerides is 0.3 mmol/L (26 mg/dL) while HDL cholesterol, apoA1, apoB and lipoprotein(a) are unaffected by fasting status; non-fasting triglycerides flagged at 175 mg/dL and fasting at 150.","url":"https://pubmed.ncbi.nlm.nih.gov/27122601/"},{"id":"ncep2001","authors":"Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults","title":"Executive summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III)","journal":"JAMA","year":2001,"pmid":"11368702","doi":"10.1001/jama.285.19.2486","type":"guideline","finding":"Triglycerides under 150 mg/dL normal, 150 to 199 borderline high; low HDL under 40 (raised from 35); metabolic syndrome criteria triglycerides 150 or more and HDL under 40 (men) or 50 (women); no HDL goal; causes of low HDL and raised triglycerides listed, including drugs.","url":"https://pubmed.ncbi.nlm.nih.gov/11368702/"},{"id":"pantoja2019","authors":"Pantoja-Torres B, Toro-Huamanchumo CJ, Urrunaga-Pastor D, et al.","title":"High triglycerides to HDL-cholesterol ratio is associated with insulin resistance in normal-weight healthy adults","journal":"Diabetes and Metabolic Syndrome: Clinical Research and Reviews","year":2019,"pmid":"30641729","doi":"10.1016/j.dsx.2018.10.006","type":"cross-sectional","finding":"In 118 normal-weight, euthyroid, non-diabetic adults in Lima, a ratio of 3 or more carried 3.16 times the prevalence of insulin resistance (HOMA-IR 2.28 or more) and 2.36 times the prevalence of hyperinsulinemia after a glucose load.","url":"https://pubmed.ncbi.nlm.nih.gov/30641729/"},{"id":"vonbibra2017","authors":"von Bibra H, Saha S, Hapfelmeier A, Müller G, Schwarz PEH","title":"Impact of the Triglyceride/High-Density Lipoprotein Cholesterol Ratio and the Hypertriglyceremic-Waist Phenotype to Predict the Metabolic Syndrome and Insulin Resistance","journal":"Hormone and Metabolic Research","year":2017,"pmid":"28597452","doi":"10.1055/s-0043-107782","type":"cross-sectional","finding":"In 1,932 non-diabetic Germans, the ratio predicted the metabolic syndrome with an area under the curve of 0.817 and optimal cutoffs (mg/dL) of 2.8 in men and 1.9 in women, better than HOMA-IR (0.761).","url":"https://pubmed.ncbi.nlm.nih.gov/28597452/"},{"id":"azarpazhooh2021","authors":"Azarpazhooh MR, Najafi F, Darbandi M, Kiarasi S, Oduyemi T, Spence JD","title":"Triglyceride/High-Density Lipoprotein Cholesterol Ratio: A Clue to Metabolic Syndrome, Insulin Resistance, and Severe Atherosclerosis","journal":"Lipids","year":2021,"pmid":"33881177","doi":"10.1002/lipd.12302","type":"cross-sectional","finding":"In three cohorts totalling 13,908 participants, a ratio above the optimal cutoff carried 6.00, 4.04 and 3.50 times the odds of the metabolic syndrome, identified insulin resistance by HOMA-IR and tracked carotid atherosclerosis severity.","url":"https://pubmed.ncbi.nlm.nih.gov/33881177/"},{"id":"bookstein1990","authors":"Bookstein L, Gidding SS, Donovan M, Smith FA","title":"Day-to-day variability of serum cholesterol, triglyceride, and high-density lipoprotein cholesterol levels. Impact on the assessment of risk according to the National Cholesterol Education Program guidelines","journal":"Archives of Internal Medicine","year":1990,"pmid":"2383160","type":"laboratory","finding":"In 51 volunteers measured three times, day-to-day variability was 20 percent for triglycerides and 10 percent for HDL cholesterol, with analytic variability contributing significantly to the HDL figure.","url":"https://pubmed.ncbi.nlm.nih.gov/2383160/"},{"id":"madsen2017","authors":"Madsen CM, Varbo A, Nordestgaard BG","title":"Extreme high high-density lipoprotein cholesterol is paradoxically associated with high mortality in men and women: two prospective cohort studies","journal":"European Heart Journal","year":2017,"pmid":"28419274","doi":"10.1093/eurheartj/ehx163","type":"cohort","finding":"In 52,268 men and 64,240 women, mortality was U-shaped against HDL, lowest at 73 mg/dL in men and 93 in women; hazard ratios 1.36 and 2.06 for men at 97 to 115 and 116 or more, 1.10 and 1.68 for women at 116 to 134 and 135 or more.","url":"https://pubmed.ncbi.nlm.nih.gov/28419274/"},{"id":"temorenga2014","authors":"Te Morenga LA, Howatson AJ, Jones RM, Mann J","title":"Dietary sugars and cardiometabolic risk: systematic review and meta-analyses of randomized controlled trials of the effects on blood pressure and lipids","journal":"American Journal of Clinical Nutrition","year":2014,"pmid":"24808490","doi":"10.3945/ajcn.113.081521","type":"meta-analysis","finding":"Across 39 randomized trials, higher free-sugar intake raised triglycerides by 0.11 mmol/L and HDL by 0.02 mmol/L, most clearly in trials with energy balance maintained and no weight difference.","url":"https://pubmed.ncbi.nlm.nih.gov/24808490/"},{"id":"mansoor2016","authors":"Mansoor N, Vinknes KJ, Veierød MB, Retterstøl K","title":"Effects of low-carbohydrate diets v. low-fat diets on body weight and cardiovascular risk factors: a meta-analysis of randomised controlled trials","journal":"British Journal of Nutrition","year":2016,"pmid":"26768850","doi":"10.1017/S0007114515004699","type":"meta-analysis","finding":"In 11 randomized trials with 1,369 participants, low-carbohydrate diets lowered triglycerides 0.26 mmol/L and raised HDL 0.14 mmol/L more than low-fat diets, but raised LDL cholesterol 0.16 mmol/L.","url":"https://pubmed.ncbi.nlm.nih.gov/26768850/"},{"id":"hasan2020","authors":"Hasan B, Nayfeh T, Alzuabi M, et al.","title":"Weight loss and serum lipids in overweight and obese adults: a systematic review and meta-analysis","journal":"Journal of Clinical Endocrinology and Metabolism","year":2020,"pmid":"32954416","doi":"10.1210/clinem/dgaa673","type":"meta-analysis","finding":"Across 73 randomized trials with 32,496 patients, each kilogram lost through lifestyle intervention lowered triglycerides 4.0 mg/dL and raised HDL 0.46 mg/dL at 12 months.","url":"https://pubmed.ncbi.nlm.nih.gov/32954416/"},{"id":"dattilo1992","authors":"Dattilo AM, Kris-Etherton PM","title":"Effects of weight reduction on blood lipids and lipoproteins: a meta-analysis","journal":"American Journal of Clinical Nutrition","year":1992,"pmid":"1386186","doi":"10.1093/ajcn/56.2.320","type":"meta-analysis","finding":"Across 70 studies, HDL rose 0.009 mmol/L per kilogram lost at stabilized weight and fell 0.007 mmol/L per kilogram during active weight loss.","url":"https://pubmed.ncbi.nlm.nih.gov/1386186/"},{"id":"kodama2007","authors":"Kodama S, Tanaka S, Saito K, et al.","title":"Effect of aerobic exercise training on serum levels of high-density lipoprotein cholesterol: a meta-analysis","journal":"Archives of Internal Medicine","year":2007,"pmid":"17533202","doi":"10.1001/archinte.167.10.999","type":"meta-analysis","finding":"Across 25 randomized trials aerobic exercise raised HDL by 2.53 mg/dL; a minimum of 120 minutes or 900 kcal a week was needed, and each extra 10 minutes per session added about 1.4 mg/dL.","url":"https://pubmed.ncbi.nlm.nih.gov/17533202/"},{"id":"kelley2011","authors":"Kelley GA, Kelley KS, Roberts S, Haskell W","title":"Efficacy of aerobic exercise and a prudent diet for improving selected lipids and lipoproteins in adults: a meta-analysis of randomized controlled trials","journal":"BMC Medicine","year":2011,"pmid":"21676220","doi":"10.1186/1741-7015-9-74","type":"meta-analysis","finding":"Across six randomized trials in up to 559 adults, aerobic exercise plus a prudent diet lowered triglycerides 10.6 mg/dL (18.2 percent) with no change in HDL cholesterol.","url":"https://pubmed.ncbi.nlm.nih.gov/21676220/"},{"id":"skulasray2019","authors":"Skulas-Ray AC, Wilson PWF, Harris WS, et al.","title":"Omega-3 fatty acids for the management of hypertriglyceridemia: a science advisory from the American Heart Association","journal":"Circulation","year":2019,"pmid":"31422671","doi":"10.1161/CIR.0000000000000709","type":"consensus","finding":"Prescription omega-3 at 4 g/day lowers triglycerides by 30 percent or more in very high triglycerides (with an LDL rise for EPA plus DHA in that group), and modestly lowers non-HDL cholesterol and ApoB in the largest trials.","url":"https://pubmed.ncbi.nlm.nih.gov/31422671/"},{"id":"brien2011","authors":"Brien SE, Ronksley PE, Turner BJ, Mukamal KJ, Ghali WA","title":"Effect of alcohol consumption on biological markers associated with risk of coronary heart disease: systematic review and meta-analysis of interventional studies","journal":"BMJ","year":2011,"pmid":"21343206","doi":"10.1136/bmj.d636","type":"meta-analysis","finding":"In 44 interventional studies alcohol raised HDL cholesterol by 0.094 mmol/L with a dose-response trend and did not affect fasting triglycerides.","url":"https://pubmed.ncbi.nlm.nih.gov/21343206/"},{"id":"holmes2014","authors":"Holmes MV, Dale CE, Zuccolo L, et al.","title":"Association between alcohol and cardiovascular disease: Mendelian randomisation analysis based on individual participant data","journal":"BMJ","year":2014,"pmid":"25011450","doi":"10.1136/bmj.g4164","type":"mendelian-randomization","finding":"In 261,991 people of European descent, carriers of an ADH1B variant who drank 17.2 percent less had lower coronary (odds ratio 0.90) and ischemic stroke (0.83) risk across all categories of consumption.","url":"https://pubmed.ncbi.nlm.nih.gov/25011450/"},{"id":"vandewiel2012","authors":"Van de Wiel A","title":"The effect of alcohol on postprandial and fasting triglycerides","journal":"International Journal of Vascular Medicine","year":2012,"pmid":"21961068","doi":"10.1155/2012/862504","type":"review","finding":"Alcohol with a fat-containing meal adds to the postprandial triglyceride peak through acute inhibition of lipoprotein lipase and increased hepatic synthesis of large VLDL.","url":"https://pubmed.ncbi.nlm.nih.gov/21961068/"},{"id":"hartley2016","authors":"Hartley L, May MD, Loveman E, Colquitt JL, Rees K","title":"Dietary fibre for the primary prevention of cardiovascular disease","journal":"Cochrane Database of Systematic Reviews","year":2016,"pmid":"26758499","doi":"10.1002/14651858.CD011472.pub2","type":"meta-analysis","finding":"In 23 randomized trials with 1,513 participants, added fiber lowered total and LDL cholesterol but did not change triglycerides (0.00 mmol/L) and slightly lowered HDL (minus 0.03 mmol/L).","url":"https://pubmed.ncbi.nlm.nih.gov/26758499/"},{"id":"reynolds2019","authors":"Reynolds A, Mann J, Cummings J, Winter N, Mete E, Te Morenga L","title":"Carbohydrate quality and human health: a series of systematic reviews and meta-analyses","journal":"Lancet","year":2019,"pmid":"30638909","doi":"10.1016/S0140-6736(18)31809-9","type":"meta-analysis","finding":"Across 185 prospective studies and 58 trials, the highest fiber consumers had 15 to 30 percent lower all-cause and cardiovascular mortality and incidence of coronary disease, stroke and type 2 diabetes, with the greatest risk reduction at 25 to 29 g a day.","url":"https://pubmed.ncbi.nlm.nih.gov/30638909/"},{"id":"kruisbrink2017","authors":"Kruisbrink M, Robertson W, Ji C, Miller MA, Geleijnse JM, Cappuccio FP","title":"Association of sleep duration and quality with blood lipids: a systematic review and meta-analysis of prospective studies","journal":"BMJ Open","year":2017,"pmid":"29247105","doi":"10.1136/bmjopen-2017-018585","type":"meta-analysis","finding":"Pooling 6 prospective studies with 30,033 participants, short sleep carried a relative risk of 1.01 for incident dyslipidemia, with heterogeneity and publication bias; no supportive evidence of a relationship was found.","url":"https://pubmed.ncbi.nlm.nih.gov/29247105/"},{"id":"spiegel1999","authors":"Spiegel K, Leproult R, Van Cauter E","title":"Impact of sleep debt on metabolic and endocrine function","journal":"Lancet","year":1999,"pmid":"10543671","doi":"10.1016/S0140-6736(99)01376-8","type":"rct","finding":"Six nights of 4 hours in bed lowered glucose tolerance and raised evening cortisol and sympathetic activity in 11 healthy young men.","url":"https://pubmed.ncbi.nlm.nih.gov/10543671/"},{"id":"babin1999","authors":"Babin PJ, Bogerd J, Kooiman FP, Van Marrewijk WJ, Van der Horst DJ","title":"Apolipophorin II/I, apolipoprotein B, vitellogenin, and microsomal triglyceride transfer protein genes are derived from a common ancestor","journal":"Journal of Molecular Evolution","year":1999,"pmid":"10368443","doi":"10.1007/pl00006528","type":"laboratory","finding":"Conserved motifs, identical residues and ancestral exon boundaries show apoB, insect apolipophorin, vitellogenin and MTP descend from one ancestral large lipid transfer module.","url":"https://pubmed.ncbi.nlm.nih.gov/10368443/"},{"id":"vanderhorst2009","authors":"Van der Horst DJ, Roosendaal SD, Rodenburg KW","title":"Circulatory lipid transport: lipoprotein assembly and function from an evolutionary perspective","journal":"Molecular and Cellular Biochemistry","year":2009,"pmid":"19130182","doi":"10.1007/s11010-008-0011-3","type":"review","finding":"Large lipid transfer proteins evolved from the earliest animals; insect high-density lipophorin acts as a reusable lipid shuttle that reloads, whereas mammalian lipoproteins taken up through LDL receptor family members are degraded in lysosomes.","url":"https://pubmed.ncbi.nlm.nih.gov/19130182/"}],"written":"2026-10-23","updated":"2026-10-23","reviewed":true,"reviewedOn":"2026-10-24","related":{"essays":[],"glossary":["ApoB","Insulin resistance","HOMA-IR","Fasting insulin","Reference range","Optimal range"],"symptoms":["high-cholesterol","weight-loss-resistance"],"entries":["triglycerides","hdl-cholesterol","fasting-insulin","homa-ir","apob","non-hdl-cholesterol","ldl-cholesterol","alt","fiber"]},"url":"https://taggehealth.com/labs/triglyceride-hdl-ratio","author":"Daniel Tagge, MD","publisher":"Tagge Health","license":"Free to read and cite with attribution to the URL above.","inPractice":"Lipids are complicated, and I do not read them as one number with one zone. I calculate the triglyceride-to-HDL ratio on every lipid panel because it costs nothing, and I read it the way I read any single marker: one marker is a clue, three are a story. A high ratio beside a fasting insulin that is high for the glucose next to it, a waist that has grown and an ALT drifting up is a story about insulin resistance, and I act on the story. A high ratio on its own earns two measurements, ApoB and fasting insulin, not a verdict; the ratio is the prompt and those are the answers, and when ApoB is in hand it is ApoB I weigh for heart risk. I keep three cautions in front of me. The ratio misleads in people of African ancestry, whose triglycerides often stay normal while they are insulin resistant, so in them I go straight to the insulin. It moves with the previous night's dinner and with the laboratory's assays, so I never compare a fasting ratio with a non-fasting one, or one laboratory with another, and I watch the direction over several draws rather than one value against a line. And a very high HDL makes the ratio look better than the person is. I will say plainly where I am ahead of the evidence: no trial has treated anyone to a ratio target, the cutoffs on this page belong to the studies that published them and I quote them as theirs, and I use the ratio to decide what to measure next, never to decide that someone is well."}