Different lipid survival value

[Guest guest]

Hi All,

Many measure various lipids in our blood tests for how they may correlate or

predict

our health risks. But, how important are the different lipids in our blood that

can

be measured for our chances of harm from their high or low levels?

The below paper seems to outline strategies to be taken regarding such issues.

Race appears not to matter.

See the pdf-available below for suggestions and information.

Barzi F, Patel A, Woodward M, Lawes CM, Ohkubo T, Gu D, Lam TH, Ueshima H; Asia

Pacific Cohort Studies Collaboration.

A comparison of lipid variables as predictors of cardiovascular disease in the

Asia

Pacific region.

Ann Epidemiol. 2005 May;15(5):405-13.

PMID: 15840555

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve & db=pubmed & dopt=Abstra\

ct & list_uids=15840555 & query_hl=38

PURPOSE: Many guidelines advocate measurement of total or low density

lipoprotein

cholesterol (LDL), high density lipoprotein cholesterol (HDL), and triglycerides

(TG) to determine treatment recommendations for preventing coronary heart

disease

(CHD) and cardiovascular disease (CVD). ... METHODS: Hazard ratios for coronary

and

cardiovascular deaths by fourths of total cholesterol (TC), LDL, HDL, TG,

non-HDL,

TC/HDL, and TG/HDL values, and for a one standard deviation change in these

variables, were derived in an individual participant data meta-analysis of 32

cohort

studies conducted in the Asia-Pacific region. The predictive value of each lipid

variable was assessed using the likelihood ratio statistic. RESULTS: Adjusting

for

confounders and regression dilution, each lipid variable had a positive

(negative

for HDL) log-linear association with fatal CHD and CVD. Individuals in the

highest

fourth of each lipid variable had approximately twice the risk of CHD compared

with

those with lowest levels. TG and HDL were each better predictors of CHD and CVD

risk

compared with TC alone, with test statistics similar to TC/HDL and TG/HDL

ratios.

Calculated LDL was a relatively poor predictor. CONCLUSIONS: While LDL reduction

remains the main target of intervention for lipid-lowering, these data support

the

potential use of TG or lipid ratios for CHD risk prediction.

.... ‘‘ANZ’’ (from Australia or New Zealand) ... In the ANZ cohorts, 60% of CVD

deaths were due to CHD; the corresponding figure for the Asian

cohorts was 28%, reflecting the relative prominence of stroke

mortality in this region. Substantial correlation was observed

between TGand TC (Pearson’s correlation coefficient 0.26),

TG/HDL and TC (0.20), HDL and TG (-0.24), non-HDL

and TG (0.34), and TC/HDL and TG (0.36); however, none

of these were so highly correlated as to present problems with

the association between each lipid variable and the risk of

CHD and CVD death were as follows (Table 1): TCd32

cohorts, 2,486,935 person-years, 2253 CHD deaths, and

5678 CVD deaths; TGd23 cohorts, 784,241 person-years,

670 CHD deaths, and 1778 CVD deaths; HDLd19 cohorts,

508,226 person-years, 623 CHD deaths, and 1437 CVD

deaths.

The relative predictive value of the lipid variables was

assessed on a total of 58,386 participants from the 17 cohorts

with information on all three measured lipids variables.

Among these cohorts, the mean age was 48 years (ranging

from 39 to 78 years, by cohort) and 49% were women

(ranging from 23% to 65%, by cohort). During 432,198

person-years of follow-up, a total of 459 CHD and 1128 CVD

deaths occurred. In the ANZ cohorts, 60% of CVD deaths

were due to CHD; the corresponding figure for the Asian

cohorts was 28%, reflecting the relative prominence of stroke

mortality in this region. Substantial correlation was observed

between TGand TC (Pearson’s correlation coefficient 0.26),

TG/HDL and TC (0.20), HDL and TG (-0.24), non-HDL

and TG (0.34), and TC/HDL and TG (0.36); however, none

of these were so highly correlated as to present problems with

co-linearity when combined in a single model. Little

correlation was observed between TC and HDL cholesterol

(Pearson’s correlation coefficient -0.04).

Associations between Lipid Variables, CHD, and CVD Death

Each of the lipid measurements was associated with the risk

of CHD death in a log-linear fashion (all p <0.05) (Fig. 1).

In both Australia and Asia, individuals in the highest fourth

of each of the lipid variables (lowest, for HDL) had

approximately twice the risk of CHD as compared with

those with the lowest levels (highest, for HDL). There was

some evidence of regional heterogeneity in the associations

for TG; this interaction was observed for CHD death, but

not for CVD death, and was of borderline statistical

significance (p = 0.05). For CVD death, the associations

for each of the lipid variables was qualitatively similar to

those for CHD death, but substantially attenuated, and with

no evidence of regional heterogeneity (Fig. 2). There was no

evidence of any heterogeneity between cohorts within Asia

or within ANZ for the association between total cholesterol,

triglycerides, or HDL, and CHD death (all p > 0.1).

Single and Joint Effects of Lipid Variables

While all the variables contributed significantly to the

prediction of CHD death, TG and the ratio TG/HDL had

large likelihood ratio chi-square statistics, of a similar mag-nitude

to those observed for HDL and TC/HDL (Table 2).

TABLE 2. Likelihood ratio chi-square statistics associated with the single and

joint

effects of the lipid variables on the risks of CHD and CVD mortality

----------------------------------------------

----CHD death CVD death

----Model Individual* Combined** Individual* Combined**

---------------------------------------------

TC 12.5 (0.006) 11.3 (0.011)

TG 32.5 (<0.001) 14.7 (0.002)

1/HDL 29.1 (<0.001) 8.6 (0.035)

Non-HDL 19.6 (<0.001) 10.9 (0.012)

TC/HDL 29.6 (<0.001) 14.6 (0.002)

TG/HDL 38.5 (<0.001) 13.6 (0.004)

TC 8.1 (0.044) 46.5 (<0.001) 9.4 (0.02) 23 (<0.001)

TG/HDL 34.3 (<0.001) 11.8 (0.008)

TG 12.7 (0.005) 42.2 (<0.001) 8.2 (0.042) 22.8 (<0.001)

TC/HDL 10.6 (0.014) 8.1 (0.04)

TG 19.8 (<0.001) 39.3 (<0.001) 10 (0.02) 20.9 (<0.001)

Non-HDL 7.1 (0.07) 6.1 (0.11)

TC 6.6 (0.08) 39.0 (<0.001) 8.6 (0.035) 23.4 (<0.001)

TG 26.7 (<0.001) 12.1 (0.007)

TG 15.4 (0.002) 44.3 (<0.001) 10.7 (0.014) 19.4 (0.004)

1/HDL 11.7 (0.008) 4.5 (0.21)

TC 15.3 (0.0016) 44.8 (<0.001) 11.8 (0.008) 20.5 (0.002)

1/HDL 32.1 (<0.001) 9.3 (0.026)

TC 9.6 (0.022) 54.1 (<0.001) 9.2 (0.026) 28.6 (<0.001)

TG 9.7 (0.021) 8.2 (0.043)

1/HDL 14.8 (0.002) 5.2 (0.16)

---------------------------------------------------

TC, total cholesterol; TG, triglycerides; HDL, high density lipoprotein

cholesterol.

All the models include study, sex, age at risk, systolic blood pressure, and

current

smoking.

*Test statistic and p-value for the effect of each lipid variable over and above

the

other variables in the model.

**Test statistic and p-value for the effect of all the lipid variables in the

model

over and above the other variables in the model.

The likelihood ratio statistics indicate that improvements

in the ability to predict CHD death could be achieved

to a similar level by adding either TG, HDL, or TG/HDL

to a model that already includes TC. When ANZ cohorts

were examined, a similar ranking of lipid variables was

observed. A similar ranking across the lipid variables was

also observed for the prediction of CVD death.

Analyses on calculated LDL cholesterol were restricted

to a subset of 57,241 individuals with TG level <4.4 mmol/

L. The overall mean (SD) LDL concentration was 3.23

(1.14). The hazard ratio for CHD death associated with

a one standard deviation higher level of usual LDL

cholesterol (HR 1.35; 95% CI, 1.13–1.61) was similar to

that of TC. The hazard ratio associated with the highest

versus the lowest fourth of calculated LDL was 1.55 (95%CI,

1.10–2.18). LDL was similar to TC in predicting CHD

risk (chi squared = 7 vs. 11), as was LDL/HDL ratio compared with

TC/HDL ratio (chi squared = 24 vs. 26). When analyses excluded

the 18% of participants in whom fasting status was not

confirmed, the results were virtually identical.

Subgroups of Participants

The interaction terms for sex and each of the lipid variables

included in the models were all non-significant,

indicating no differences in the association with CHD by

gender. A similar ranking was observed for the likelihood

ratios, such that TG and TG/HDL were best predictors of

CHD risk for both men and women. There was also no

evidence of heterogeneity in the associations with CHD by

age. Because each study did not necessarily contribute to all

age groups, the results of analyses ranking lipid variables

according to ability to predict CHD risk by age were highly

dependent on the age cut-point used. This suggests that the

results are unreliable and that the study is underpowered to

adequately examine the question of differences in risk

prediction by age.

Interactions between Lipids

There were no significant interactions between any of the

lipid variables examined, indicating that the effect of each

variable on the risk of CHD death was not modified by the

level of the other(s). Figure 3 illustrates an example of this

with TC and TG/HDL, which, when combined in a single

model, had the highest likelihood ratio statistic. The slopes

of the association between TC and CHD death are not

significantly different across increasing levels of TG/HDL.

DISCUSSION

The results of this meta-analysis indicate that total

cholesterol, LDL, HDL, and triglycerides are each in-dependent

risk factors for coronary heart disease, and the

association of each with the risk of CHD death is of a similar

magnitude. However, as a single lipid variable, TG is the

best predictor of CHD risk, to similar degree as HDL and

the derived TC/HDL ratio. The TG/HDL ratio was even

more informative. Combination of TC with TG and HDL

separately provided the largest likelihood ratio chi-square

statistic. However, whereas there was little to gain from

addition of another lipid measurement to TG or HDL, this is

not the case for TC. While these results do not inform

about biological mechanisms, they reflect the possible role

of these variables as markers of cardiovascular risk.

As the principal lipoprotein considered atherogenic,

LDL remains the primary target for intervention and

determines recommendation for lifestyle change or drug

treatment in most lipid lowering guidelines (1, 19, 20). This

position has been reinforced by the randomized trial

evidence of the efficacy of LDL lowering in reducing CHD

risk (21–23). However, given the complex metabolic

interactions between lipoprotein particles, proven effec-tiveness

of lowering a particular lipid variable does not

necessarily equate with utility in risk prediction. The

current analyses indicate that LDL alone or LDL/HDL ratio

provides less information about CHD risk than TC alone

or TC/HDL ratio, and even less than HDL, non-HDL, TG,

or TG/HDL ratio. These conclusions are based on calculated

LDL, which is most commonly used in routine clinical

practice. Whether the results are applicable to directly

measured LDL is unknown.

Much interest relating to the use of lipids in risk

prediction has centered on trying to capture information

simultaneously about cholesterol carried on both unfavor-able

and protective lipoproteins. The lipid ratios TC/HDL

and TG/HDL, in particular, attempt to encompass this

information, and the former has found a place in many lipid

lowering guidelines (19). Consistent with other studies, the

current analyses indicate that TC/HDL was more useful

than TC alone, both in terms of strength of association and

goodness-of-fit (6, 24, 25). The superiority of TC/HDL over

LDL/HDL has been reported elsewhere, and possibly reflects

incorporation of triglyceride-rich lipoproteins (e.g., VLDL)

in the former measure (6, 8). TG/HDL ratio is relatively

novel, but in at least one other study has been shown to

identify a particularly high risk group (26). Our analyses

suggest that TG/HDL is better than TG alone and that it

provided better risk prediction than TG when each was

separately combined with TC. Non-HDL is another derived

variable that has recently been advocated as a representa-

tion of cholesterol carried on atherogenic particles, and

is recognized in current North American guidelines as

a secondary target for intervention among individuals with

elevated triglyceride levels (1, 27). Some data have

suggested that non-HDL is superior to LDL-cholesterol,

total cholesterol, and the ratio of either to HDL-cholesterol

in conveying information about CHD risk (4). However, we

found that while non-HDL was better than TC, it was not

as good as the TC/HDL ratio in explaining CHD risk.

While the use of total cholesterol, LDL, and HDL in risk

prediction has been well accepted, the role of triglycerides is

less well-defined. This is because of the difficulty in

establishing an independent association between triglycer-ides

and cardiovascular risk in epidemiological studies,

although more recent investigations have supported this (2,

28), particularly after adjustment for regression dilution

(15). Other studies have raised the potential importance of

TG, both as a marker of CHD risk, as well as for predicting

response to therapy in reducing this risk (29–32). Any

controversy about the independent nature of the association

between triglyceride level and coronary disease should not

detract from its possible importance as marker of CHD risk.

Unlike some previous studies, we did not find any reliable

evidence of differences in the utility of lipid variables in

subgroups defined by gender or by age. In particular, TG was

similarly associated with the risk of CHD death in men and

women, and explained risk to a similar extent. In women,

the associations for TC and LDL may be confounded by

menopausal status (33, 34); however, this was not measured

in the individual studies and cannot be accounted for. Other

analyses have suggested that LDL cholesterol is of limited

value in stratifying coronary risk among the elderly (8).We

found that LDL was a relatively poor predictor of risk among

the entire study population, and we could not reliably

demonstrate that this variable performed worse among the

elderly. However, all subgroup analyses are hampered by

small numbers of events, thus limiting the conclusions that

can be drawn. A further important finding of this meta-analysis

is the lack of any clinically meaningful statistical

interactions between the lipid variables, although, again, we

have limited power to assess this. While some other reports

have suggested effect modification, particularly between

total cholesterol or LDL and triglycerides or TG/HDL ratio,

these were in even smaller studies or involved post-hoc

analyses of clinical trials (30).

Despite the large size of this meta-analysis, there are some

important limitations. Foremost is the lack of standardiza-tion

in ascertainment of baseline information and outcomes.

While we cannot systematically account for differences

between studies, the demonstrated lack of statistical

heterogeneity in the results between cohorts provides

some reassurance that this is unlikely to have introduced

substantial bias.

The current analyses support current guidelines that

advocate measurement of total cholesterol, HDL, and

triglycerides. In most individuals, LDL can be calculated

with this information, but the knowledge of calculated LDL

does not appear to add any value to knowing total

cholesterol. The findings relating to TG and TG/HDL as

a marker of CHD risk warrant further investigation in other

similarly large studies, even if LDL reduction remains the

primary target of intervention.

Al Pater, PhD; email: old542000@...

____________________________________________________

Sports

Rekindle the Rivalries. Sign up for Fantasy Football

http://football.fantasysports.