BSc (Honours) in Computer Science and Mathematics
Cotton Tree Medical Centre, 27 Cotton Tree Parade, Cotton Tree (Queensland)Australia 4558
Phone: 07 54432880 Fax: 07 54438770 Email: Pampollard@universal.net.au

Abstract

Concerns were raised that total cholesterol readings do not reflect cardiac (cholesterol) risk in female patients, particularly pre-menopausal ladies. Consequently for three years Dr Pollard specifically requested the breakdown of the patient's cholesterol. The data used came from patients treated by Dr Pollard between 1996-2000, thus forming a random sample. The data was categorised by sex, age, total cholesterol reading and risk ratio- where the risk ratio is derived from HDL and LDL readings. The study looks at the strengths and weaknesses of three standard risk calculation guidelines. Two of the guidelines found 17% of the patients required drug treatment. The third guideline - Lipid Risk Ratio, produced an extra 7% who were distributed in the hazy area of cholesterol readings between 5 and 7. Analysis of this group revealed the importance of the HDL and LDL levels, especially in younger female patients.

Introduction

Over the years it was observed that total cholesterol readings were extremely variable, especially in female patients. An old rule of thumb uses your chronological age. A decimal point placed between the digits indicates your acceptable cholesterol reading. (ie. If a patient’s age is 53, then 5.3 is an acceptable cholesterol reading.) This acknowledged that cholesterol levels rose with age. However the problem became more complex when High Density Lipid (HDL) and Low Density Lipid (LDL) readings were introduced in the late 1980’s. The local pathology laboratory had taken the evidence from the Multiple Risk Factor Intervention Trial [1] and derived a Lipid Risk Ratio (LRR) which represented the ratio between the protective cholesterol-HDL and atherogenic cholesterol-LDL, with a ratio of 1 being the mean. Consideration of treatment was recommended if the ratio was greater than 1. It was observed that females with high total cholesterols (ie. above 6) often had low risk ratios.

Besides the Lipid Risk Ratio there exists another two predominate cholesterol guidelines, the Pharmaceutical Benefits Schedule and the National Heart Foundation’s. The Pharmaceutical Benefits Schedule (PBS) [2] guidelines predominantly referred to total cholesterol being age and sex dependent but acknowledges that a low HDL in total cholesterol is a risk and requires drug treatment. The National Heart Foundation’s (NHF) [3] guidelines included total cholesterol and LDL as determinants in their recommendations.

Please note that this study only looked at cholesterol and all other risk factors of heart disease should be considered for each individual patient.

Method

Between 1996 and 2000 patients who required a blood test had a total cholesterol with HDL and LDL included. Only initial readings were included in the survey because some patients required treatment and follow up blood tests. The data was categorised for sex, age, total cholesterol, HDL, LDL and LRR. Each guideline was applied :-

Pharmaceutical Benefits Schedule (PBS)

• Cholesterol >6.5 and HDL <1
• Cholesterol >7.5 in 35-75 year old males
• Cholesterol > 7.5 in 50-75 year old females
• Cholesterol >9 in <50 and >75 year old females

National Heart Foundation (NHF)

• Cholesterol > 7
• LDL >5

Queensland Medical Laboratory (LRR)

• Lipid Risk Ratio > 1 (See Appendix 1 for the full Lipid Risk Ratio calculation criteria.)

Results

A sample of 430 females was generated from the original results supplied by the local pathology laboratory (ie. the first entry with a cholesterol and a HDL, LDL reading). When the PBS criteria was applied to this sample, 28 patients were found to require treatment. When the NHF criteria was applied, then 71 patients were found to need treatment, with the PBS group being within this group. (See Figure 1.) Therefore 43 extra patients were picked up for treatment. When the LRR criteria was applied, 77 patients were found to have a risk greater than 1. Of these 46 were in common with the NHF group and only 12 of these were in common with both the NHF and PBS groups. This left 31 other patients requiring treatment. (See Figure 2.) In total, using a combination of all criteria, 102 patients required treatment.

Figure 1.

Figure 2.

Because patients believe their cholesterol reading should be below 5 or 6, the data was looked at as a line graph. The 430 female patients were graphed according to their cholesterol level (less then 5 and greater than 5, 6, 7 and 7.5 respectively). The group of patients with cholesterol readings less than 5, generally, would not require treatment, whereas the groups with cholesterol greater than 7 would have been treated under the PBS and NHF guidelines. This left a hazy group in the 5 to 7 range.

Interestingly when the LRR patients were plotted onto this line graph, 61% were in this hazy area. (See Figure 3.) More surprising was that 5% (4 patients) had a cholesterol reading less then 5 and would not have being considered to need treatment with such a low level of cholesterol. This point was also concluded in a male based study by Danish researchers which showed that patients with a low cholesterol and a low HDL were at significant risk of a heart attack and required aggressive treatment. [4]

Figure 3.

The 430 female sample and the patients with LRR greater than 1 were then subdivided into three age groups, less than 50, between 50 and 75, and over 75, with the same cholesterol groups as above. (See Figure 4). Because the influence of oestrogen raises HDL levels, only 42% of the LRR group in the pre-menopausal group (age < 50) were in the hazy area. However, reflecting the total sample finding of 61%, the group aged between 50 – 75 had 59% of the LRR group in the hazy area.

In the older post–menopausal group (age > 75), where the influence of oestrogen has declined or been lost, there were 79% of the LRR group in the hazy area.

Figure 4.

Age < 50

Age 50 – 75

Age > 75

The survey also generated a sample of 49 males. The PBS criteria found 4 and the NHF criteria found an additional 1. The LRR found 8, with only 3 being in common with the other groups. This sample was too small to form any valid conclusions.

Conclusion

By applying the three prevalent guidelines the group of patients requiring treatment was reduced to102 patients. However the difficulty lies in knowing which guideline to use, since the LRR and the NHF guidelines had only 45% of patients in common. Of the patients with LRR > 1, 61% of them fell in the hazy total cholesterol area of 5 to 7. However 5% of patients with LRR > 1 also had cholesterol reading less than 5. This means that patients with cholesterol less than 5 may still be at risk if they have a low HDL reading; a factor only taken into account by the LRR guideline. (PBS only looks at the HDL reading if the cholesterol reading is greater than 6.5).

It can be concluded that the younger a patient is and especially if they are pre-menopausal, then higher cholesterol levels don’t have to indicate increased risk of heart disease, but as patients become more post-menopausal, then lower cholesterol levels can constitute increased risk.

The hazy area is not well covered by either the PBS or NHS guidelines. The LRR gave more useful information and assisted the doctor in deciding when to provide lifestyle and dietary advice or to intervene with drug therapy.

Patient perception that their cholesterol reading should be below 5 or 6 is a common difficulty dealt with in general practice. With breakdowns being available it is possible to show people that if there is a lot of protective cholesterol in their total cholesterol reading then they aren’t at an increased risk of heart disease. This is particularly true with female patients, especially pre-menopausal women (as it is now known that oestrogen increases levels of HDL). It is comforting to reassure these patients and not instigate drug therapy.

The male sample was too small for conclusive answers but would be an area for future research.

Also in the future I would like to investigate further what happens to the ratio of HDL and LDL in total cholesterol when drug therapy is started. Again the same problem, total cholesterol levels are advised, but are we keeping the LRR down?

References

1. Queensland Medical Laboratory. West End Qld.4101. Biochemistry Department – Dr. Charles Appleton. Lipid Risk Index.

2. Commonwealth of Australia, (2001) Schedule of Pharmaceutical Benefits. p111, J.S. McMillan Printing Group, N.S.W.

3. Heart Foundation of Australia http://www.heartfoundation.com.au/prof/index_fr.html

4. Mansberg, Dr G. (2001) Medical Observer. p5, Medical Observer Pty. Ltd. Sydney.

Appendix 1

Lipid Risk Ratio Calculation Formulas were derived by QML from “Multiple Risk Factor Intervention Trial” New England Journal from the late 1980’s.

To Calculate Lipid Risk Ratio.

Assume Female HDL Mean: 1.3 mmol/L
Assume Male HDL Mean: 1.1 mmol/L
Assume Female/Male LDL Mean: 4.0 mmol/L

Lipid Risk Ratio = HDL Risk *LDL Risk

If HDL < 0.4mmol/L:                 Then no risk is calculated
If cholesterol is < 4.5mmol/L: Then no risk is calculated

To calculate HDL risk.

Female
If HDL is < 1.3 mmol/L: Then HDL Risk = (1+(1.3-HDL))
If HDL is ≥ 1.3 mmol/L: Then HDL Risk = (1/(1+HDL-1.3)))

Male
If HDL is < 1.1 mmol/L: Then HDL Risk = (1+(1.1-HDL))
If HDL is ≥ 1.1 mmol/L: Then HDL Risk = (1/(1+HDL-1.1)))

To calculate LDL Risk.

Male and female
If age is > 65 years:   Then LDL Risk =1.0
LDL is > 4.0mmol/L:  Then LDL Risk = 1+((LDL-4.0)*0.4)
If LDL is ≤4.0 mmol/L:  Then LDL Risk= 1/(1+(4.0-LDL)*0.4)
If age 56-65 years:  Then LDL Risk = ((LDLRisk-1)*(65-age)/10)+1

If LDL Risk is <0.1 :      Then LDL Risk = 0.1