While treatments for some metabolic disorders have been fully researched in HIV-negative populations, there is little information available about the effectiveness of these treatments among people on antiretroviral therapy. Currently no strategy has been proven wholly effective for the treatment of lipodystrophy, lipoatrophy and metabolic changes among people on antiretroviral therapy.

Thus this section contains experimental treatment information drawn from clinical trials, anecdotal reports and treatment strategies suggested by leading doctors in the field. Trial data on the effectiveness of switching anti-HIV drugs is discussed in detail in Treating body fat and metabolic changes - switching drugs.

Why treat fat and metabolic changes?

Some clinicians have questioned the need to treat lipodystrophy if an individual continues to maintain an undetectable viral load on treatment, and have described the manifestations as a cosmetic problem.

A lot of people with HIV do not see it this way. People with HIV frequently refer to the way in which lipodystrophy and lipoatrophy mark them out as people with HIV, increasing the stigma of the infection once more.

There are also concerns that the metabolic changes apparent on HAART may increase the risk of heart disease in some individuals, reducing the long-term benefits of HAART. Metabolic changes may also cause diabetes, pancreatitis and other life-threatening conditions and thus should be treated.

Treating central fat accumulation and 'buffalo hump'

There is no proven treatment for body fat irregularities associated with anti-HIV treatment. However, a number of treatment strategies have been used with varying success: the anti-diabetes drug metformin, a synthetic form of human growth hormone, anabolic steroids, and liposuction.

Human growth hormone

One of the most promising treatments is recombinant (synthetic) human growth hormone (HGH). New York doctor Gabriel Torres has investigated its use to reduce 'buffalo hump' and abdominal fat. He initially found that 11 of 13 treated patients had substantial improvements in fat accumulations. However, several people who stopped HGH treatment experienced a return of fat deposits, suggesting that HGH may need to be taken for an indefinite period. Subsequent reports by Dr Torres indicate that individuals treated early with HGH had improvement in abdominal fat mass, breast size and buffalo hump; wasting in the face and limbs was not affected by HGH (Havlir 2001).

Two larger, randomised studies of HGH presented at the 2002 International AIDS Conference confirmed that HGH is effective in reducing abdominal fat and boosting lean body mass or muscle (Moyle 2002; Kotler 2002). For instance, Dr Graeme Moyle of the Chelsea and Westminster Hospital, London, reported that 6mg of HGH daily produced an average increase in lean body mass of 5.2kg while overall weight increased by 2.91kg and trunk fat declined by 1kg. American researcher Dr Don Kotler reported that visceral abdominal fat (that is, fat surrounding organs rather than fat under the skin) fell significantly in people treated with 4kg HGH daily for 12 weeks.

Other studies have found a partial response in some people treated with HGH. As with the Torres study, response is most likely with buffalo hump and/or abdominal fat. Growth hormone secretion has been shown to be lower in lipodystrophic people with central fat accumulation. HGH has already been used with success in the treatment of abdominal fat accumulations seen in other hormonal disorders. Increases in lean body mass have also been reported (Tai 2002).

Dr Kotler has argued that HGH is unique as a therapy for lipodystrophy because, unlike anabolic steroids, it preferentially depletes central fat tissue to a greater degree than subcutaneous fat in the face, arms and legs. Ongoing research is assessing the long-term efficacy and safety of low dose HGH.

A key disadvantage with HGH is the frequency of side-effects. For example, Dr Moyle reported that one-third of participants receiving the 6mg daily dose reported joint pain, and a third also experienced hand and foot swelling. HGH is also associated with increases in insulin resistance in the short term. In a study of 7 individuals with HIV-associated lipodystrophy who received 3mg per day for six months, glucose levels and insulin resistance rose significantly in the early months but had returned to baseline by month six. One patient with high baseline insulin levels developed diabetes during the study, suggesting that HGH may not be an appropriate treatment for central fat accumulation when insulin levels are high. Patients in this study also lost an average of 0.7kg of sub-cutaneous fat, which may be unwelcome in those with severe fat loss (although lean muscle mass increased by 5.4 kg during the same period) (Lo 2001).

Human growth hormone has a controversial history in the HIV field due to its very high price when it was marketed as a treatment for AIDS wasting in the US. According to US online pharmacy drugstore.com the monthly cost of a 12 week course of HGH (6mg daily) would be US$6917.40 (£4717.91). Its cost makes widespread access to rHGH through the National Health Service in Britain unlikely. For further discussion see Human growth hormone in Drugs used by people with HIV.

Diabetes drugs

The anti-diabetes drug metformin has been effective in reducing abdominal, visceral fat and insulin resistance. See the section on insulin resistance below for details.

Steroids

Anabolic steroids are another experimental treatment for lipodystrophy as well as a standard treatment for HIV-related wasting. French researchers reported on a man receiving treatment with AZT/3TC who developed a buffalo hump and insulin resistance nine months after beginning treatment. He received intramuscular testosterone cypionate biweekly for four months and lost abdominal fat and gained lean muscle mass (Saint-Marc 1998). Furthermore, studies have found that anabolic steroids produce weight and lean body mass increases in people with HIV-related wasting. Steroids may disguise the visible signs of lipodystrophy rather than stop or reverse loss of fat tissue.

A randomised study of oxymetholone treatment in 92 individuals with weight loss due to HIV wasting or lipodystrophy found that the anabolic steroid had no impact on total body fat after 16 weeks of follow-up, although weight and muscle increased (Hengge 2002).

Liposuction and surgery

Removal of abdominal fat by liposuction is regarded as risky because liposuction is a technique designed to remove fat under the skin, rather than the harder visceral fat which often develops with lipodystrophy. However, it appears to be an effective strategy against 'buffalo hump'.

Several groups have now reported on the use of liposuction and other surgical techniques to remove buffalo humps. San Francisco researchers have reported on 25 liposuctions carried out in 23 patients; whilst early results of buffalo hump removal were described as excellent or good by visual examination in 24 of 25 cases; six month follow-up revealed recurrence in seven cases and poor outcome in three cases. Major complications of surgery occurred in five cases (including infection, anaemia and pancreatitis) and nine patients experienced minor complications (DeWeese 2003).

New York researchers reported on liposuction in ten patients with buffalo humps. Liposuction initially appeared somewhat successful (although without complete disappearance of the fat pad), but five buffalo humps subsequently recurred. Two patients, both immunised twice before against pneumococcus, developed pneumococal pneumonia and bacteremia within three months of liposuction (Piliero 2003).

Italian investigators followed 18 patients (14 of whom were men) who had either liposuction (15) or surgical fat removal (3) to treat buffalo hump. As well as experiencing changes in their body shape, all the patients in the Italian case series suffered from restricted neck movement, abnormal posture, and back pain (Gervasoni 2004).

Buffalo hump recurred in only one person after an average post-surgery follow-up of 19 months. The relapse patient had received liposuction and was taking a regimen including d4T, 3TC and efavirenz.

This study favoured the use of liposuction over surgery. The disadvantages of surgical removal of fat included scarring, long operating time, a higher risk of post-operative complications and a long recovery time. By contrast liposuction was found to be simpler and more efficient, having the advantage of good hump correction, shorter operating time, less scarring, the risk of fewer complications, and quicker recovery.

In particular, this study highlighted the advantages of the 'wet form' of liposuction. This technique involves the injection of very large volumes of a solution containing lidocaine and epinephrine into the fat mass in order to allow the fibrous component to be broken down more easily by a surgical scoop called a curette. This technique has several advantages: it enhances local anaesthesia, reduces post-operative pain, reduces blood loss, and ensures a smoother post-operative skin surface.

Niacin

A B vitamin known as niacin is also being used to treat metabolic and fat disorders, with encouraging preliminary results. Niacin is given to people with elevated cholesterol to increase levels of `good` HDL cholesterol. In an open label study at Kaiser Permanente in San Francisco, sixteen participants received an average of 3.00g a day. After a mean duration of one year, 81% of patients had experienced reductions in intra-abdominal fat as measured by a single slice abdominal CT scan (the method agreed to provide an objective measure of changes in body fat). The average reduction in those who experienced improvement was 27%, and the degree of fat loss was significantly associated with the degree of increase in HDL cholesterol and a reduced Total Cholesterol/HDL cholesterol ratio (Fessel 2002).

Not all patients were able to tolerate niacin, due to side-effects at high doses which include flushing, tingling and burning sensations, especially in the upper body, as well as nausea, diarrhoea and headaches. Doses of 2.00g or higher should not be attempted without medical advice, and people with irregular heart beats should consult their doctor before taking large doses of niacin, due to the potential for altered heart rate. Another study, in fourteen patients, found that after 14 weeks of niacin 2000mg, insulin sensitivity had significantly worsened. Body composition was not measured, but central fat accumulation might be expected to worsen if insulin sensitivity declines (Gerber 2003).

Treating facial fat loss

Several cosmetic procedures have been tried in people with facial wasting, such as the injection of a substance called polylactic acid or New Fill. See Treating facial wasting for a discussion of research into these procedures, as well as information about how you might access these experimental procedures in the United Kingdom.

Treating fat loss: switching drugs

See Treating body fat and metabolic changes - switching drugs in Anti-HIV therapy: Body fat and metabolic changes whilst on treatment for further details.

Treating fat loss: other treatment approaches

A new class of anti-diabetes drugs, called thiazolidinediones or glitazones, has been tested as a treatment for lipodystrophy, and fat wasting in particular. Recent research has dashed hopes that the glitazones may reduce fat wasting associated with antiretroviral therapy. Nevertheless, this drug class may have some utility in HIV-infected people in improving insulin sensitivity and liver functioning.

The glitazones are insulin-sensitising agents which, in HIV-negative people, encourage the storage of fatty acids in peripheral fat tissue. Another name for the glitazones is PPAR-gamma agonists. This refers to the activity of the glitazones in stimulating PPAR-gamma an important receptor in the development and functioning of fat cells.

The first drug of the class on the market, known as troglitazone (Rezulin) was tested in people with an inherited form of lipodystrophy called Dunnigan's syndrome, and found to improve lipoatrophy. However, due to several toxicity-related deaths, troglitazone was withdrawn from the market in May 2000. Subsequently, two other drugs in this class have been given marketing approval in Europe for the treatment of diabetes: rosiglitazone (Avandia) and pioglitazone (Actos).

A small randomised study of 27 individuals found some improvement in fat loss after rosiglitazone treatment. The patients were randomised to receive either oral rosiglitazone (4mg daily) or placebo for three months. At the end of the study, the patients who had received the drug showed a 26% improvement in their sensitivity to insulin, measured by a standard test called an insulin clamp. This is compared to a 7% decrease in patients assigned to the placebo group (p = 0.02).

The patients taking rosiglitazone also showed significant increases in peripheral fat, measured by computerised tomography. Fat measurements in the leg increased by 24% in the treatment group but only 2% in the placebo group (p = 0.02). However, the drug did not cause any significant changes in the volume of subcutaneous or visceral fat tissue in the abdomen, or in body mass index.

Patients treated with rosiglitazone reported a significant improvement in lipoatrophy compared to the control group (p = 0.02), although the report does not distinguish between body regions. It is not possible, therefore, to determine whether the drug may improve fat loss from the face or from the limbs (Hadigan 2004).

Another study, which compared rosiglitazone with metformin, found that individuals randomised to receive rosiglitazone were significantly more likely to gain both subcutaneous and visceral fat (+2kg) than the metformin group (-1kg) over 24 weeks. A total of 47% of individuals taking rosiglitazone said that they thought that there had been an improvement in their lipodystrophy compared to 22% of individuals receiving metformin (Van Wijk 2005). However metformin had a more substantial corrective effect on triglyceride and total lipid levels.

However these results are contradicted by a randomised, placebo-controlled study conducted in Australia. This study found that rosiglitazone had no impact on fat wasting. One hundred and eight HIV-positive people with lipoatrophy were randomised to rosiglitazone 4mg twice daily or placebo for 48 weeks. The study was powered to detect as little as a half kilogram difference in limb fat between the groups but no difference was seen. Limb fat increased by 0·14kg and 0·18kg in the rosiglitazone and placebo groups, respectively (Carr 2004). Moreover, rosiglitazone had no significant impact on any other measure of lipodystrophy.

The non-existent impact on lipoatrophy seemed to be at odds with the finding that rosiglitazone boosted levels of adiponectin - a cellular messenger involved in fat cell maturation and function. Rosiglitazone is also known to boost the activity of PPAR-gamma, which is known to be down-regulated by d4T.

Researchers do not understand exactly why these effects failed to translate into an increase in fat cells. They have speculated that the negative impact of therapy may destroy new fat cells, or that too few PPAR-gamma receptors may exist as targets for rosiglitazone, or that lipoatrophy may not be the result of interference with PPAR-gamma after all (Moyle 2004).

Background therapy had no impact on response to rosiglitazone, although there was some suggestion that ongoing use of AZT or d4T was associated with a poorer response to rosiglitazone. Comparing the presence or absence of AZT or d4T among people taking rosiglitazone, there was a significant difference in fat gain at week 24 (no AZT or d4T +0.48 kg versus AZT or d4T -0·06 kg; p=0·05) but this difference was not maintained at week 48. The authors dismissed this finding as either an error or a short-term effect with no clinical relevance.

On a positive note, rosiglitazone did improve insulin sensitivity and liver function. Unfortunately, the most common side-effects seen among those taking rosiglitazone were elevated triglycerides and elevated cholesterol. The Finnish study, described below, also found that rosiglitazone caused significant increases in cholesterol and triglyceride levels.

A precursory study was conducted in Finland involving 25 men and five women with lipodystrophy on stable HAART (Sutinen 2002). Participants were randomised in a blinded fashion to receive either 8mg rosiglitazone or matching placebo for the 24 week study period. Lipodystrophic parameters were measured at baseline and at 24 weeks by a series of sixteen MRI scans around the body, spectroscopy of liver fat, skinfold thickness, serum leptin (a correlate of body fat mass), bioimpedence analysis, and self-assessment. Blood lipid and liver enzyme levels were also determined.

At the end of the 24 week treatment period, there were no changes observed in any morphologic parameter in either treatment arm. In other words, there was no evidence that rosiglitazone improved body fat changes. Self-assessment showed improvements in both fat loss and gain, but this only reached statistical significance in those receiving the inactive placebo. Rosiglitazone appeared to improve insulin resistance.

At the time of presentation, these negative findings were attributed to inadequate dosing, inadequate treatment duration, inadequate powering, and the small sample size. However, the Australian study provides confirmation that this earlier study was, in fact, accurate.

Another study which reported a negative finding with a glitazone involved the drug pioglitazone. A small, observational study involving nine HIV-infected people with lipodystrophy, this study found no significant changes in body fat or metabolic measures, although there were trends towards greater fat mass and lower triglycerides (Calmy 2002).

Earlier work suggested that the glitazones would have a positive effect on fat wasting in people taking antiretroviral therapy. Two small, non-randomised studies reported significant increases in limb or facial fat in a small number of patients (Visnergarwala 2001; Gelato 2002), and a 12-week, controlled study found some evidence that rosiglitazone could encourage expansion of body fat despite ongoing antiretroviral therapy.

Conducted in 28 HIV-positive individuals with lipoatrophy and hyperinsulimia, subcutaneous fat increased by 8% in the treatment group compared with negligible change in the placebo group (p = 0.001). After a further three months of open-label rosiglitazone treatment at 8mg/day, those who had received rosiglitazone throughout the study had experienced a 12% increase in subcutaneous fat mass from baseline, and self-reported lipoatrophy improved significantly (p = 0.03). Insulin sensitivity also improved significantly at month 3 and this was sustained at month 6 (Hadigan 2003).

Recently, a dietary supplement called NucleomaxX has also been shown to be effective in restoring fat loss in patients taking AZT or d4T. NucleomaxX is an extract of sugar cane that can raise levels of uridine in the body, which may block the toxicity caused by NRTIs and allow fat cells to regenerate.

Two studies have shown some benefits of the supplement in patients with fat loss who were taking AZT or d4T. One demonstrated that patients taking NucleomaxX alongside an anti-HIV drug regimen containing AZT or d4T gained more subcutaneous fat than patients taking a placebo supplement (Sutinen 2005). The second, open-label study showed an improvement in subcutaneous fat loss over the body, although this was not coupled with improvements in levels of mitochondrial DNA in fat tissue (McComsey). Larger, longer-term studies of NucleomaxX are underway.

Treating insulin resistance and diabetes

Insulin resistance is seen in `Syndrome X' - a metabolic disorder with similar symptoms to HAART-associated lipodystrophy that occurs in HIV-negative people. Syndrome X may be corrected in some cases by the restoration of insulin sensitivity, using an anti-diabetes drug called metformin.

There is now evidence that metformin may be effective in reducing insulin resistance and abdominal fat in HIV-infected people on HAART. French doctors reported that metformin reduced insulin resistance, reduced triglyceride levels and improved abdominal fat distribution in a randomised study of 25 people receiving PIs (Saint-Marc 1999a). HIV wasting expert, Dr Carl Grunfeld, has also reported that metformin has shown the best results in the treatment of HAART-associated insulin resistance and diabetes, producing an associated decrease in body fat.

There is also evidence from randomised, controlled studies which suggests metformin is an effective treatment for insulin resistance and diabetes, and related metabolic and body fat disorders, seen in HIV-infected people on treatment. For instance, a team from Boston conducted a randomised, controlled study of metformin in 26 HIV-infected individuals with fat redistribution, high insulin levels and/or abnormal glucose tolerance. After three months, insulin levels, weight and blood pressure fell significantly in those treated with metformin compared to controls. The treated group also experienced a reduction in waist size and reduced visceral abdominal fat (Hadigan 2000).

A randomised, placebo controlled comparative study of metformin or gemfibrozil in patients with abdominal fat accumulation and triglyceride levels above 5.2mmol/l found that after one year of treatment, there was no significant improvement in fat redistribution, triglyceride levels or insulin resistance, nor was there any difference between the arms (Martinez 2002). This study should be treated with caution because of the small numbers involved - 15 in each arm.

A third study randomised 37 individuals with baseline fasting insulin level of above 104pmol/l and evidence of fat redistribution to receive metformin at a dose of 500mg twice daily, with dose escalation to 850mg twice daily after two weeks, or the same metformin regimen in combination with an exercise programme consisting of an hour of cardiovascular and weight training three times a week for twelve weeks. The two groups of patients had comparable insulin profiles and cardiovascular disease risk factors at baseline.

Patients in the exercise arm experienced a greater fall in their median systolic (p = 0.012) and diastolic (p = 0.001) blood pressure than individuals in the metformin only arm. In addition, the waist-to-hip ratio improved to a greater extent in the exercise/metformin arm than in the metformin-only arm, as did the fasting insulin (p < 0.05). Thigh muscle area increased to a greater extent in the exercise/metformin arm than in the metformin-only arm (p = 0.015), and so did exercise time (p = 0.045). Both subcutaneous adipose tissue (p = 0.049) and visceral adipose tissue (p = 0.063) tended to decrease in the exercise arm as opposed to the metformin-only arm (Driscoll 2004).

The 2003 British HIV Association guidelines recommend that insulin resistance should be treated with metformin.

The new class of anti-diabetes drugs called glitazones, discussed above, may also be effective for treating diabetes or insulin resistance in HIV-positive people. The glitazones are unlikely to interact with PIs and studies to determine the efficacy of glitazones in the treatment of PI-related diabetes are ongoing (see above).

Discontinuing protease inhibitor treatment may also reduce insulin resistance (Walli 2001).

Human growth hormone (HGH) can also restore insulin sensitivity, and some researchers believe that disruption of insulin-like growth factor 1 and HGH production within the body is at the root of the disorders now being seen. However, HGH may precipitate diabetes in a small proportion of people.

Mineral and vitamin supplements may protect against diabetes in the general population (Ford 2001) but whether supplements have a protective effect in people taking anti-HIV treatments is unknown.

Treating high lipids

The 2003 British HIV Association treatment guidelines recommend action to treat lower lipid levels in people with cholesterol above 6.5mmol/l or triglycerides above 8mmol/l, but in the absence of evidence about the risk of heart disease in HIV-positive individuals, it is unclear whether this advice should apply to everyone with lipid elevations, or just those with other risk factors for heart disease.

Treatment should consist of lipid-lowering drugs, such as statins or fibrates. For further information, see Lipid-lowering drugs in Drugs used by people with HIV: Lipid-lowering drugs.

Risk factors include:

• Age (men - above 45, women - above 55 or premature menopause with no oestrogen replacement).

• Family history of heart disease.

• Current smoker.

• High blood pressure.

• LDL cholesterol below 0.9mmol/l (35mg/dl).

• Diabetes.

HDL cholesterol above 1.55mmol/l (60mg/dl) is assumed to cancel out one other risk factor when assessing the need to intervene.

It is crucial that dietary changes (e.g. reducing fat intake) do not reduce absorption of anti-HIV drugs, so standard cholesterol-lowering advice is not always appropriate. In the United Kingdom, dietitians have recommended that individuals increase their intake of monounsaturated fat (from sources like olive oil) and reduce saturated fat intake, and cut sugar intake in order to reduce triglyceride levels. Omega-3 fish oils, present in oily fish like sardines, mackerel, salmon and pilchards, may offer some protection against heart disease. A small amount of alcohol each day may also improve levels of HDL cholesterol.

If at all possible, individuals with elevated lipids should stop smoking or cut down drastically. The drug buproprion (Zyban) is now available in the United Kingdom to assist people who are attempting to stop, as is nicotine replacement therapy, but many clinics will require individuals to obtain this prescription from their GP.

Dietary changes

Dietary intervention is normally recommended before any other intervention to control high lipid levels in adults without HIV. Similarly, dietary changes are fundamental to the management of adult-onset diabetes. However, there is little evidence that dietary interventions alone are successful in managing lipid elevations, diabetes or lipodystrophy in people with HIV.

The most encouraging evidence comes from a single case study. A man with lipodystrophy, who had been taking nevirapine/d4T/3TC for more than two years, had some features of lipodystrophy and metabolic abnormalities reversed after a four-month structured exercise programme combined with a low fat, high fibre diet. Specifically, the regime reduced visceral fat, cholesterol and insulin resistance, but did not affect the severe wasting in the man's arms and leg (Roubenoff 2002). While this case study is encouraging news, it should be treated with caution. At most, the case study indicates that some people may benefit from this type of exercise and dietary regime.

Recently a Spanish team reported that a low-fat diet in people on HAART with high lipids had only moderate success. Less than half of participants complied with the dietary schedule at six months and the effects were modest. Among the compliant participants, the average cholesterol reduction was only 11% while triglycerides fell by a mean of 23%. After six months, those on a low-fat diet reported an average weight loss of 2kg (Barrios 2002).

A comparative study of dietary advice versus pravastatin and dietary advice found that dietary advice alone had no significant impact on levels of total cholesterol, LDL cholesterol or triglycerides, whereas treatment with the lipid lowering drug pravastatin did result in a significant reduction in lipid and triglyceride levels (Moyle 2001). See below for further discussion of this study. In contrast, another small study found dietary advice produced a trend towards lower lipids of up to 7% after four weeks, and the addition of omega-3 fatty acid supplements further reduced triglycerides levels (Peabody 2002).

Diabetes dietary guidelines may be relevant to people with HIV who have diabetes, insulin resistance or high blood sugar. In the past, a high carbohydrate and low fat diet was recommended to all patients with diabetes but recent evidence has established that some fats actually improve blood fat profile and diabetes control in HIV-negative people. Specifically, monounsaturated fatty acids (which come from olive oil, rapeseed and peanut oils, avocados, and some nuts) and polyunsaturated fats (from corn, sunflower, safflower and soybean oils) are recommended. A diet rich in unrefined carbohydrates (from fresh fruits and vegetables, and whole grains) and high in fibre is encouraged to improve blood sugar control and lower blood lipid concentrations.

According to American diabetes guidelines, other types of fats should still be minimised. Less than 10% of total energy intake should come from saturated fats (found in butter, full fat milk, animal fat and lard). A dietary cholesterol intake of less than 30 grams a day is recommended, and trans-unsaturated fatty acids should be minimised. Trans-fatty acids are often included in margarines made from vegetable oils, and can detected by the words "Hydrogenated fat" in the labelling. Trans fatty acids are also found in many ready prepared foods, biscuits, cakes and pizzas where vegetable oil is used. For the European diabetes dietary guidelines, click here.

It is crucial that dietary changes (e.g. reducing fat intake) do not reduce absorption of anti-HIV drugs, so standard cholesterol-lowering advice is not always appropriate for people with HIV. If you are considering changes to your diet, discussion with your doctor and/or a dietitian is recommended. In the UK, dietitians have recommended that individuals increase their intake of monounsaturated fat (from sources like olive oil) and reduce saturated fat intake, and cut sugar intake in order to reduce triglyceride levels. Omega-3 fish oils, present in oily fish like sardines, mackerel, salmon and pilchards, may offer some protection against heart disease. A small amount of alcohol each day may also improve levels of HDL cholesterol.

There are many misconceptions about the impact of dietary changes on lipid levels, largely fuelled by misleading advertising by food manufacturers. A review of the evidence by the British Dietetic Association highlights the following points:

• Reduction of saturated fat intake reduces the risk of cardiovascular disease, but a reduced fat diet needs to be followed for at least two years, according to large population studies, before it has any long-term impact on your risk of heart disease.

• Fat reduction has no significant effect on the future risk of death from heart attack in people who have already had a heart attack.

• The lipid lowering effect of dietary fat reduction is of the order of 3-6%.

• It is not clear whether monounsaturated fats (olive oil, rapeseed oil) are better than polyunsaturated fats at reducing lipid levels.

• There is no evidence from well designed trials that vitamin supplements, garlic or plant sterols (as found in certain types of specially designed margarine) have any significant effect on lipid levels.

• The levels of soy protein and soluble fibre needed to have a significant impact on cholesterol levels are unrealistically large.

Exercise

Exercise is also known to have an impact on reducing lipid levels. At a minimum, individuals should aim to increase their heart rate for thirty minutes at least three times a week, either by brisk walking, cycling, jogging or swimming. More structured exercise may bring greater benefits: a study of resistance training in men with lipodystrophy showed that triglyceride levels normalised in 9 of 11 individuals who lifted weights at least three times a week (Yarasheski 2001). This is probably because larger muscles consume more triglycerides as fuel, thus removing them from the circulation.

A prospective 10 week study of aerobic exercise and resistance training in 6 HIV-positive individuals with lipodystrophy showed that total cholesterol and triglyceride levels were reduced by 17.6% and 25% respectively. This reduction was statistically significant, and there was a trend towards a reduction in the total cholesterol: HDL cholesterol ratio. There was also a significant improvement in waist-hip ratio (a possible indicator of central fat accumulation), but body fat percentage declined, suggesting that whilst exercise may improve lipid disturbances and central fat accumulation, it will not help peripheral fat wasting. This study did not use DEXA scanning to check whether the reduction in waist-hip ratio was associated with reduced central fat or subcutaneous fat (Jones 2001).

In HIV-negative people insulin sensitivity can be improved by regular exercise, but the effects of exercise (and diet) on insulin resistance in HIV infection have not been reported. A small observational study of an intensive exercise regimen and dietary advice in 15 HIV-infected women produced weight loss but had virtually no effect on insulin resistance (Engelson 2002b). Of note, calorie intake was restricted to 1200mg a day. Weight loss averaged 6kg in this study, almost all fat, but some of this fat was subcutaneous rather than visceral.

Drug therapy for high lipid levels

There is considerable controversy over the significance of cholesterol elevations on HAART, with some researchers arguing that we should expect to see growing numbers of heart attacks and other cardiovascular problems in patients taking protease inhibitors (PIs), while others think it will take years before we know for sure if relatively short-term increases in lipids will translate into an epidemic of heart disease among people with HIV.

See Heart disease and HAART in Anti-HIV therapy: Body fat and metabolic changes whilst on treatment for further discussion of this evidence.

Models developed by Dr Matthias Egger of the University of Bristol show that there are big differences in risk according to age, lifestyle, and lipid profiles between individuals. Using data from a lipodystrophy study, Dr Egger notes that a 50 year old male smoker with diabetes and severe lipodystrophy has a 14% risk of some sort of coronary event within five years, but the risk is considerably smaller for a younger non-smoker with increased lipid levels but no diabetes.

In HIV-negative individuals, 45 men needed to be treated for five years in order to avert one heart attack or diagnosis of angina or death from heart disease in a large randomised study of pravastatin treatment in men with average baseline cholesterol of 7mmol/L, and the study found that pravastatin treatment reduced the risk of death from cardiovascular causes by 22% (Shepherd 1995). In this study, the effects of pravastatin according to baseline risk factors have not been analysed.

A study from Houston, Texas, showed that in patients with high lipid levels, lipid-lowering therapy was only partially successful. Sixty-three consecutive patients were analysed. An average cholesterol reduction of 19% was reported on the first lipid-lowering regimen (predominantly fibrates), and LDL cholesterol levels fell by just 5%. Only 16% of patients who continued protease inhibitor therapy achieved target levels of LDL and total cholesterol after more than one year of lipid-lowering treatment. Presenting the study, Dr Fahmida Visnegarwala said that management of dyslipidemia alone without correcting the underlying metabolic disturbances may not be effective, but a disappointingly small proportion of non-HIV patients (typically less than 40%) ever reach the lipid goals set out at the beginning of lipid-lowering therapy.

However, a randomised Italian study found that lipid-lowering treatment with either a statin or fibrate was more effective than switching from a protease inhibitor to efavirenz or nevirapine in controlling lipid levels (Calza 2005). The study randomised 132 patients with elevated lipid levels to switch from their protease inhibitor-containing regimen to nevirapine or efavirenz, or to stay on their PI-containing regimen and add either pravastatin (Lipostat) or bezafibrate (Bezalip).

Patients treated with lipid-lowering drugs experienced a 44% decrease in triglycerides, compared to a mean decrease of 18% in the NNRTI-treated group. Total cholesterol levels fell more substantially in the patients treated with lipid-lowering drugs than in the NNRTI group too (41 vs. 19%), and a similar pattern held true for low-density lipoprotein (LDL) cholesterol.

With these cautions, people with the following risk factors are more likely to be encouraged to take lipid-lowering drugs and pursue the interventions discussed in previous sub-sections on the basis of current knowledge:

• High blood pressure (hypertension).

• A prior history of cardiovascular disease.

• A family history of heart disease.

• Silent myocardial ischemia.

Treating high lipids: specific drugs

In HIV-negative people with high lipids, several types of drugs are used to lower lipid levels. These include:

• CoA reductase inhibitors, or statins, which reduce cholesterol production in the liver but may interact with protease inhibitors.

• Fibrates, which reduce levels of triglycerides and LDL cholesterol.

• Nicotinic acid or niacin, which reduces the production of triglycerides and cholesterol.

• Fish oils, which contain fatty acids of the omega-3 group which may reduce LDL cholesterol and increase HDL cholesterol at high doses.

The effectiveness of these drugs among HIV-infected people on antiretroviral therapy remains experimental, although the studies summarised below suggest that these treatments may be moderately effective at reducing lipid levels. There is little evidence that lipid-lowering agents have any effect on body fat changes.

A group of drugs called the statins are commonly used to reduce lipid levels. People taking protease inhibitors should discuss drug interactions with their HIV clinician when considering statins because most of the statins are metabolised in a similar way to protease inhibitors - by the cytochrome P450 - which means drug interactions are likely. Cytochrome P450 is a super-family of similar enzymes or proteins which break down drugs in the liver. 3A4 is a member of the P450 family which plays an important role in breaking down statins and protease inhibitors as well as other drugs.

Key statins, atorvastatin, cerivastatin, and simvastatin all use the P450 3A4 pathway. Drug interaction studies have found that blood concentrations of simvastatin and atorvastatin are substantially increased in individuals on ritonavir/saquinavir and nelfinavir and should not be co-administered (Fichtenbaum 2002). Changes in the concentrations of the protease inhibitors are not considered clinically relevant.

Research to date indicates that another statin called pravastatin (metabolised by the P450 3A pathway rather than 3A4) appears to be the safest statin drug for use in combination with protease inhibitors. Exposure to pravastatin is reduced by half when taken concurrently with protease inhibitors (Fichtenbaum 2002). Several studies have found improvements in cholesterol and triglyceride levels after six months treatment with pravastatin, although cholesterol did not normalise when baseline levels were very high (Moyle 2001; 2001b; Baldini 2000; Calza 2002; Negredo 2002c). However, a randomised study in 16 individuals with lipid elevations on HAART found that whilst non-atherogenic LDL and VLDL particles declined after 8 weeks of therapy, other atherogenic particles (large VLDL and small LDL) did not (Stein 2002).

At the Ninth European AIDS Conference in Warsaw in 2003, a team from London's Chelsea & Westminster hospital presented a paper comparing these two cholesterol-lowering drugs in people on both PI- and NNRTI-based antiretroviral therapy. Out of 179 people (102 on atorvastatin and 77 on pravastatin) 82% had a modest decrease in LDL cholesterol: a mean of 1.26mM on atorvastatin and 1.30mM on pravastatin. However, 56% of those on atorvastatin managed to decrease their cholesterol levels to below 6.6 mmol/L compared with 32% of those on pravastatin, suggesting that atorvastatin is somewhat more effective (Smith 2003).

The newest statin on the market, rosuvastatin (Crestor), has been shown to reduce total cholesterol by 21% and triglycerides by 31% in 16 Italian patients with baseline cholesterol above 240mg/dl. Rosuvastatin is not metabolised by the cytochrome p450 pathway, so the risk of drug interactions or increased side effects is limited (Calza 2005b).

United States guidelines on the management of metabolic disorders in HIV infection propose that elevated cholesterol levels should be treated with either of these two options:

• Pravastatin (20 to 40mg daily as a starting dose).

• Atorvastatin (10mg daily as a starting dose).

The guidelines also suggest other options to the two recommended statins, noting that fluvostatin (20 to 40mg daily) is a "reasonable alternative." They add that if statins cannot be used, for example due to drug-drug interactions, then another class of drugs, fibrates, might be considered.

Some doctors argue that since clinical trials in HIV-negative people show that it can take five or six years of statin treatment before any clinical benefit becomes evident, younger people with moderately elevated lipids but no other risk factors should be encouraged to begin statin treatment.

People with antiretroviral therapy-related metabolic disorders tend not to have high triglycerides without also having elevated LDL cholesterol. Sometimes, however, LDL can just be on the wrong side of the upper limit of normal and under these circumstances - when triglyceride levels are above 5.65mM - then fibrate therapy should be used. Fibrates work in several ways, the main effect being to reduce very low density lipoproteins (VLDL) in the blood which are largely made up of triglycerides. The two recommended drugs are gemfibrozil (600mg twice daily 30 minutes before meals) or micronised fenofibrate (54 to 160 mg daily).

An Italian study compared fibrates with statins for the treatment of hypertriglyceridaemia (Calza 2003). One of three different fibrates (bezafibrate, gemfibrozil or fenofibrate) or one of two statins (pravastatin or fluvastatin) were given to 113 PI-based antiretroviral therapy recipients with high triglycerides for a year. Since the numbers on each were small - between 18 and 25 on any one drug - the conclusions reached were no surprise: the fibrates were more effective at reducing triglyceride levels.

Several studies published last year found that both gemfibrozil and fenofibrate not only reduced triglyceride levels but also raised HDL ("good") cholesterol levels (Aberg; Palacios). However, only a minority of those enrolled in these trials actually achieved normal levels of these beneficial lipids.

The fibrates and the statins produce moderate reductions in lipid levels. Generally, gemfibrozil or pravastatin reduce total cholesterol by about 20% and triglycerides by up to 40% in HIV-infected people on treatment (Calza 2002; Moyle 2001b; Negredo 2002c; Baldini 2000) although some reports have indicated greater falls (Hewitt 1999). However, a randomised placebo-controlled study of gemfibrozil in patients with hypertriglyceridemia on HAART showed modest effects on triglyceride levels (-1.22mmol/L after 16 weeks) and only one patient in the gemfibrozil group achieved normalisation of triglyceride levels (Miller 2002).

Combining gemfibrozil with a statin also appears to be safe and moderately effective. For example, the combination of gemfibrozil with another lipid-lowering agent, atorvastatin, has been moderately effective at reducing cholesterol and triglyceride levels among people on protease inhibitors (Henry 1998c, also see Negredo 2002c).

The latest BHIVA guidelines place their focus squarely on the possible interactions between statins and PIs. They agree with the US guidelines that the use of simvastatin and lovastatin in people on PI- or delavirdine-containing HAART is not recommended.

• Atorvastatin levels could be doubled in people on PIs, but both suggest that this drug can be used with caution.

• Pravastatin appears to be safe with PIs, but whereas the UK guidelines say it appears to have a low likelihood of PI interactions, the US guidelines caution that pravastatin may need to be increased when taken along with ritonavir-containing regimens.

• Fluvastatin does not appear to have any PI-related drug-drug interactions.

• Any of the statins are probably safe in efavirenz or nevirapine-containing regimens.

• The US guidelines also suggest that levels of fibrates may be reduced in ritonavir-containing regimens, but otherwise no interactions are known at this time.

Also see Drugs used by people with HIV: Lipid-lowering drugs.

Vitamin supplements such as acetyl-L-carnitine and niacin are also being tried as treatments for lipid abnormalities and lipodystrophy. Niacin is known to reduce triglycerides and LDL or 'bad' cholesterol in non-HIV-infected people with diabetes and artery disease. As mentioned above, niacin has improved lipid levels in people with HIV-related lipodystrophy (Fessel 2002; Gerber 2003), although side-effects are a concern. Carnitine supplementation is primarily being tested as a treatment for peripheral neuropathy (nerve damage) associated with anti-HIV drugs, and there is little evidence about its ability to reduce lipid levels. A small, observational study of L-carnitine 1000mg twice daily for three months in twelve HIV-infected people reduced total cholesterol but had no effect on triglycerides or body fat (Mauss 2001).

A supplement known as Cholestin, which is made from fungus fermented on red rice, has been shown to reduce total and LDL or 'bad' cholesterol in people with HIV-related high lipids when 1200mg is taken twice daily (Keithley 2002). Although marketed as a dietary supplement, it contains natural statins including the agent lovastatin. It is manufactured by Pharmanex Inc. and can be bought over the internet in the UK and Europe from in-house pharmacy websites. A month's supply costs around £40. The supplement is not available on prescription, and its sale has been banned in the USA

There have also been reports of normalisation of lipids following discontinuation of antiretroviral therapy.

See the research sections below for detailed summaries of key studies.

What can you do?

General advice aimed at reducing lipid levels and reducing the chance of heart disease may be considered:

• Exercise regularly, stop smoking, and use alternative forms of birth control to the contraceptive pill (e.g. condoms, caps)

• Resistance and aerobic exercise may increase muscle and body mass.

• Discuss monitoring your triglycerides, cholesterol and blood glucose with your doctor. Sophisticated monitoring of the heart is not generally warranted unless other symptoms or tests suggest heart problems.

• Ask your doctor about monitoring for fat wasting and lipodystrophy.

• Changes to diet, such as reducing fat and sugar intake, may reduce a person's chance of diabetes, elevated blood lipids and heart disease. Of course, dietary changes may be difficult given that all protease inhibitors except indinavir are taken with a high-fat meal. Consult an HIV dietitian to ensure you don't inadvertently lose weight.

• Mild blood fat elevations plus low HDL cholesterol may not respond to a low-fat diet. Ask your dietitian about mono-unsaturated fats.

Research into drug treatments

Hadigan (2000) randomised 26 HIV-infected, non-diabetic people with fat redistribution, high insulin levels and abnormal glucose tolerance to metformin (500mg twice daily) or placebo. After 3 months treatment, the metformin group had improved metabolic and fat parameters: mean weight fell by -1.3 kg vs +1.1 kg (p=0.005); waist circumference fell by -1.1 vs +1.1 cm (p=0.02) mean insulin AUC 120 minutes after glucose tolerance test fell by -20349 vs -2875 pmol/L (p=0.01), and distolic blood pressure fell by -5 vs +5 mm Hg (p=0.009); visceral abdominal fat (VAT) fell by -1115 vs +1191 mm2 (p=0.08), with a proportional reduction in subcutaneous fat (SAT). The VAT/SAT ratio did not change significantly. Total insulin levels fell by about 20%. Metformin was associated with diarrhoea. There was a trend towards stabilisation of lipid levels in the metformin group.

Visnegarwala (2001) analysed 63 patients treated for HAART-related metabolic and fat disorders. Fibrate treatment produced an average cholesterol reduction of 19%, with LDL cholesterol levels falling by just 5%. Only 16% of patients who continued protease inhibitor therapy achieved target levels of LDL and total cholesterol after more than one year of lipid-lowering treatment.

Martinez (2002) randomised 66 HIV-infected people with lipodystrophy to metformin (850mg twice daily) or gemfibrozil (600mg twice daily) or placebo twice daily. Measurements at baseline and every three months showed slight decreases in weight, body mass index, LDL cholesterol, body fat and muscle but there was no significant difference between the three arms.

Gelato (2002) treated 8 HIV-positive people with insulin resistance with rosiglitazone (8mg daily) for 6-12 weeks. Two were also taking atorvastatin or gemfibrozil. Insulin sensitivity rose from an average of 3.8 to 5.99 mg/glucose/kg, still well below the sensitivity seen in the HIV-negative control group. Visceral adipose tissue decreased by 21% and subcutaneous fat rose by 25%. Fasting insulin levels fell significantly. One person ceased rosiglitazone due to liver abnormalities.

Sutinen (2002) randomised 25 men and 5 women with lipodystrophy on stable HAART to receive either 8mg rosiglitazone or matching placebo for 24 weeks. Lipodystrophic parameters were measured at baseline and at 24 weeks by a series of sixteen MRI scans around the body, spectroscopy of liver fat, skinfold thickness, serum leptin (a correlate of body fat mass), bioimpedence analysis, and self-assessment. Blood lipid and liver enzyme levels were also determined. At 24 weeks, there were no changes observed in any morphologic parameter in either treatment arm.

Domingo (2002) conducted a randomised, double-blind study of metformin (500mg twice daily) versus gemfibrozil (900mg daily) in 37 men and 11 women with lipodystrophy. Metformin improved markers of fibrinolysis (clearance of blood clots from circulation) in comparison to gemfibrozil and placebo.

Calmy (2002) treated 9 people with PI-related diabetes with pioglitazone for 6 months. Although 4 individuals reported improvement, there was no significant change in fat mass, triglycerides or leptin levels. Average percentage fat mass rose from 10% to 11% (p=0.1).

Bonnet (2002) conducted a retrospective study of the efficacy of statins and fibrates to treat high lipids in 106 people taking antiretroviral therapy. At 3 months, total cholesterol had fallen significantly by -0.9mmol/L in the statin group and total cholesterol and triglycerides fell significantly by -0.7mmol/L and -2.3mmol/L in the fibrate group.

Calza (2002) treated 106 people with elevated lipids, unresponsive to dietary and exercise interventions, with fibrates or statins for 12 months. Statins reduced triglycerides by 33% and cholesterol by 25%, with no significant difference between pravastatin (20mg daily) and atorvastatin (20mg daily). Fibrates reduced triglycerides and cholesterol by an average of 41% and 22%, respectively, again with no difference between fibrates (benafibrate 400mg daily, gemfibrozil 600mg twice daily, and fenofibrate 200mg daily).

Negredo (2002c) a 17% reduction in total cholesterol and a 41% in triglycerides in 70 HIV-infected people with elevated lipids after 24 weeks of treatment with pravastatin (with 32 also taking gemfibrozil). 64% achieved some improvement, 30% achieved normal values while 20% had no change in total cholesterol.

Moyle (2001b) compared dietary advice or 40mg pravastatin daily in 31 HIV-infected individuals with high cholesterol. Total cholesterol fell by 17% in the pravastin group and by 4% in the dietary group. HDL cholesterol rose slightly in both groups. LDL cholesterol fell by about 20% in the pravastin group. LDL cholesterol remained above normal, possibly due to high baseline levels.

Baldini (2000) treated 19 individuals with 20mg pravastatin. Cholesterol fell by 20 and triglycerides by 37%. One quarter of participants reached normal cholesterol levels.

Schliefer (2000) treated 13 people with high lipids with bezafibrate 400 mg daily for an average of 8.8 months. Average triglycerides and cholesterol fell from 1800 mg/dl and 374 mg/dl at baseline to 516mg/dl and 331mg/dl at week 4 (p=0.004, p=0.006). At week 36, average triglycerides and cholesterol were 391mg/dl and 256mg/dl. Three stopped treatment after switching to a PI-sparing regimen and 1 person developed gastrointestinal side-effects.

Walli (2000) treated 6 people with HAART-associated diabetes and insulin resistance with troglitazone (400mg once daily). Two people had improvement and 2 had full reversal of insulin resistance, while improvements in body fat changes and lipid levels were also reported.

Hewitt (1999) reviewed patient records of 8 men on protease inhibitor combinations who developed high triglycerides (median from 298 to 1,803mg/dl). The median nadir after treatment with gemfibrozil was 300, an 83% reduction. However, triglycerides rose in five men who were switched to nelfinavir. Researchers warned that nelfinavir and gemfibrozil may interact and cause a rebound of the high triglycerides.

Fichtenbaum (2002) tested drug interactions between PIs and statins in HIV-negative people. Ritonavir/saquinavir (400/400mg twice daily) and nelfinavir (1250mg twice daily) interacted with pravastatin (40mg once daily) reducing total exposure to pravastatin by an average of 50%. Simvastatin also interacted with saquinavir/ritonavir and nelfinavir to increase exposure to simvastatin by 30-fold. Atorvastatin levels were also significantly increased. PI changes were not considered clinically relevant.

Curry (1999) treated 12 people for elevated cholesterol and triglycerides after a mean of 14 months of protease inhibitor therapy. Ten were treated with 200mg fenofibrate once a day, and 2 were treated twice daily. Mean cholesterol fell from 11.3 to 8.3mmol/l and mean trigs fell from 14.5 to 7.6mmol/l. Body fat changes were reduced in only one participant.

Saint Marc (1999a) randomised 27 patients with central adiposity to receive either 850mg tid metformin or placebo after eight weeks of observation. In the metformin group, mean serum triglycerides decreased by 30.1% after eight weeks on treatment, compared with no significant change in the placebo group. Plasma glucose, insulin and C peptide fell significantly in the metformin group and returned to normal levels in 3/14. There were no significant differences between the two groups in total cholesterol. Visceral abdominal adipose tissue fell by 37.5% in the metformin group and increased by 10.5% in the control group. Mean total body weight loss in the metformin group was 2.8kg at week 16. Two patients withdrew from the metformin group owing to severe diarrhoea and abdominal cramps. NOTE: caution should be exercised when combining metformin with nucleoside analogue treatment owing to increased risk of lactic acidosis attached to NRTIs and to metformin, especially in those with kidney problems.

Henry (1998c) analysed strategies for reducing lipid levels. Of 133 people on protease inhibitors, 44 were treated for high lipids with either: diet and exercise, or gemfibrozil, or atorvastatin, or atorvastatin plus gemfibrozil. Exercise/diet had negligible effect on cholesterol and triglycerides. 19 people on gemfibrozil had poor responses and had atorvastatin added. Atorvastatin plus gemfibrozil appeared to be most effective with mean cholesterol falling 30% and triglycerides falling 60% in 6 months. Despite concerns that side-effects would be more likely with these drug combinations, there were no cases of myopathy or raised liver enzymes.

Research into hormone treatment

For a discussion of human growth hormone (HGH) see Human growth hormone in Drugs used by people with HIV: Anti-wasting treatments.

Hengge (2002) conducted a randomised, double-blind study of oxymetholone (50mg twice or three times daily) versus placebo in 92 HIV-infected people with unintended weight loss while on treatment. 12 people did not complete 12 weeks of treatment. At 12 weeks, the mean weight gain was 3.7kg in the twice daily group, 3.1kg in the thrice daily group and 0.97 in the placebo group (p < 0.001). Body cell mass, lean body mass and body mass index increased in the treated groups only. 7.4% of the twice daily group and 21.4% of the thrice daily group developed grade 3 or 4 toxicity.

For other research into the use of anabolic steriods in the treatment of lipoatrophy and HIV-associated wasting see Wasting syndrome - overview and Wasting syndrome - key research in Symptoms and illnesses: A to Z of illnesses.

Oral reported that injections with a synthetic form of the hormone leptin (naturally produced by fat cells) in nine women with diabetes-related lipodystrophy lowered triglycerides levels by 60% after four months of treatment. Discussion of this paper is published in the 20 June 2002 edition of the New England Journal of Medicine. Bodary has reported that leptin affects blood clotting in mice. This association is thought to explain the increased risk of heart attack and stroke in obese people with high levels of leptin.

Research on vitamins, diet and exercise

Gerber (2003) treated 14 HIV-infected men on antiretroviral treatment with high cholesterol or triglycerides with up to 2000mg niacin (Niaspan) daily for 14 weeks. Median cholesterol fell from 245 to 220mg/dl (p<0.005), and triglycerides from 489 to 406mg/dl. However insulin sensitivity worsened.

Barrios (2002) monitored triglycerides and cholesterol in 230 HIV-infected patients on HAART who were asked to follow a low-fat diet for 6 months. All had high lipids prior to commencing the diet. Dietary compliance at 3 and 6 months was 36% and 45%, respectively. Among compliant people, cholesterol fell by 10% while trigs fell by 23% at 6 months. Among non-compliant people, cholesterol fell by only 2% but trigs fell by 9%. Compliant patients experienced an average weight loss of 2kg at 6 months.

Peabody (2002) conducted a randomised, double-blind study of omega-3 fatty acid supplements in 34 HIV-infected people with triglycerides above 3mmol/L. Dietary counselling was given at baseline and omega-3 fatty acid supplements (300mg/day) commenced at week 4. Dietary information and lipid data were collected at baseline, and weeks 4 and 12. 28 people (21 male) completed the protocol. Four weeks after dietary counselling, trglycerides fell by 7%, HDL cholesterol increased by 6%, LDL decreased by 6%, and total cholesterol decreased by 4%. Weeks 4-12 produced no significant difference in lipid parameters between the supplement and placebo groups, but there was a trend towards an effect on triglycerides (10% vs 3% decrease).

Fessel (2002) reported an observational study of niacin (3000mg/day) to treat 16 HIV-infected people with increased intra-abdominal fat and lipid abnormalities. After treatment with niacin for a median of 364 days intra-abdominal fat fell by 17% (=0.01). 13 people experienced a loss of abdominal fat (mean 27%). Decreasing intra-abdominal fat was associated with increased HDL cholesterol.

Engelson (2002b) reported that 15 HIV-infected obese women with insulin resistance underwent a 12 week program of resistance and aerobic exercise and calory reduction. The average weight loss was 6kg. Muscle mass was maintained while subcutaneous and visceral fat tissue was lost, but there was no effect on insulin resistance. Mynarcik (2000) has shown that insulin resistance is related to loss of subcutaneous fat in HIV-positive people, unlike insulin resistance seen in HIV-negative people.

Loignon (2001) conducted an open-label study of L-carnitine in 16 HIV-infected adults who had been taking antiretroviral therapy for an average of 7.2 years. L-carnitine (3g daily) was taken for an average of 8.9 months (range 2.3-19 months). At baseline, mean triglyceride levels were 5.67mmol/L (range 2.28-8.74, normal range 0.5-2.0). Triglycerides fell by 39% after one month of treatment. After two months of treatment, 15% of people had triglycerides below 2.3mmol/L and 54% were below 3mmol/L. There was no significant change in mean cholesterol measures in this study.

Mauss (2001) treated 12 people with HIV-associated lipodystrophy with L-carnitine (1000mg twice daily) for three months. There was no impact on body fat or triglycerides but total cholesterol levels fell significantly.

Keithley (2002) conducted a randomised, double-blind, placebo-controlled study of Cholestin (1.2g twice daily) in 14 HIV-infected people for 8 weeks. One person in each group failed to complete the study. Mean fasting total cholesterol fell by -30.8 in the treatment group and increased by 7.7 in the untreated group (p=0.01). Mean LDL cholesterol also fell significantly in the treatment group (-32.2 vs +26.3, p=0.01). HDL cholesterol and triglyceride levels did not change significantly in either group. Cholestin contains 'naturally occurring statins' including lovastatin and its status as a dietary supplement has been disputed.

Roubenoff (2002) reported the case of a 44-year-old white male who had been taking d4T/3TC/nevirapine for 2.5 years with suppressed viral load and stable CD4 levels. The man had gained 12 kg in weight, and experienced a wasting of the arms and legs, the accumulation of fat around the girth and on the breasts, and had developed large bulging veins in the limbs. Doctors prescribed an exercise programme lasting 4 months involving stretching, 20 minutes of vigorous cardiovascular exercise and 40 minutes weight training for the major muscle groups 3 times a week. At the same time, the man was asked to consume at least 25 grams of dietary fibre daily and follow a diet in which 15% of his total caloric intake was derived from protein and 30% from fat, with the remainder derived from carbohydrates. During the study period, the proportion of calories that the man derived from saturated fat fell from 18% to 11% while his average daily calorie consumption increased by approximately 400 calories per day, to around 2800 calories. At the end of the 4-month programme, there had been a noticeable improvement in the man's body shape, with the loss of over 8kg of fat, a 10% reduction in waist to hip ratio and the disappearance of 62 cm2 of visceral fat from the abdomen (as measured by CT scan). Levels of LDL cholesterol had fallen by -30%, fasting insulin by -3.5% and insulin resistance by over -15%. In addition, visceral fat which had accumulated around his internal organs had fallen by 54%. However, there had been no improvement in the lipoatrophy in his limbs, although his body strength increased by over 50% during the study.

Jones (2001) reported on the impact of exercise training on hyperlipidemia and body composition in 6 men with lipodystrophy over 10 weeks, compared to a matched control group. Exercise consisted of 20 minutes of cardiovascular work (cycle ergometry) followed by 1 hour of resistance exercise three times a week. Total cholesterol and triglycerides were reduced significantly after ten weeks in the exercise group (`19.8% and 25% respectively, p<0.005) but remained unchanged in the control group. HDL and LDL cholesterol levels did not change. Total body fat decelined by 3% whilst body weight increased by 3.6kgs.

Yarasheski (2001) investigated the impact of a 16 week course of weight-lifting exercise on body mass and lipids of 17 HIV-infected men on HAART. At baseline, 11 had high triglycerides. While lean mass (muscle) increased, total adipose (fat) tissue and distribution were not changed. Fasting triglycerides normalised in 9/11 with elevated levels at baseline. Cholesterol levels were unaffected.

Brinkman (1999) reported that vitamin supplements such as riboflavin (vitamin B2) and L-carnitine may be useful in treating people with high lactate levels in the blood. Lactic acidosis, which may results from damage to the mitochondria or energy producers in cells, has been associated with malnutrition and is a rare side-effect of NRTIs.

Tenzif (1999) compared 52 HIV-infected people who undertook cardiovascular and resistance training (1-2 hours a day, 3 days a week) and 38 HIV-infected controls. Groups were matched for food intake, HIV treatment and risk factors for HIV (gender and age data not provided). At week 10, the exercise group had significant increases in body cell mass and lean body mass, as well as quality of life and overall health.

Magro (1999) reported that aerobic exercise (3 x 40 min sessions per week) and a low fat diet were effective in reducing triglyceride levels to normal levels in 20% of individuals with PI-associated hyperlipidemia in a group of 87 patients (57 men, 29 women).

References

See References - body fat and metabolic changes.