Food and diet

Do tomatoes really compare to statins?

Cooked tomatoes may “have the same benefits as statins”, the Daily Mail has reported, and The Daily Express said that “pizza can be healthy” because cooked tomatoes may be the key to a healthier lifestyle.

These reports are based on review of previous research examining how cholesterol and blood pressure may be affected by lycopene, the chemical that gives tomatoes and other fruits their red colouring. Analysing the combined results of 14 trials found lycopene had no effect on cholesterol overall, although a subanalysis of doses at or above 25mg a day found it reduced cholesterol by a small amount compared to no lycopene.

Lycopene also reduced systolic blood pressure, but not diastolic blood pressure. Both readings are equally relevant in the assessment of a person’s blood pressure.

Importantly, not all of the included trials had robust designs, with limitations that meant the results were not necessarily due to lycopene having some effect. Also it is not clear whether the observed differences in cholesterol or systolic blood pressure would have had any effect on health outcomes, as cardiovascular disease development and related events were not assessed in these short-term trials.

Despite the review concluding that the effect ‘is comparable to the effect of low doses of statins in patients with slightly elevated cholesterol levels’, lycopene use has not been directly compared with statins, so such conclusions cannot be made. Tomatoes and other lycopene-containing fruits can still contribute to our recommended five daily portions of fruit and vegetables. However, based on this study it should not be assumed they possess the effects reported in newspapers based on this study.

Where did the story come from?

The study was carried out by researchers from The University of Adelaide in Australia, supported by the Primary Health Care Research Evaluation Development (PHCRED) Program and funded by the Australian Government of Health and Ageing.

The study was published in the journal, Maturitas.

In suggesting that tomatoes and lycopene offer the same effect as statins the media has not considered that the research has not directly compared the effects of lycopene with those of statin treatment. Therefore, it cannot be said that lycopene ‘is as good as statins’, although it does appear that media suggestions of this type are based on statements in the research paper itself, which says the effects are similar.

Also the Daily Express’ headline suggesting that pizzas can be healthy does not take account of the high saturated fat, salt and sugar content of many pizzas, which would not make a healthy option even with the inclusion of tomato paste.

What kind of research was this?

This was a systematic review and meta-analysis that aimed to identify controlled studies that had investigated the effect of lycopene on blood lipids (soluble fats), and blood pressure. Lycopene is the chemical responsible for the red colouration of tomatoes, watermelon and other fruit. It is thought to have antioxidant effects and prevent the oxidation of low-density-lipoprotein (bad cholesterol) and prevent atherosclerosis – the fatty build-up in arteries that causes cardiovascular disease.

A systematic review is the best way of examining the global literature for trials that have examined the effects of a particular intervention. However, the reliability of its findings will depend on its methods and the characteristics of the studies that it includes.

This review chose to include studies that were not randomised controlled trials, the most robust study design for addressing this sort of question. By including other, less-robustly designed studies it means that the results may have been affected by differences between the groups, which have nothing to do with their lycopene intake.

Any meta-analysis also carries some inherent limitations if the individual studies it is combining vary in their methods and design, such as through their inclusion criteria, intervention methods, follow-up period and assessment of outcomes.

What did the research involve?

The researchers searched the PubMed and Cochrane databases for studies published between 1955 and 2010 that had examined the effects of lycopene on blood lipids or blood pressure. To be eligible trials had to be in English, be a diet- or placebo-controlled trial, used a standard natural lycopene dose, have an intervention period for at least two weeks and report average blood lipid levels (total cholesterol, HDL, LDL, triglycerides) or blood pressure levels before and after the intervention.

They looked at the effect of lycopene on cholesterol and blood pressure levels, applying methods that took into account ‘heterogeneity’ (differences) between the individual trials’ results and the nature of individual trials’ results. Subgroup analyses were carried out to see whether the dose of lycopene (less than or greater to 25mg daily) affected cholesterol, and whether the effect on blood pressure was influenced by baseline blood pressure (whether or not the person had hypertension or not at the start of the study).

What were the basic results?

Fourteen studies met inclusion criteria, 12 investigating the effect of lycopene on total cholesterol and four also looking at blood pressure. Seven studies had a control group that either had a placebo or a lycopene-free-diet, while the remaining studies investigated lycopene-rich and lycopene-free periods in the same person. Trials used lycopene-containing tomato products, watermelon juice or tomato extract capsules (in one study these capsules also contained other carotenoid rich extracts), with lycopene doses ranging from 4-44mg daily.

Treatment periods ranged between two and six weeks, with one trial using a six-month intervention period. Six of the seven cholesterol studies were investigating people with high cholesterol, and two of the four blood pressure trials were investigating people with hypertension.

When the 12 studies on cholesterol were pooled (694 people) lycopene treatment had no effect on cholesterol compared to control treatment (mean difference from control -0.87 mg/dl, 95% confidence interval [CI] -4.12 to +2.38). However, when the researchers separately analysed those trials according to lycopene dose, they found that only doses at or above 25mg a day affected cholesterol and blood pressure.

Compared to control interventions these doses significantly reduced both total cholesterol (401 people analysed: mean difference from control −7.55 mg/dl, 95% CI -13.70 to -1.40) and low-density-lipoprotein (253 people analysed: mean difference −10.35, -15.99 to -4.71).

Meta-analysis of all four studies of blood pressure found that, overall, lycopene reduced systolic blood pressure - the upper blood pressure reading reflecting arterial pressure when the heart is contracting (209 people analysed: mean change compared to control −5.60mmHg, 95% CI -10.86 to -0.33).

There was no effect on diastolic blood pressure (the lower blood pressure reading reflecting arterial pressure when the heart is filling with blood).

How did the researchers interpret the results?

The researchers conclude that their meta-analysis suggests that lycopene taken in doses of 25mg daily or greater is effective in reducing LDL and total cholesterol. They say this effect ‘is comparable to the effect of low doses of statins in patients with slightly elevated cholesterol levels’. They call for more research to confirm suggested beneficial effects on total serum cholesterol and systolic blood pressure.

Conclusion

This review of the effects of lycopene on cholesterol and blood pressure has some limitations and this means that it cannot conclusively tell us whether lycopene has any effect on lowering cholesterol or blood pressure. It definitely cannot tell us whether lycopene has any effect on the risk of heart disease. The points to note include:

  • The study was a systematic review, which is the best way to identify and summarise all relevant literature on a particular question. Unfortunately, it was limited by the fact that it included non-randomised studies. By including these less robustly designed studies it meant that the results may have been affected by differences between the groups, which have nothing to do with their lycopene intake.
  • Any meta-analysis carries some inherent limitations if there is variation in the methods of the individual studies, such as their inclusion criteria, intervention methods, follow-up period and assessment of outcomes. In this case, the studies varied in their methods of allocating participants, the doses of lycopene consumed, the form in which lycopene was provided, and the duration of the treatment, among other things.
  • Trials were mostly of a small sample size, and this may have affected the ability of even randomly assigned studies to achieve balanced groups. Larger studies would be able to provide more-robust findings.
  • There was no effect of lycopene on cholesterol overall. Only subanalysis by dose revealed an effect of higher doses. There was also no effect on diastolic blood pressure, only systolic blood pressure, and both readings are equally relevant in the assessment of a person’s blood pressure.
  • It is not clear whether the observed differences in cholesterol or systolic blood pressure would have had any effect at all on health outcomes, as cardiovascular disease development and related events had not been assessed.
  • Despite the paper concluding that the effect ‘is comparable to the effect of low doses of statins in patients with slightly elevated cholesterol levels’, lycopene use has not been directly compared with statins so such conclusions cannot be made.

Overall, it is not possible to say conclusively from this review that tomatoes or lycopene have any definite effect on cholesterol or blood pressure, and whether this is clinically significant. Large randomised controlled trials will be needed to further investigate this question. In the interim, tomatoes and other lycopene-containing fruits can still contribute to our recommended five daily portions of fruit and vegetables.


NHS Attribution