“Med diet is a life-saver,” reports the Daily Express today, which goes on to claim that “eating a Mediterranean diet can halve the risk of the most dangerous form of skin cancer”. It says the findings from an Italian skin disease hospital might explain why skin cancer is lower among Mediterranean people than those in northern Europe, the US and Australia.
The study behind this news report looked at patients with malignant melanoma at a hospital in Rome. There were limitations within the design of this study, most notably that it relied on participants accurately recalling their lifetime sun exposure and diet over the past year.
The researchers themselves are cautious to draw definite conclusions from their results, only going as far as stating that factors present in the Mediterranean diet ‘might’ protect from cutaneous melanoma, and that the results warrant further research. Until then, excessive sun exposure remains the most important risk factor for skin cancers, although adopting elements of a Mediterranean-style diet (rich in vegetables, fruit, fish, nut and seed oils) might promote good overall health.
Where did the story come from?
Dr C Fortes and colleagues from the Istituto Dermopatico dell’Immacolata in Rome carried out this study. The research was funded by the Istituto Dermopatico dell’Immacolata, Istituto di Ricovero a Caraterre Scientifico, and by the Italian Ministry of Health. The study was published in the peer-reviewed medical journal, The International Journal of Epidemiology.
What kind of scientific study was this?
This was case-control study based in the inpatient wards of a dermatological hospital in Rome (IDI San-Carlo). All Caucasian patients aged over 18 who were admitted to the hospital with malignant cutaneous melanoma (skin cancer) between May 2001 and May 2003 were included as cases.
People who did not have skin cancer were used as controls and selected from the same hospital during the same time period and were matched to the cases according to age and gender.
Trained interviewers then asked the cases and controls about their, medical history, smoking status, family history of skin cancer and skin type, hair and eye colour, exposure to sun, sunburn history and use of sun beds/lamps and other sociodemographic information.
The number of "pigmented lesions" on the body (excluding scalp and pubic region) were recorded. Acquired melanocytic lesions were defined as those of a certain size that were not identifiable as freckles, sun spots, other benign lesions or pigmented basal cell carcinomas; i.e. they were acquired "melanocytic nevi". These were counted and recorded as none, few (1-24), moderate (25 to 59) or many (60 or more).
Other lesions were also counted, including “atypical naevus” (none or 1), sunspots (none, limited to one body area, two body areas, or more than two body areas).
- Other skin characteristics were recorded:
Skin type was classified using the Fitzpatrick system (which assesses tanning and burning tendency). Skin, eye and hair colour were also recorded.
- Sun exposure at ages pre-12 years, 12-18 years and over 18 years was assessed (based on average hours of sun exposure per day), as were lifetime sunburn episodes (none, 1, 2-6, 6 or more) and use of sun protection.
A food frequency questionnaire was then used to assess the participants’ diet in the year leading up to the interview. It was this that was used to determine whether a person’s diet was a Mediterranean one.
The researchers then used a statistical modelling technique (called logistic regression) to examine the association between diet and skin cancer, while taking into account other factors that may influence this relationship (sunburns, skin type, sun exposure, sex, age, education, number of pigmented lesions).
What were the results of the study?
304 cases and 305 controls agreed to participate, and were interviewed and clinically examined. The researchers collected various pieces of information from the participants in an attempt to rule out non-dietary factors that might cause melanoma. This process found that:
- More of the cases with melonoma had fair hair and fair skin compared with controls.
- Compared to people with dark brown or black hair, those with blonde or red hair were more likely to have melanomas.
- Other factors associated with increased risk of melanoma were skin types I and II compared to III and IV, presence of freckles, light coloured eyes, having many sunspots, sunburn episodes in childhood, spending time outdoors, and use of sun beds or lamps.
When the researchers took some of these factors into account, namely sex, age, education, hair colour, skin types, number of pigmented lesions, presence of freckles and sunburn in children, they found that:
- Consumption of vegetables (five or more times a week) and fruits (more than once a day) reduced likelihood of skin cancer.
- Other protective elements of diet included a high consumption of cruciferous vegetables (brassicas), leafy green vegetables, carrots, citrus fruit, fresh herbs, nuts, salad and exclusive use of olive oil for dressing (although some of these results were not very precise).
- High consumption of fish rich in n-3 fatty acids was also protective, as was consumption of shellfish.
- Tea drinking was also protective.
- There was no association with alcohol, high meat intake, liver, offals, cheese, butter, eggs and milk.
- There was little difference between men and women in these results.
In another analysis, the researchers also took into account BMI and use of specific medications, and found that these did not affect the risk estimates.
What interpretations did the researchers draw from these results?
The researchers conclude that after careful control for individual characteristics and sun exposure, the Mediterranean dietary profile (use of fresh herbs, citrus, cruciferous and dark green vegetables, and high consumption of fish rich in n-3 fatty acids and shellfish) might give protection against cutaneous melanoma.
What does the NHS Knowledge Service make of this study?
This case control study has found an association between diet and skin cancer, but there are conflicting results from other studies about this relationship. Given its design, the study has some limitations that should be kept in mind when interpreting these results:
The main limitation of this study is the possibility that bias (systematic error) associated with recollection has been introduced (namely ‘recall bias’). The participants were asked to describe their exposure to sunlight, etc. from years, possibly decades before, which could lead to confusion over details or instances.
Additonally, the participants were asked to report incidences of sunburn when they were under 12 years old, aged 12-18, and during their adult years. It is unlikely that people correctly recall their exact age during all cases of sunburn.
Recall bias could also have affected the responses to the questionnaire, as dietary questions were based on consumption in the year before admission. It is unclear how accurate recall of diet over 12 months would actually be, and there may have been differential recall between cases and controls.
There may also have been a systematic difference in the way that cases reported their sun exposure compared to the controls, given that they have had skin cancer in their adulthood.
This study did not assess portion sizes, therefore it is not possible to make extrapolations about the amount of each foodstuff that showed protective associations.
The findings of this study support a link between foods that are rich in polyphenols and n-3 fatty acids and a reduction in risk of malignant melanoma. However, even the researchers are cautious in their conclusion, saying that “some dietary factors present in the Mediterranean diet ‘might’ protect from cutaneous melanoma”.
This caution is likely to relate to the fact that, as a case control study, its results cannot prove causal links between exposures and outcomes. It is therefore important to investigate these claims in prospective studies.
Until then, excessive sun exposure remains the most important risk factor for these types of skin cancers, and those with fairer skin types should be particularly careful. However, the balanced intake of vegetables, fruit, fish, nut and seed oils in a Mediterranean-style diet can help promote general health.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
Daily Express, 13 October 2008
The Daily Telegraph, 13 October 2008
Links to the science
Int J Epidemiol 2008; 37: 1018-1029