Adherence to Mediterranean diet and risk of esophageal, gastric, pancreas, lung and breast cancer: a prospective cohort study

  • Topic: Combination of cancers
  • Institution: Maastricht University
  • Country: Netherlands
  • Status: Ongoing

Scientific abstract

(View plain language abstract)

Background

The traditional Mediterranean diet (MD) is characterised by a high intake of fruits, vegetables, legumes, whole grains, fish, and monounsaturated fat, moderate alcohol intake and low intake of refined grains, red meat, and sweets. The MD has been associated consistently with decreased overall, cardiovascular and cancer mortality. Adherence to a MD pattern is also related to lower risk of cardiovascular disease incidence, as was confirmed in randomised control trials (RCTs). Besides this cardioprotective effect, there is increasing interest in the potential cancer preventive effect, i.e. the association between MD-adherence and cancer incidence. This is especially the case for tumor types with poor prognosis, like esophageal, gastric, lung and pancreas cancers. However, very few prospective studies have been published on this matter, and there is no consistent evidence yet, notably for subtypes. For breast cancer, it may be that particularly estrogen receptor negative (ER-) subtypes are related to MD-adherence. The large Netherlands Cohort Study (NLCS) with long follow-up provides an excellent opportunity to study the relationship between adherence to MD and these cancers.

Hypothesis and objectives

  1. Adherence to the MD pattern is associated with a lower risk of total cancer in men and women, and with a lower risk of esophageal, gastric, pancreas, lung, and breast cancer.
  2. The MD is differentially associated with risk of subtypes of the abovementioned cancers:
    1. Esophageal squamous cell carcinoma and gastric noncardia adenocarcinoma are stronger inversely related with MD-adherence than esophageal and gastric cardia adenocarcinoma.
    2. Associations with lung cancer vary according to histologic subtype.
    3. ER- breast cancer subtypes are inversely associated with MD, and ER+ cancers not.
    4. The association with MD-adherence and total cancer risk is more inverse in women than in men, and is more inverse in smoking-related types of cancer.
  3. The inverse relationships with cancer are stronger when alcohol intake is excluded from the MD-adherence score.
  4. To compare models with (relative) MD-scores with models using scores for adherence to dietary WCRF/AICR recommendations, based on absolute cutoffs.

Settings and methods

The NLCS uses a case-cohort design and includes 120,852 subjects aged 55-69 years at baseline in 1986, when they completed a questionnaire on diet and other lifestyle and medical risk factors for cancer. Record linkages with cancer registries are used for cancer incidence follow-up. After 20.3 years of follow-up, 32,249 total, 463 esophageal, 1,036 gastric, 766 pancreatic, 3,902 lung and 3,357 incident breast cancers were detected. After data-entry of the questionnaires of newly detected cases, two scores for (relative) MD-adherence (aMED, mMED) will be constructed and compared. Multivariable adjusted hazard ratios for the various endpoints will be estimated using Cox proportional hazards survival analysis adapted to the case-cohort design, to test the hypotheses.

Impact

This cohort study with high statistical power will provide evidence on MD-adherence and risk of esophageal, gastric, pancreas, lung, and (ER-) breast cancer, as well as total cancer. If the MD is indeed related to reduced risk, the impact on risk of these cancers by promoting and adhering to MD may be substantial.

Plain language abstract

Background

The traditional Mediterranean diet (MD) is characterised by a high intake of fruits, vegetables and legumes, whole grains, fish, and monounsaturated fat, moderate alcohol intake and low intake of refined grains, red meat, and sweets. The MD can protect against heart disease, as has been shown in intervention studies. Besides this protective effect, there are several recent scientific indications that adhering to MD pattern might also protect against cancer. This concerns especially cancer types with a poor prognosis, such as esophageal, stomach, pancreatic, and lung cancer. For these cancer types primary prevention is of utmost importance, given the very poor prognosis. However, there are still very few prospective studies available on this, and there is no consistent pattern of results available yet, let alone on specific subtypes of these cancers. It is also interesting that a particular subtype of breast cancer, the so-called estrogen receptor negative (ER-) type (which has a worse prognosis than ER+ breast cancer) may be related to the MD pattern as well.

Aims and objectives

We aim to investigate whether adhering to a MD pattern will lower the risk of total cancer in men and women, and whether it lowers specifically the risk of the following cancer types: esophageal, stomach, pancreas, lung, and breast cancer. We will also investigate whether the influence of the MD pattern is different for various subtypes of esophageal, stomach, pancreas, lung, and breast cancer, and whether it is associated more strongly with smoking-related cancers. We will also test whether specific dietary components of the MD pattern have a dominant influence of the observed associations with cancer. Finally, we will compare the MD pattern with existing dietary World Cancer Research Fund International recommendations in terms of potential cancer prevention.

How it will be done

The prospective Netherlands Cohort Study (NLCS) began in September 1986 among 120,852 men and women aged 55-69 years who completed a self-administered questionnaire on dietary habits and other risk factors for cancer. Follow-up for cancer occurrence in this cohort is being accomplished by record linkage to the Netherlands cancer registry and with PALGA, a nationwide Dutch pathology database. After 20.3 years of follow-up, we observed 32,249 new cases of total cancer (19,779 men, 12,470 women); 3,357 breast cancer; 3,902 lung; 463 esophageal; 1,036 stomach; and 766 pancreatic cancer. The full original (1986) questionnaires of these subjects will be entered into our computer database, and from that the average intake of nutrients and foods will be calculated. A score will be constructed that indicates the adherence to the MD pattern. In the statistical analyses, we will assess the association of MD adherence with the 20.3 year risk of the mentioned cancer types, and test the other objectives.

Potential impact

This cohort study with high statistical power will provide evidence on MD-adherence and risk of esophageal, stomach, pancreas, lung, and (ER-) breast cancer, as well as total cancer. If the MD is indeed related to reduced risk, the impact on risk of these cancers by promoting and adhering to the MD may be substantial.