Encyclopedia
In the Greek segment of the EPIC cohort study (EPIC-Greece), the changes in dietary habits of 23,505 participants regarding several food items/groups (vegetables, legumes, fruits, nuts, dairy, cereal, meat, fish/seafood, olive oil) were recorded repeatedly over time and compared to the baseline assessment (1994–1997), using a short, qualitative, follow-up questionnaire.
- food questionnaire
- nutritional behavior
- follow-up dietary data
- EPIC
- ordinal regression
1. Introduction
Dietary habits vary over time [1][2][3][4][1,2,3,4] and are correlated with a wide range of individual characteristics, such as biological, demographic, psychological, situational and socio-cultural on an interpersonal level [5]. The association between dietary behavior and health has been studied intensively over the last decades, the literature in the area of nutritional epidemiology is very extensive and the evidence of association can vary from weak to strong depending on the health outcome studied [6][7][8][6,7,8]. Therefore, it is of high public health importance to recognize trends in the dietary habits of a population. Due to the heterogeneity of dietary habits between different population subgroups, it is also of high priority to study the nutritional trends in association with socioeconomic, demographic and health factors as part of well-designed policy plans and strategies to address nutritional trends towards more unhealthy patterns of consumption, and specify the subgroups that should be the main focus of these policies.
The National Diet and Nutrition surveys/Food Observation surveys [1][9][10][1,9,10] and long-term cohort studies [11][12][13][11,12,13] aim to assess nutritional trends either in representative samples of the national population or among cohort participants. Among the cohort studies with repeated assessment of dietary habits, most use a unique quantitative questionnaire (i.e., the same as baseline) or a shorter version of the baseline questionnaire for follow-up information. It is less common to use qualitative follow-up questions to evaluate changes in the dietary habits of the study population [14]. Using a qualitative dietary follow-up questionnaire has the advantage of being simpler to administer on a large scale. With the use of appropriate statistical models, the information provided by qualitative follow-up questionnaires can be used to study the association between a large number of sociodemographic, somatometric and health status factors, as well as for the assessment of nutritional trends.
This less common qualitative questionnaire approach was explored in order to meaningfully capture the long-term changes in the dietary preferences of the EPIC-Greece cohort participants, who were asked about dietary changes compared to baseline during follow-up.
2. The EPIC-Greece Study
European Prospective Investigation into Cancer and Nutrition (EPIC) is a multicenter prospective cohort study in 10 European countries, investigating the association of several risk factors, with a special focus on diet with cancer and other health outcomes [15]. Many publications of this major study have contributed to our knowledge of nutrition and health [8][16][17][8,16,17]. The Greek segment of EPIC started in 1994, after the successful implementation of a pilot study during the period 1991–1993. The recruitment of volunteers was completed in 1999. All volunteers gave their informed consent for inclusion before they participated in the study. The study was conducted in accordance with the Declaration of Helsinki. The EPIC-Greece cohort consists of 28,572 adults, aged between 25 up to 86 years old, recruited from all over Greece on a voluntary basis, and representing a broad range of sociodemographic factors. To enhance the response rate, a campaign informing potential participants was conducted via television, radio and announcements at the local community centers. At enrollment, detailed information on the participants’ demographic and socioeconomic characteristics, physical activity anthropometric measurements, medical and reproductive history was recorded via a face-to-face interview with the volunteer at the local health centers. Dietary intake at baseline was interviewer-administered via a detailed semi-quantitative food frequency questionnaire (FFQ) [18].
3. Follow-Up
After the initial enrollment, between 1994 and 1997, 26,579 participants were followed-up every 3–4 years in order to update the information on lifestyle factors, including diet, and on the health status. There were no exclusion criteria. The follow-up of participants between 1997 and 2011 was conducted mainly by telephone interviewer-administered questionnaires (92.34% of the interviews) and to a small extend by mail (7.15%) or by personal interviews (0.51%). When the participant was not found, information on vital status was obtained by the next of kin. Information on the occurrence of cancer and cardiovascular diseases was confirmed by hospital records.
4. The Follow-Up Dietary Questionnaire (FU-DQ)
In the first follow-up period (1997–2002), the questionnaire did not include questions on food consumption. From the second follow-up onwards (after 2002), a dietary follow-up questionnaire (FU-DQ) was designed and administered to 23,505 participants. The FU-DQ included questions on the consumption of the following food items/groups: potatoes, vegetables, legumes, fruits, dairy, low-fat dairy, nuts, cereal, total red meat, lamb/goat, poultry, fish/seafood, eggs, olive oil, seeds oil, margarine, butter, sugar and confectionery products, and alcoholic and non-alcoholic beverages. For all food items/groups excluding alcoholic beverages, individuals were asked to qualitatively report the change in consumption when compared to their consumption at baseline. The questions regarding the change in food consumption included seven pre-specified, categories: “No consumption (None)”, “Much less”, “Less”, “The same”, “More”, “Much”, “More” and “Do not remember”.
5. Study Sample
The EPIC follow-up is a dynamic, continuous process that started in 1997 while the follow-up dietary questionnaire (FU-DQ) started in 2002. In the present analysis we include the dietary follow-up until 2011. Within that period of 14 years, each participant was interviewed approximately every 3–4 years. The first follow-up period was in 1997–2002, the second in 2002–2007 (first FU-DQ period) and the third was in 2007–2011 (second FU-DQ period).
