Abstract
Objective
The consumption of potatoes is increasing worldwide, but few studies have assessed the association between potato consumption and mortality, particularly in Mediterranean countries. We therefore investigated whether potato consumption is associated with higher risk of death in a large cohort of people living in South Italy.
Design
Longitudinal.
Setting
Community-dwelling.
Measurements
2,442 participants coming from MICOL and NUTRIHEP studies aged more than 50 years at baseline were followed-up for 11 years. Dietary intake was assessed by means of a Food Frequency Questionnaire. Potato consumption was categorized in quintiles according to their daily consumption (< 3.95, 3.96-8.55, 8.56-15.67, 15.68-22.0, and > 22.0 g/day). Mortality was ascertained through validated cases of death. The association between potato consumption and mortality was assessed through Cox's regression models, adjusted for potential confounders, and reporting the data as hazard ratios (HRs) with 95% confidence intervals (CIs).
Results
The 2,442 eligible participants were prevalently males (54.6%) and aged a mean of 64.3±9.3 years. During the 11-year follow-up, 396 (=16.2%) participants died. After adjusting for 12 potential baseline confounders, and taking those with the lowest consumption of potatoes as the reference group, participants with the highest consumption of potatoes did not have an increased overall mortality risk (HR=0.75; 95%CI: 0.53-1.07). Modelling the potato consumption as continuous (i.e. as increase in 10 g/day) did not substantially change our findings (fully-adjusted HR=0.93; 95%CI: 0.84-1.02).
Conclusion
Overall potato consumption was not associated with higher risk of death in older people living in a Mediterranean area. Future studies are warranted to elucidate the role of potato consumption on all-cause and cause-specific mortality.
Key words: Potato, mortality, Mediterranean diet, cohort study, FFQ
Introduction
Potatoes, a typical Western food, are rich in starch and have a high–glycemic index. However, compared with other common carbohydrate sources, potatoes have a low energy density due to their high-water content (1) Moreover, potatoes provide other important micronutrients that literature has associated with a decreased risk of morbidity and mortality (2). For these reasons, the literature regarding potato consumption and medical conditions and mortality is still equivocal.
In three large North American cohorts, greater consumption of potatoes was associated with a significant higher risk of diabetes, also after taking in account several potential confounders (3). The association between potato consumption and incident diabetes was confirmed in the Nurses' Health Study, another large North American cohort study (4). However, a study conducted in Sweden among almost 70,000 individuals failed to find any significant associations between higher potato consumption and the risk of cardiovascular disease (CVD) morbidity and/or mortality (5). A recent systematic review, including five large observational studies, did not support any evidence linking higher potato intake and the risk of developing obesity, type 2 diabetes or CVD (6). Finally, to the best of our knowledge, only one study reported the association between potato consumption and overall mortality finding that the consumption of potatoes was not associated with higher mortality risk overall 8 years of follow-up (7).
Even if all these studies advanced our knowledge regarding potato consumption and possible higher risk of medical conditions/mortality, all of them were made among North American or North European individuals, that probably have significant different dietary patterns than other countries, such as Mediterranean ones.
Since recent reports show that potato consumption is increasing worldwide and that the heartiest potato eaters are Europeans (8) and limited literature is available regarding potatoes and overall mortality in Mediterranean countries, this study aimed to investigate whether higher consumption of potatoes is associated with any increased risk of death in individuals participating in two south Italian cohort studies.
Methods
Participants
This research is based on two longitudinal studies, including participants aging > 50 years, with a common baseline evaluation in 2005 conducted at the Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS) “Saverio de Bellis” (Castellana Grotte, Italy).
The first one is called NUTRIHEP (Nutrition Hepatology) with fully details have been published elsewhere.[9, 10] In brief, in collaboration with 12 General Practitioners (GPs) working in Putignano (Puglia, Italy) and, after testing the hypothesis that the sex-age group specific mean age was the same among the general population and subjects of the GP clinics (p= 0.15), a random sample was drawn (for those individuals 18 years old and older) from the GP patients' list. The NUTRIHEP study included 2,301 subjects (systematic population random sample from age range: 18–95 years).
The other study is called MICOL (Multicentrica Italiana Colelitiasi, Italian, Multicentric study for cholelithiasis). MICOL is population-based study started in 1985 and consisted in a systematic random sample of Castellana Grotte population aged 30 or older. For the aims of our research, we used the data of the third assessment made in 2005 as baseline evaluation. The population enrolled in 2005 (third assessment) consisted of 1,708 subjects (among 1,942 initially contacted or re-called), 950 males and 758 females.
Both studies and the proposal of this research were approved by the Institutional Review Board (Ethical Committee) of IRCCS De Bellis and written informed consent was obtained from each participant before entering the study.
Measurements
The survey visit consisted of the administration of a standardized questionnaire, including validated food frequency questionnaires, medical history and examination, and anthropometric measurements. All assessments were performed at baseline and in the same day.
A trained nurse recorded weight and height and body mass index (BMI) was calculated and reported in Kg/m2. Finally, blood pressure was also recorded.
Several medical conditions were recorded in both cohorts, including self-reported information regarding hyperuricemia, acute myocardial infarction, hypertension, dyslipidemia, diabetes, and gastric ulcer. Hyperuricemia was defined as serum uric acid over 7 mg/dl in men and 6 in women or use of allopurinol (11).
Regarding nutritional parameters, two different questionnaires were administered. In the NUTRIHEP, the EPIC (European Prospective Investigation on Cancer) Food Frequency Questionnaire (FFQ) was given (12). In the MICOL study, a validated 96 food items FFQ was given (13).
Exposure: potato consumption
Potato consumption was recorded through the standardized and validated questionnaires mentioned before. For the aims of our study, we divided the population in five categories according to their daily potato consumption, as follows: < 3.95, 3.96-8.55, 8.56-15.67, 15.68-22.0, and > 22.0 g/day.
Outcome: mortality
The time of observation from date of enrolment to death, migration or end of study, whichever occurred first, was recorded for each subject. The follow-up of the subjects ended at December 31, 2016. A copy of the official death certificate of all deceased participants was obtained. Mortality rates were obtained by dividing the number of deaths occurring during the study by the accumulated number of person-years.
Statistical analyses
For descriptive purposes Smoking habits was categorized as current vs. past/never. Occupation was recorded as manager/professionals, craft, agricultural or sales workers, elementary occupations, housewife, pensioners, jobless, or missing information. For analytical purposes, Occupation was categorized as the most frequent job (i.e. elementary occupations) vs. others and Education as primary vs. higher levels.
For the aims of our study, we reported energy daily intake (in Kcal/day), alcohol intake (in g/day) and the adherence to Mediterranean diet. To identify Dietary Patterns a factor analysis was applied to grams of food groups consumed per day, using principal component factor analysis (PCFA) as the extraction method. Before running PCFA, it was necessary to collapse the food items of the FFQ into 15 major food groups. Major foods groups were chosen not only to provide comprehensive representation of the local diet but also by taking into consideration glycemic index and nutritional composition of foods (see Supplementary Material Table 1). Then, the PCFA allows us to describe the variance-covariance structure among food groups in terms of a few underlying unobservable and randomly varying factors (14). The Kaiser-Meyer-Olkin (KMO) index and Bartlett's sphericity test were used to evaluate factorability of the correlation matrix. An orthogonal varimax rotation was then applied to the factor loading matrix (15). We chose three factors to retain based on criteria of ensuring a factor eigenvalue greater than 1 and factor interpretability. To account for parsimony and plausibility of the factors, Bayesian Information (BIC) and Akaike information (AIC) criteria were employed (15). Factor scoring coefficients were then calculated for each study participant by applying the regression method (14) and categorized for adherence (low, medium, and high) according to tertiles of factor scores.
Table 1.
Characteristics of the participants classified according to their frequency of potato consumption
| < 3.95 g/day (n=517) | 3.96-8.55 g/day (n=649) | 8.56-15.67 g/day(n=542) | 15.68-22.0 g/day(n=248) | > 22.0 g/day (n=486) | P value for trend1 | |
|---|---|---|---|---|---|---|
| General characteristics | ||||||
| Age (years) | 64.6 (9.6) | 65.2 (9.3) | 63.6 (9.0) | 63.4 (8.8) | 64.0 (9.5) | 0.02 |
| Females (n, %) | 240 (46.4) | 295 (45.5) | 214 (39.5) | 133 (53.6) | 227 (46.7) | 0.52 |
| Smoking (current) (n, %) | 59 (11.5) | 75 (11.6) | 54 (10.0) | 13 (5.3) | 31 (6.4) | <0.0001 |
| Elementary occupations2 (n, %) | 177 (34.5) | 249 (38.5) | 192 (35.6) | 74 (30.0) | 145 (29.8) | <0.0001 |
| Primary school (n, %) | 283 (55.2) | 383 (59.3) | 289 (53.4) | 132 (53.4) | 264 (54.3) | 0.06 |
| Nutritional parameters | ||||||
| Energy intake (Kcal/day) | 1829 (646) | 1941 (613) | 2097 (688) | 1987 (602) | 2285 (684) | <0.0001 |
| Alcohol drinks (g/day) | 12.3 (21.7) | 13.6 (23.6) | 16.9 (27.6) | 10.5 (21.5) | 12.3 (24.8) | 0.18 |
| High adherence to Mediterranean diet (n, %) | 137 (26.5) | 205 (31.6) | 218 (40.29 | 108 (43.5) | 265 (54.5) | <0.0001 |
| Medical conditions | ||||||
| BMI (Kg/m2) | 29.1 (5.1) | 29.4 (5.0) | 29.0 (5.3) | 28.3 (4.7) | 28.8 (4.8) | 0.03 |
| Systolic blood pressure (mmHg) | 131.4 (18.4) | 131.6 (18.4) | 131.6 (17.6) | 133.3 (17.7) | 132.1 (16.3) | 0.17 |
| Diastolic blood pressure (mmHg) | 77.2 (9.9) | 77.2 (9.8) | 78.9 (9.6) | 79.3 (9.7) | 80.1 (9.6) | <0.0001 |
| Presence of any disease3 (n, %) | 154 (29.8) | 176 (27.1) | 122 (22.5) | 50 (20.2) | 100 (20.6) | <0.0001 |
Notes: The data are presented as means (with standard deviations) for continuous variables and number (with percentage)
P values for trends were calculated using the Jonckheere-Terpstra test for continuous variables and the Mantel-Haenszel Chi-square test for categorical variables;
This variable was categorized as manager/professionals, craft, agricultural or sales workers, elementary occupations, housewife, pensioners, jobless, and missing information. The most common job was reported as descriptive information;
The medical conditions commonly recorded were: hyperuricemia; acute myocardial infarction; hypertension; dyslipidemia; diabetes; gastric ulcer; Abbreviations: BMI: body mass index.
Normal distributions of continuous variables were tested using the Kolmogorov-Smirnov test. Data are shown as means±standard deviations (SDs) for quantitative measures, and frequency and percentages for all discrete variables. P values for trends were calculated using the Jonckheere-Terpstra test for continuous variables and the Mantel-Haenszel Chi-square test for categorical variables.
Incidence rates are reported as number of deaths for 1,000 persons-years. The proportional hazards assumption was checked by plotting the Schoenfeld residuals versus time without any violation and then Cox's regression analyses were performed (16). The basic model was adjusted for age (as continuous) and sex. In addition to age and sex, the fully adjusted model adjusted for: project (MICOL vs. NUTRIHEP); Occupation; Educational level, Systolic and Diastolic Blood Pressure at baseline (both as continuous); adherence to Mediterranean diet (as continuous); BMI (as continuous); presence of any disease (yes vs. no); alcohol intake (g/day, as continuous); total energy intake (Kcal, as continuous). These covariates used for adjustment were factors significantly different across potato consumption categories or significantly associated with death according to univariate analysis (considering a p-value<0.10 in both cases). In secondary analyses, we reported the association between potato consumption (modelled as continuous, i.e. as increase in 10 g/ day of potato consumption) and mortality. Multi-collinearity among covariates was assessed using the variance inflation factor, with a score of 2 leading to the exclusion of a variable, but no parameter was excluded for this reason. In all the analyses, Cox's regression analysis data are reported as hazard ratios (HRs) with 95% confidence intervals (CIs).
To test the robustness of our analyses, sensitivity analyses were conducted evaluating the interaction between potato consumption and selected factors (e.g. sex, presence/absence of diseases at baseline, BMI above or below 25 Kg/m2) in predicting mortality, but no moderator emerged as significant.
All the analyses were performed using SPSS 17.0 for Windows (SPSS Inc., Chicago, Illinois). All statistical tests were two-tailed and statistical significance was assumed for a p-value <0.05.
Results
Sample selection
Among the 2,301 individuals included in the NUTRIHEP and the 1,708 included in the MICOL studies (n=4,009 participants), 2,687 were older than 50 years. From this sample, we excluded 84 participants without information regarding potato consumption, 97 having implausible calorie intake (< 800 or > 4,200 Kcal/day), and 64 with no information regarding vital status during follow-up, obtaining 2,442 eligible participants.
Descriptive baseline characteristics
The 2,442 eligible participants were prevalently males (1,333, =54.6%) and aged a mean of 64.3 years (±9.3; range: 50-93). The average daily potato consumption was 16±18 g (range: 0–275 g/day).
Table 1 reports the participants' characteristics by baseline potato intake. Those consuming more potatoes (i.e. Q4, > 22 g/day) were significantly older (p for trend=0.02) and less frequently smokers (p for trend<0.0001) than other participants. The participants with higher potato consumption had a higher adherence to Mediterranean diet compared to other people included in this analysis (p for trend<0.0001) and a lower BMI (p for trend=0.03).
Regarding medical conditions, people with a higher potato consumption had a significant higher diastolic blood pressure ((p for trend<0.0001), but they did not differ regarding systolic values (p for trend=0.17) (Table 1). Finally, those consuming more potatoes had a significant lower presence of any disease (Q4=20.6 vs. Q1=29.8%, p for trend <0.0001).
Potato consumption and mortality
During a median follow-up period of 11 years, 396 people (=16.2% of the baseline population) died.
Table 2 illustrates the association between potato consumption and mortality. The incidence of mortality did not differ across categories of potato consumption (p=0.84). Using a Cox's regression analysis, adjusted for 12 potential baseline confounders, and taking those with the lowest consumption of potatoes as the reference (< 3.96 g/day), the participants with the highest consumption of potatoes (> 22.0 g/day) did not have any increased risk of mortality (HR=0.75; 95%CI: 0.53-1.07, p=0.11; Table 2).
Table 2.
Association between potato consumption and mortality
| Incidence (95% CI) per 1,000 persons-year | Basic-adjusted HR(95% CI)1 | P value | Fully-adjusted HR(95% CI)2 | P value | |
|---|---|---|---|---|---|
| < 3.95 g/day | 17 (14-21) | 1 [reference] | 1 [reference] | ||
| 3.96-8.55 g/day | 17 (14-21) | 0.94 (0.71-1.24) | 0.65 | 0.90 (0.68-1.19) | 0.45 |
| 8.56-15.67 g/day | 15 (12-19) | 1.00 (0.74-1.36) | 0.99 | 0.94 (0.69-1.30) | 0.72 |
| 15.68-22.0 g/day | 12 (8-16) | 0.81 (0.53-1.25) | 0.34 | 0.75 (0.48-1.16) | 0.19 |
| > 22.0 g/day | 15 (11-18) | 0.85 (0.61-1.17) | 0.32 | 0.75 (0.53-1.07) | 0.11 |
Basic-adjusted model included as covariates age (as continuous) and sex;
Fully-adjusted model included as covariates: age (as continuous); project; sex; job (manager/ professionals, craft, agricultural or sales workers, elementary occupations, housewife, pensioners, jobless, and missing information); educational level (primary school, secondary school, high school, graduated, illiterate, and missing information); systolic and diastolic blood pressure at baseline (both as continuous); adherence to Mediterranean diet (as continuous); body mass index (as continuous); presence of any disease (yes vs. no); alcohol intake (g/day, as continuous); total energy intake (Kcal, as continuous). Abbreviations: CI: confidence intervals; HR: hazard ratio.
Modelling the potato consumption as continuous (i.e. as increase in 10 g/day) did not substantially change our findings (fully-adjusted HR=0.93; 95%CI: 0.84-1.02; p=0.11).
Discussion
In this study, including more than 2,000 participants followed-up for over 10 years, we found that overall potato consumption was not associated with a higher risk of mortality, after taking in account several potential baseline confounders. Dietary patterns were also extracted from the data to account for complex combinations of micro and macronutrients.
In these people living in South Italy and so in an area in which Mediterranean culture is widely diffused, the consumption of potatoes was, however, very limited. In fact, having in mind that one potato weights about 100g (8), the mean intake in our cohorts was only 16g/daily suggesting that this food is rarely used. However, in other countries, potato consumption is really elevated. For example, recent data suggest that American citizens consumed about 115.6 pounds of white potatoes per year (17). It is noteworthy that in the USA of which about two thirds are French fries, potato chips or other frozen or processed potato products (17). However, also in Mediterranean countries, the consumption of potatoes is increasing suggesting that to have information regarding if this food could increase (or not) overall mortality could be of importance (18).
It is noteworthy that in our cohort, higher intake of potatoes was associated with some protective factors for mortality, particularly higher adherence to Mediterranean diet (19, 20), lower smoking rate, lower presence of chronic diseases and lower adiposity level (21).
The literature regarding potato consumption and mortality is very limited being, to the best of our knowledge, only one study assessing this potential relationship and reporting that among North American people, overall potato consumption was not associated with any increased risk of death, even if higher consumption of French fries was (7).
Several reasons could explain the lack of association between higher potato consumption and mortality risk. First, potatoes are rich in fibers, vitamins and micronutrients that could have counterbalanced the detrimental effects of their high glycemic index (2). Second, our study was made in a Mediterranean area that is deeply different in dietary patterns from North America and North Europe in which the other studies assessing potato consumption and medical conditions/mortality were made. For example, in the Osteoarthritis Initiative, people with a higher consumption of potatoes had a significant lower adherence to Mediterranean diet, whilst in the current study we observed the contrary (22). Moreover, most fat intake in this area is almost exclusively as olive oil. In line with this observation, we can finally hypothesize that in our participants the consumption of potatoes was associated with higher intake of vegetables (23). and extra-virgin olive oil (24) and these factors could further reduce the overall mortality risk. Moreover, we can hypothesize that in this population the use of fried potatoes is very limited, whilst it is likely that is unhealthy food could exponentially increase the risk of future metabolic and cardiovascular conditions and consequently mortality. However, this last hypothesis remains speculative because we don't have specific information regarding how potatoes were cooked in our study, representing a possible limitation of our findings.
The findings of our study must be interpreted within its limitations. The main one is that we were not able to assess cause-specific mortality. Second, the medical conditions were self-reported and could have introduced some level of bias. Regarding this point, due to the differences between the two cohorts included in this research, we were not able to include any information regarding important medical conditions, such as cancer. Fourth, nutritional intake recorded in older people could have suffered from selective and potentially inaccurate recall and this may have influenced our results. Moreover, the questionnaires used were different between the two studies and Physical activity data which is a well-known major confounding factor between diet and health are missing in our study
Finally, as we did not consider changes in dietary habits between baseline and the follow-up, this too could have introduced bias.
In conclusion, our data suggest that overall potato consumption was not associated with higher risk of death in older people living in a Mediterranean area. Since overall potato consumption is increasing worldwide, future studies are warranted to elucidate the role of potato consumption on allcause and cause-specific mortality.
Conflict of interest
none.
Authors' contribution
OAR, NV, CMG analyzed the data and wrote the paper; GV, CAM, NA, CA collected the data and analyzed the questionnaires; NM and MG gave a critical revision of the final version. All authors read and approved the final manuscript.
Funding
none.
Ethical Standards
All procedures were conducted in accordance with the guidelines in the Declaration of Helsinski and approved by the Ethics Committee of the IRCCS Saverio de Bellis.
Electronic supplementary material
Supplementary material is available for this article at https://doi.org/10.1007/s12603-018-1018-4 and is accessible for authorized users.
The MICOL study group also includes: Correale Mario, De Michele Giampiero, Mastrosimini Anna, Iacovazzi Palma, Laboratory of Clinic Pathology, IRCCS Saverio de Bellis, Castellana Grotte, Bari, Italy; Chiloiro Marisa, Noviello Marisa, Unit of Radiology, IRCCS Saverio de Bellis, Castellana Grotte, Bari, Italy; Burattini Osvaldo, Pugliese Vittorio, « Laboratory of Biostatistic and Epidemiology, IRCCS Saverio de Bellis, Castellana Grotte, Bari, Italy. The NUTRIHEP study group also includes: Caterina Bonfiglio, Laboratory of Epidemiology and Biostatistics, National Institute of Gastroenterology-Research Hospital, IRCCS “S. de Bellis Silvana Elba, Josè Petruzzi, Giampiero Buongiorno: Unit of Gastrenterology 1, IRCCS “S. de Bellis Elsa Lanzillotta, Unit of Clinical Pathology, IRCCS “S. de Bellis.
FOODS GROUPS COMPOSITION
References
- 1.Anderson GH, Soeandy CD, Smith CE. White vegetables: glycemia and satiety. Advances in nutrition (Bethesda, Md) 2013;4:356S–367S. doi: 10.3945/an.112.003509. 10.3945/an.112.003509 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Camire ME, Potatoes and Human Health, 2016, 685–704
- 3.Muraki I, Rimm EB, Willett WC, Manson JE, Hu FB, Sun Q. Potato Consumption and Risk of Type 2 Diabetes: Results From Three Prospective Cohort Studies. Diabetes care. 2016;39:376–384. doi: 10.2337/dc15-0547. 10.2337/dc15-0547 PubMed PMID: 26681722. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Halton TL, Willett WC, Liu S, Manson JE, Stampfer MJ, Hu FB. Potato and french fry consumption and risk of type 2 diabetes in women. The American journal of clinical nutrition. 2006;83:284–290. doi: 10.1093/ajcn/83.2.284. 10.1093/ajcn/83.2.284 PubMed PMID: 16469985. [DOI] [PubMed] [Google Scholar]
- 5.Larsson SC, Wolk A. Potato consumption and risk of cardiovascular disease: 2 prospective cohort studies. The American journal of clinical nutrition. 2016;104:1245–1252. doi: 10.3945/ajcn.116.142422. 10.3945/ajcn.116.142422 PubMed PMID: 27680993. [DOI] [PubMed] [Google Scholar]
- 6.Borch D, Juul-Hindsgaul N, Veller M, Astrup A, Jaskolowski J, Raben A. Potatoes and risk of obesity, type 2 diabetes, and cardiovascular disease in apparently healthy adults: a systematic review of clinical intervention and observational studies. American Journal of Clinical Nutrition. 2016;104:489–498. doi: 10.3945/ajcn.116.132332. 10.3945/ajcn.116.132332 PubMed PMID: 27413134. [DOI] [PubMed] [Google Scholar]
- 7.Veronese N, Stubbs B, Noale M, et al. Fried potato consumption is associated with elevated mortality: an 8-y longitudinal cohort study. The American journal of clinical nutrition. 2017;106:162–167. doi: 10.3945/ajcn.117.154872. 10.3945/ajcn.117.154872 PubMed PMID: 28592612, PMCID 5486204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Food, Agriculture Organization of the United N. Potato world: Production and Consumption -International Year of the Potato 2008.
- 11.Zhu Y, Pandya BJ, Choi HK. Prevalence of gout and hyperuricemia in the US general population: The National Health and Nutrition Examination Survey 2007–2008. Arthritis and rheumatism. 2011;63:3136–3141. doi: 10.1002/art.30520. 10.1002/art.30520 PubMed PMID: 21800283. [DOI] [PubMed] [Google Scholar]
- 12.Kroke A, Klipstein-Grobusch K, Voss S, Moseneder J, Thielecke F, Noack R, Boeing H. Validation of a self-administered food-frequency questionnaire administered in the European Prospective Investigation into Cancer and Nutrition (EPIC) Study: comparison of energy, protein, and macronutrient intakes estimated with the doubly labeled water, urinary nitrogen, and repeated 24-h dietary recall methods. The American journal of clinical nutrition. 1999;70:439–447. doi: 10.1093/ajcn/70.4.439. 10.1093/ajcn/70.4.439 PubMed PMID: 10500011. [DOI] [PubMed] [Google Scholar]
- 13.Leoci C, Centonze S, Guerra V, Cisternino A G M. Validazione e riproducibilità di un questionario di frequenza alimentare semi-quantitativo autosomministrato. (Reliability and validity of a self administered semi-quantitative food frequency questionnaire.). G Ital Nutr Clin Prev. 1993;2:58–59. [Google Scholar]
- 14.Edefonti V, Bravi F, Garavello W L, Vecchia C, Parpinel M, Franceschi S D, Maso L, Bosetti C, Boffetta P, Ferraroni M. Nutrient-based dietary patterns and laryngeal cancer: evidence from an exploratory factor analysis. Cancer Epidemiology and Prevention Biomarkers. 2010;19:18–27. doi: 10.1158/1055-9965.EPI-09-0900. 10.1158/1055-9965.EPI-09-0900 [DOI] [PubMed] [Google Scholar]
- 15.Edefonti V, Decarli A L, Vecchia C, Bosetti C, Randi G, Franceschi S D, Maso L, Ferraroni M. Nutrient dietary patterns and the risk of breast and ovarian cancers. International journal of cancer. 2008;122:609–613. doi: 10.1002/ijc.23064. 10.1002/ijc.23064 PubMed PMID: 17764109. [DOI] [PubMed] [Google Scholar]
- 16.Grambsch PM, Therneau TM. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika. 1994;81:515–526. 10.1093/biomet/81.3.515 [Google Scholar]
- 17.Service UDoAER (2012) Vegetables & Pulses: Potatoes http://www.ers.usda.gov/topics/crops/vegetables-pulses/potatoes.aspx Accessed 10 February 2017
- 18.Tur JA, Romaguera D, Pons A. Food consumption patterns in a mediterranean region: does the mediterranean diet still exist. Annals of nutrition & metabolism. 2004;48:193–201. doi: 10.1159/000079754. 10.1159/000079754 [DOI] [PubMed] [Google Scholar]
- 19.May AM, Struijk EA, Fransen HP, et al. The impact of a healthy lifestyle on Disability-Adjusted Life Years: a prospective cohort study. BMC medicine. 2015;13:39–40. doi: 10.1186/s12916-015-0287-6. 10.1186/s12916-015-0287-6 PubMed PMID: 25858161, PMCID 4362634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Limongi F, Noale M, Gesmundo A, Crepaldi G, Maggi S. Adherence to the Mediterranean Diet and All-Cause Mortality Risk in an Elderly Italian Population: Data from the ILSA Study. The journal of nutrition, health & aging. 2017;21:505–513. doi: 10.1007/s12603-016-0808-9. 10.1007/s12603-016-0808-9 [DOI] [PubMed] [Google Scholar]
- 21.Veronese N, Li Y, Manson JE, Willett WC, Fontana L, Hu FB. Combined associations of body weight and lifestyle factors with all cause and cause specific mortality in men and women: prospective cohort study. BMJ. 2016;355:i5855–i5856. doi: 10.1136/bmj.i5855. 10.1136/bmj.i5855 PubMed PMID: 27884868, PMCID 5122318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Veronese N, Stubbs B, Noale M, et al, Fried potato consumption is associated with elevated mortality: an 8-year longitudinal cohort study, The American journal of clinical nutrition, 2017 [DOI] [PMC free article] [PubMed]
- 23.Kok FJ, Kromhout D. Atherosclerosis—epidemiological studies on the health effects of a Mediterranean diet. European journal of nutrition. 2004;43(1):I/2–5. doi: 10.1007/s00394-004-1102-x. 10.1007/s00394-004-0432-z PubMed PMID: 15052492. [DOI] [PubMed] [Google Scholar]
- 24.Guasch-Ferre M, Hu FB, Martinez-Gonzalez MA, et al. Olive oil intake and risk of cardiovascular disease and mortality in the PREDIMED Study. BMC medicine. 2014;12:78. doi: 10.1186/1741-7015-12-78. 10.1186/1741-7015-12-78 PubMed PMID: 24886626, PMCID 4030221. [DOI] [PMC free article] [PubMed] [Google Scholar]
Uncited references
- 9.Cozzolongo R, Osella AR, Elba S, et al. Epidemiology of HCV infection in the general population: a survey in a southern Italian town. Am J Gastroenterol. 2009;104:2740–2746. doi: 10.1038/ajg.2009.428. 10.1038/ajg.2009.428 PubMed PMID: 19638964. [DOI] [PubMed] [Google Scholar]
- 10.Osella A D, Mdel P, Cozzolongo R, et al. Overweight and Obesity in Southern Italy: their association with social and life-style characteristics and their effect on levels of biologic markers. Revista de la Facultad de Ciencias Medicas (Cordoba, Argentina) 2014;71:113–124. [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
FOODS GROUPS COMPOSITION