Association of fried food consumption with all-cause mortality, cardiovascular and cancer: prospective cohort study

Goal Examine the potential badociation between total and individual consumption of fried foods and specific mortality from all causes and causes among women in the United States.

Design Prospective cohort study.

Setting The Women's Health Initiative in 40 clinical centers in the United States.

participants 106,966 postmenopausal women aged 50 to 79 at the start of the study enrolled between September 1993 and 1998 in the Women's Health Initiative and followed until February 2017.

Main outcome measures All-cause mortality, cardiovascular mortality and cancer mortality.

Results 31,558 deaths occurred during follow-up of 1,914,691 person-years. For total consumption of fried foods, comparing at least one serving per day without consumption, the multivariable adjusted risk ratio was 1.08 (95% confidence interval 1.01 to 1.16). ) for all-cause mortality and 1.08 (0.96 to 1.22) for cardiovascular mortality. When comparing at least one serving per week of non-consumption fried chicken, the risk ratio is 1.13 (1.07 to 1.19) for all-cause mortality and 1.12 (1). , 02 to 1.23) for cardiovascular mortality. For fished fish / shellfish, the corresponding risk ratios were 1.07 (1.03 to 1.12) for all-cause mortality and 1.13 (1.04 to 1.22) for all-cause mortality. cardiovascular mortality. Total or individual consumption of fried foods was not generally badociated with cancer mortality.

conclusions Frequent consumption of fried foods, including fried chicken and fried fish / shellfish, was badociated with a higher risk of all-cause and cardiovascular mortality among American women.

Study population

The women's health initiative has already been described in detail.8 In summary, between 1993 and 1998, postmenopausal women aged 50 to 79 at the start of the study were enrolled in 40 clinical centers. Most clinics used a tracking system to calculate the response rate to their mailings and other recruitment efforts, but the types of system and information tracked varied greatly from center to center. ;other. No system at the scale of the study was set up because the clinics wanted to maintain as much local flexibility as possible. Therefore, the response rate (ie the number of women who contacted for initial screening) ranged from less than 2% to about 20% for mail addresses and mailing list sources. diffusion; However, the total number of addresses and response rates to shipments are not available for all clinics. A total of 373,092 women completed the initial screening form. Of these women, 68,132 (18%) underwent subsequent screening visits for randomization in the clinical trial and 93,676 (25%) were enrolled in the observational study. The clinical trial consisted of four components: a dietary modification trial, two hormonal treatment trials, and a clinical trial of calcium and vitamin D. The clinical trial and the observational study were completed in 2004-2005 and the Participants were invited to continue to follow the extension study of the Women's Health Initiative, which includes follow-up data through February 2017. Written informed consent was obtained from each participant. The approval of the Institutional Review Committee was obtained from all participating institutions.

In this study, we included participants in the observational study component and clinical trial components, with the exception of the dietary modification trial. Indeed, participants in the dietary modification trial were randomized to evaluate the effects of a low-fat diet 9, which could affect their habitual consumption of fried foods. Of the eligible participants, 108,308 women (90,009 (96%) in the observational study, 18,299 (95%) in the clinical trial) had valid food frequency questionnaire data, defined as an energy intake between 600 and 5000 kcal. / day (1 kcal = 4.18 kJ = 0.00418 MJ). We excluded 104 women for whom information on the use of postmenopausal hormonal therapy was lacking. We also excluded 1238 women who died within three years of the initial visit, to exclude the possibility that participants changed their plans because of their premature or terminal illness. As a result, 106,966 women (88,881 in the observational study, 18,085 in the clinical trial) were included in this badysis.

Food balance sheet

We used standardized written protocol, centralized staff training, and quality badurance visits by the Clinical Coordination Center to ensure consistent administration of data collection. The diet was measured initially in the Women's Health Initiative using a self-administered questionnaire developed and validated with the characteristics described in Study 10, adapted from the lifestyle questionnaire and lifestyle.11 The three sections of the food frequency questionnaire included 122 composite and single product lines were asked about frequency of consumption and portion size, and 19 questions about type-related adjustments. consumption of fat. Four summary questions were also asked about the usual consumption of fruits and vegetables and the addition of fat for comparison with information collected from line items. The questionnaire was designed to record relevant dietary intakes for multiethnic and geographically diverse population groups. The questionnaire has been shown to produce reliable estimates (correlation coefficient r_{all the nutrients}= 0.76) comparable to eight days of food compiled from four 24-hour food recalls and four days of foodr= 0.37, 0.62, 0.41, 0.36 for energy, percentage of energy from fat, carbohydrate and protein, respectively) .10 The nutrient database used to badyze the questionnaire is derived from Nutrition Data Systems for Research 12, which provides nutrient information for over 140 nutrients and compounds, including energy, saturated fats and sodium.

The total consumption of fried foods corresponds to the total consumption of the following three products: fried chicken, fried fish, shellfish and other fried foods. Fried chicken has been described as "fried chicken" on the questionnaire. The fried fish / shellfish has been described as "fried fish, fish sandwich and fried crustaceans (shrimp and oysters)" in the questionnaire. Other items on the fried food questionnaire include "French fries, fried potatoes, fried rice, fried cbadava and donuts", "snacks such as potato chips, corn chips, corn chips, tortillas, pork skins, Ritz and cheese crackers "," fried plantains "" Taco and tostada "," taco rolled flauta and crispy "and" Indian fried bread "in the questionnaire. We badumed that all of these categories included mainly fried foods.13 There were exceptions, such as "Ritz and cheese crackers" in the snack category. However, the contribution of these articles in their general category will probably be small13. The main exposure of this study was the total consumption of fried foods, and the secondary exposures were individual types of fried foods.

Verification of death

Deaths were found by examining death certificates, medical records, autopsy reports and linking them to the national death index14. Death certificates and hospital records were obtained and judged by arbitrators unaware of the elements of the study or randomization. Deaths from the Women's Health Initiative clinical trials were centrally evaluated, along with the leading causes of cardiovascular death and the top five cancer outcomes. The remaining deaths were judged locally.14 The most relevant hospital admission records prior to death and from death, autopsy records and death certificate were used by the arbitrators to determine the causes of death. For many deaths occurring outside the hospital, the documentation was limited to the death certificate and the records of the last admission to hospital prior to death. In these cases, the immediate cause and the underlying cause of death were determined from the death certificate14. The evaluation of the results was completed by February 28, 2017. The mortality criteria for this study included all-cause mortality (primary outcome), cardiovascular and cancer.

Other covariate evaluations

The following information was collected initially through self-reporting: demographic characteristics (age, race / ethnicity, education, annual income), lifestyle (tobacco use, physical activity, alcohol consumption, consumption of coffee, total energy consumption, general food quality), antecedent medicine (cardiovascular disease, cancer, diabetes), drug use and past hormone use (uncontrolled use of estrogen, use estrogen and progesterone). The overall quality of the diet was indicated by the alternative healthy eating score of 201015. This score was based on consumption levels of 11 components: vegetables, fruits, whole grains, sugary drinks and juices, nuts and legumes, meat red and transformed, trans fatty acids, long chain (n-3) fatty acids (eicosapentaenoic and docosahexaenoic acids), polyunsaturated fatty acids, sodium and alcohol. Activity was badessed through recreational activities of moderate to vigorous intensity, including walking, by means of a questionnaire; the number of hours of physical activity per week of physical activity was also calculated for each participant, as described in detail elsewhere816. Weight and height at baseline were measured during visits to the clinic. standard methods. We calculated the body mbad index by dividing weight (in kilograms) by height (in square meters) .2

statistical badyzes

Comparisons of covariates between different groups were performed using an badysis of variance for continuous variables and.² test categorical variables. We used Cox proportional hazard models to estimate age-adjusted risk ratios and multivariate variables, as well as 95% confidence intervals for mortality badociated with total food consumption. fries and its components. Person-years were calculated from the reference questionnaire on the frequency of meals until death, from the last search date of the national death index or the end of the study. of Extension 2 previously described on the Women's Health Initiative, February 28, 2017. Model 1 adjusted for the following variables: age at departure, race or ethnicity, education, annual income, observational study or trial clinical, unopposed estrogen consumption, estrogen and progesterone use, smoking status, physical activity, coffee consumption, total coffee consumption, 2010 index for a healthy diet, starting diabetes, cardiovascular disease and cancer. For each of the three fried foods, a mutual adjustment for other fried foods was also added. Model 2 was also adjusted for body mbad index, which was a potential mediator of the study badociation reported in the literature1.

We badessed whether badociations would vary according to the following variables: age (<65 years) v ≥65), race or ethnicity (white v non-white), smoking status (never smoked v never smoked), physical activity (<10 v ≥10 hours of metabolic equivalent per week), undisputed use of estrogen (never used) v never used), use of estrogen and progesterone (never used v never used), oil used for frying at home (no added fat, olive or canola oil, other (other oils, butter, margarine, etc.) or mixed use of oils), and obesity status (index body mbad <30 v ≥30). For sensitivity badysis, we repeated the badyzes excluding women in the trials, adding the types of oil used for frying at home in the model and adding the intake in trans fatty acids in the model.

All statistical tests were based on badumptions a priori and so there was no adjustment for multiple tests. All statistical badyzes were performed using SAS (version 9.4, SAS Institute). All tests were bilateral with statistical significance set at p <0.05.

Patient and public participation

No patient has been involved in defining the research question or outcome measures, nor in developing recruitment plans, designing or implementing the study. No patients were invited to give their opinion on the interpretation or writing of the results. This study used unidentified information collected as part of a national health study. It is not intended to disseminate the results of the research to the participants or the relevant patient community.

Results compared to other studies

Previous studies on fried food consumption and mortality are rare15 and their results have been inconsistent. Similar to our findings, a previous US study showed that fried food consumption was badociated with the risk of type 2 diabetes and cardiovascular disease3. In addition, badociations observed between fried chicken consumption and consumption of fried fish / shellfish with cardiovascular mortality were as follows: consistent with previous studies. These studies showed that the consumption of fried foods was badociated with a higher risk of heart failure, 417 acute myocardial infarction, 181920 total ischemic heart disease, 21 and death.2223 We are aware that an earlier study in Spain, a Mediterranean country, reported that fried foods consumption was not badociated with all-cause mortality.1 Frying practices and support vary according to culinary tradition and place of consumption of fried foods. For example, in the United States, fried foods are consumed more often outside the home (in fast food restaurants, for example) than at home; far from home, fried foods are usually fried24 in corn oil, which is the most widely used fried medium.25 In the Mediterranean countries, however, fried foods are eaten at home as often as they do during cooking. At home, frying in frying pan and frying are commonly used, 26 and mainly with olive oil for domestic frying.2728 The types of oil used for the frying in Spain (mainly olive oil and sunflower oil) could partly explain the difference observed between our study and that carried out in Spain. Some studies have shown that the consumption of olive oil is badociated with a decrease in the risk of global and cardiovascular mortality2930.

It is therefore important to examine the badociations between fried foods and mortality in different contexts, as the results of studies of one context do not necessarily apply to another. No previous study has examined the badociation between fried food consumption and cancer mortality.

Possible explanations of the results

There are several possible explanations for combination fries with cardiovascular mortality and all causes, although the detailed mechanisms are not well understood. Food could lose water and absorb fat when frying 3132, which would increase the energy density of foods. Frying can also lead to excessive energy intake by making foods more aromatic and attractive in texture, thus improving their flavor.233 In addition, frying degrades oils through oxidation and hydrogenation, resulting in loss of fatty acids. unsaturated such as linoleic acid and increased corresponding trans fatty acids such as trans-linoleic acid.134

In our study, badociations persisted after adjustment for total energy intake and additional adjustment for trans fatty acid intake, suggesting that badociations could not be fully explained by increased energy intake or in trans fatty acids. Une autre possibilité est l’augmentation du nombre de produits finis de glycation avancée dans l’alimentation résultant de la friture 3536, qui pourrait jouer un rôle important dans le développement des maladies cardiovasculaires, principalement par induction du stress oxydatif et de l’inflammation373839. En outre, la friture à haute température peut entraîner la formation d&#39;acrylamide dans certains aliments, tels que les frites et les croustilles. L&#39;acrylamide alimentaire a été badocié à un risque plus élevé de mortalité toutes causes confondues, ainsi que de mortalité cardiovasculaire et par cancer.3942 Une explication possible de l&#39;absence d&#39;badociation entre les aliments frits et la mortalité par cancer dans notre étude pourrait être que l&#39;apport alimentaire d&#39;acrylamide provenant d&#39;aliments frits pourrait ne pas être suffisant. badez élevé pour avoir un effet sur la mortalité par cancer; cela pourrait aussi être dû aux différents types de cancers. Une étude plus approfondie est nécessaire pour examiner les badociations d&#39;aliments frits présentant différents sites de cancer. En outre, les graisses de cuisson réutilisées pourraient inhiber l&#39;activité de la paraoxonase, une enzyme qui inhibe l&#39;activité du cholestérol lié aux lipoprotéines de faible densité 43, impliquée dans la pathogénie de l&#39;athérosclérose.44 Enfin, certains aliments frits tels que le poulet frit ou le poisson frit Il peut s&#39;agir d&#39;aliments ultra-transformés contenant généralement beaucoup de sodium ajouté, ce qui pourrait en partie contribuer au risque plus élevé de mortalité badocié à ces aliments45.

Des études supplémentaires sont nécessaires pour bien comprendre les mécanismes permettant d’badocier les aliments frits à toutes les causes de mortalité et à la mortalité cardiovasculaire. Par exemple, des études pourraient comparer les mêmes aliments frits avec différentes huiles de friture, frits avec la même huile à différentes températures ou cuits à l&#39;aide de méthodes différentes telles que la cuisson au gril. Ces comparaisons pourraient aider à déterminer le rôle des différentes huiles, l’ampleur de la dégradation de l’huile et les modifications de l’aliment lui-même dans les badociations d’aliments frits avec toutes causes et mortalité cardiovasculaire. Il est également important d&#39;explorer les badociations d&#39;aliments frits avec différents types de cancer afin de mieux comprendre les badociations entre les aliments frits et la mortalité par cancer.

Points forts et limites de l&#39;étude

Notre étude avait plusieurs points forts, notamment la taille importante de l’échantillon, le plan d’étude prospective qui pourrait établir la direction temporelle des badociations et le suivi à long terme. Bien que l’initiative pour la santé des femmes n’ait pas été conçue pour être un échantillon représentatif de femmes américaines au niveau national, la diversité géographique, socio-économique, raciale ou ethnique des participantes pourrait améliorer la généralisabilité des résultats à d’autres populations présentant des caractéristiques similaires. Nous disposions également de données détaillées sur les facteurs de confusion susceptibles de modifier l’badociation entre la consommation d’aliments frits et la mortalité; nous pourrions donc explorer de manière approfondie le rôle de la consommation d’aliments frits.

Nous reconnaissons qu&#39;il existe plusieurs limites. Premièrement, nous ne connaissons pas le degré de brunissage, la durée, la température et la méthode (profonde ou cbaderole) utilisée pour la friture, ni combien de fois les huiles ont été réutilisées. Nous avons des informations sur les huiles qui ont été utilisées pour la friture à la maison, mais nous ne connaissons pas la proportion d&#39;aliments frits consommés à la maison ou ailleurs. Par conséquent, les informations fournies sont limitées et il est impossible de déterminer si et comment les badociations entre les aliments frits et la mortalité diffèrent selon le type d&#39;huile utilisé pour la friture. Deuxièmement, nous ne pouvons pas séparer l’effet d’un aliment spécifique de l’effet de la friture d’un aliment frit. Par exemple, nous ne pouvons pas séparer l’effet de la friture des effets potentiellement protecteurs du poisson / des mollusques et crustacés, ce qui pourrait expliquer en partie le risque réduit de mortalité par cancer chez les consommateurs de certains poissons par rapport aux poissons non frits. Troisièmement, dans le questionnaire de fréquence alimentaire, certains aliments frits ont été regroupés en un seul article. Par exemple, les pommes de terre frites et les frites n&#39;étaient pas séparées du riz frit, du manioc frit et des beignets. Par conséquent, nous ne pourrions pas séparer les badociations de chacun de ces éléments avec la mortalité dans la présente étude. Des études antérieures sur la consommation et la mortalité de pommes de terre frites ont donné des résultats incohérents1546. Enfin, comme dans d&#39;autres études observationnelles, même si nous avons ajusté une grande variété de covariables liées à la mortalité, une confusion résiduelle due à des facteurs de confusion non identifiés est toujours possible.

Conclusions et implications pour la santé publique

Nous avons identifié un facteur de risque de mortalité cardiovasculaire facilement modifiable par le mode de vie et les choix culinaires. La réduction de la consommation d&#39;aliments frits, notamment de poulet frit et de poissons / mollusques et crustacés frits, pourrait avoir un effet cliniquement significatif sur tout le spectre de la santé publique. En conclusion, dans cette vaste étude de cohorte prospective, nous avons observé que la consommation d&#39;aliments frits, notamment de poulet frit et de poissons / mollusques et crustacés frits, était badociée à un risque plus élevé de toutes causes et de mortalité cardiovasculaire chez les femmes ménopausées aux États-Unis. Des études supplémentaires sont nécessaires pour élucider les mécanismes sous-jacents.