Matina Kouvari1, Venetia Notara1, Demosthenes B. Panagiotakos1*, Moscho Michalopoulou1, Eleni Kalli1, Margarita Verdi1, Nikos Vasileiou1, Evangelos Papataxiarchis1, Despoina Tzanoglou1, Georgia Kouli1, Yannis Mantas2, Yannis Kogias3, Petros Stravopodis4, George Papanagnou5, Spyros Zombolos6 and Christos Pitsavos7
Objective: The aim of this work, and under the context of the GREECS study, was to evaluate the association between sweets consumption and the 10 year (2004-2014) incidence of Acute Coronary Syndrome (ACS), among cardiac patients. Methods: From October 2003 to September 2004, 2,172 ACS consecutive patients from 6 major Greek hospitals were enrolled; during 2013-2014, the 10-year follow-up was performed in 1,918 patients (88% participation rate); development of fatal or non-fatal ACS was the outcome of interest and recorded through medical records or registries. Among others, sweets (i.e., cakes, chocolate, traditional pastries, Lenten sweets, pies and other common Greek sweet treats) consumption at baseline examination was assessed using a semi-quantitative food frequency questionnaire. Results: A positive association was observed between sweets consumption (at least one portion per week vs. other) and ACS incidence after taking into account various potential confounders, including diabetes mellitus and obesity [OR=1.23, 95% CI (0.99, 1.53) p=0.060]. However, after stratified the analysis by education status (i.e., ≤ 9 vs. >9 years of school) the above association remained significant only among patients with higher educational status [OR=1.50, 95%CI (0.93, 2.40), p=0.095]; moreover, no significant interaction effect was observed between patients’ financial status and sweets consumption, on the tested outcome (p for interaction =0.56). Conclusion: Moderation or even avoidance of sweets consumption, in the context of healthier dietary habits, seems to be of high importance to reduce the risk of re-current cardiac events and improve disease prognosis among cardiac patients.
Acute coronary syndrome; Risk; Cardiovascular; Sweets; Socioeconomic; Education; Nutrition
Almost nobody can resist sweets; it is a "guilty" pleasure, a temptation to which many people succumbs. Indeed, the western dietary pattern, which is mainly characterized by high consumption of various unhealthy food choices, it is also characterized by increased sweets consumption. The last few decades, accumulating data suggest that the developing countries, as well, experience a shift in their food habits towards the western model [1,2]. Dietary guidelines to reduce cardiovascular risk clearly suggest avoiding sweets consumption due to their great amount of sucrose, trans /saturated fatty acids, high energy-density and glycemic index/load . In accordance to these guidelines, the Mediterranean dietary pattern, that has long been studied as a prototype with numerous health benefits, recommends occasional consumption of sweets (i.e., monthly basis) and in small amounts . However, products with added sugars are preferable for the reason that they cost less, they are tasteful, and give the feeling of satiety. Additionally, humans are genetically predisposed to prefer sweet tastes from the early stages of their life; newborn babies seem to like the sweet flavor of their mothers’ breast milk. What is more, evolution has discriminated the types of foods initially preferred or rejected by making the sensory system detect calorie-rich foods and full of carbohydrates, as these, apart from representing the primary source of energy, offer a powerful hedonic appeal thanks to their palatability. On the contrary, the bitter ones, recognized as potentially poisonous, are rejected. The aforementioned claims are supported by certain brain and hormone mechanisms concerning not only the pediatric but also the adult population, which may lead to potential addiction equivalent to tobacco or alcohol [5,6]. Furthermore, this innate preference is empowered due to the worldwide sweet consumption either as a part of tradition or as a result of technological improvements in food industries .
Sweets have been examined over a long period with regard to some chronic diseases, such as diabetes mellitus and obesity, because of their great amount of sucrose, trans /saturated fatty acids, high energy-density and glycemic index/load, characteristics which lead to a rather harmful combination [8- 10]. Nevertheless, studies as regards cardiac outcomes are rare and poorly understood. Although the recent American Heart Association (AHA) Guidelines in the primary cardiovascular disease (CVD) prevention , highlight the adverse effect of sweets consumption on CVD in the context of an unhealthy diet, studies examining the independent, of overall dietary habits, effect of sweets in the prevention of heart disease are lacking [11,12]. Furthermore, the potential interaction effect of the socio-economic component and sweets consumption, on CVD outcomes has never been studied. Therefore, the aim of the present work was to investigate the association of sweets consumption on the 10-year ACS incidence among cardiac patients in relation to their educational and financial status. The working research hypothesis was that ACS patients, mainly those of low educational or financial status, who regularly (e.g., weekly) consumed sweets, had higher risk of a recurrent ACS incidence compared with those who avoided eating sweets.
The GREECS (GREEK acute Coronary Syndrome) is a prospective observational study that established in 2003 and aimed to evaluate the role of various risk factors on the CVD development and prognosis among ACS patients . During October 2003 - September 2004, n=2,172 consecutive patients with discharge diagnosis of ACS (i.e., acute myocardial infarction (AMI) or unstable angina (UA)) that were hospitalized in the cardiology clinics or the emergency units of six major General Hospitals in Greece (i.e., Hippokration hospital in Athens and the general prefectural hospitals in Lamia, Karditsa, Halkida, Kalamata and Zakynthos island) were enrolled into the study (participation rate varied from 80% to 95%) (1649 men, 65 ± 13 years old and 523 women, 72 ± 11 years old, p<0.001). The hospitals were selected in order to represent populations with various socio-economic, cultural and regional characteristics. By the exception of the metropolitan city of Athens - where there are several other hospitals -, all the other hospitals cover the whole population of the aforementioned regions, including urban and rural areas. The hospitals were selected in order to represent populations with various socio-economic, cultural and regional characteristics.
Acute myocardial infarction was defined by typical rise and gradual fall (troponin) or more rapid rise and fall (CK-MB) of biochemical markers of myocardial necrosis with at least one of the following: (a) ischemic symptoms, (b) development of pathologic ST waves on the ECG, (c) ECG changes indicative of ischemia (ST segment elevation or depression) or (d) coronary artery intervention (e.g., coronary angioplasty) . Unstable angina was defined by the occurrence of one or more angina episodes, at rest, within the preceding 48-hours, corresponding to class III of the Braunwald classification . Braunwald classification was initially introduced to divide the patients with unstable angina according to their clinical characteristics and to provide better prognostic stratification. Medical information was retrieved through patient’s hospital records.
The study was approved by the Medical Research Ethics Committee of the participated Institutions and was carried out in accordance with the Declaration of Helsinki (1989) of the World Medical Association. All patients were informed about the aims and procedures of the study and signed an informed consent.
Socio-demographic and lifestyle characteristics studied at baseline, included (among others): age, sex, physical activity, dietary and smoking habits, years of school and financial status. Specifically, sweets consumption, which is the main exposure factor of the present work, was recorded through a validated Food Frequency Questionnaire (FFQ) as follows: never/rare consumption, 1-2 portions, 3-5 portions and >5 portions, per week; because of the small number of cases in the “3-5 portion” and “>5 portion” classes, the groups were merged into two, i.e., never/rare and often/very often (even 1 portion per week), for the statistical analyses. The studied sweets were: bakery (i.e., cakes/ sweetbreads), chocolate products, traditional Greek pastries, such as baklava, kadayif (angel-hair pastry), galaktoboureko (custard-filled pastry), bougatsa (creamy semolina/custard pudding), pies with fruits or jams, common sweet treats like olive oil/butter cookies, chocolates, lokums, louloumadhes (ballshaped doughnuts), spoon-sweets (like syropi-fruit preserves), rizogalo (rice-milk pudding), Lenten (taxini/semolina halva), as well as honey macaroons, kourabiedes, and honey-roll turnover (diples)]. Light sweet products or products with other sweeteners (i.e., sweeteners without calories etc) were not included. In addition, as far as the educational and financial level is concerned, the former was measured in years of school and classified as low (i.e., ≤ 9 years of school) and intermediate/high or academic (>9 years of school or technical/vocational or academic institutes). The use of the specific cut-off values was due to the fact that 9 years represents the obligatory education versus the up to 9-year education which represents subjects who either continued until the level of technical/vocational education or led to an academic level. Furthermore, concerning patients’ financial status, mean annual income of during the last three years was recorded through self reports; regarding people who were unemployed, the basic monthly allowance they received from the Social Service Office was considered to calculate annual income. Financial status was then classified as low/moderate (annual income<12.000€) and good/very good (>12.000€), following the tax policy of the Ministry of Economics.
Overall dietary habits were evaluated through a validated dietary index, the MedDietScore (range 0–55);  higher values of Med Diet Score indicate greater adherence to the Mediterranean diet (arbitrarily, values between 27-41 may consider as “moderate” adherence). To evaluate physical activity status of the patients during the past year a modified version of a self-reported questionnaire provided by the American College of Sports Medicine was used;  physical activity was defined as any engagement in activities of at least 3 times per week and for at least 30 minutes. Years of smoking were recorded; current smokers were defined as those who smoked at least one cigarette per day or have stopped smoking during the past 12 months, while the rest smokers were defined as past smokers. The rest of the patients were defined as never or rare smokers. Medical history included the baseline assessment of history of hypertension, hypercholesterolemia, diabetes mellitus and CVD (i.e., prior to the baseline CVD event), as well as the pharmaceutical management of these conditions. Body mass index (BMI) was calculated as weight (in Kg) divided by height (in m) squared. Overweight was defined as BMI between 25 and 29.9 kg/m2, while obesity as BMI greater than 29.9 kg/ m2 (no underweight cases, i.e., BMI<18.5 kg/m2, were observed). Further details about the aims, measurements and baseline procedures of the GREECS study may be found elsewhere [13,18].
Endpoints at 10-year follow-up
The endpoints studied in the 10-year follow-up were recurrent fatal or non-fatal ACS events. In particular, the development of a new AMI, angina pectoris, other identified forms of ischemia (WHO-ICD coding 410-414.9, 427.2, 427.6), heart failure of different types and chronic arrhythmias (WHO-ICD coding 400.0-404.9, 427.0 -427.5, 427.9-), were recorded. All patients were interviewed by using a standard questionnaire. During the 10-year (2004-14) follow-up, of the n=2172 initially enrolled patients, n=1,918 were found (88% participation rate). The rest n=254 (11.7%) of the patients that were lost in the 10-year follow-up were considered as censored. No differences were observed between those participated in the 10-year followup and those lost in all baseline clinical and lifestyle factors (all p-values>0.50).
Continuous variables are presented as mean values ± standard deviation, while categorical variables are presented as absolute (n) and relative frequencies (%). Associations between normally distributed continuous variables (MedDietScore, body mass index and age) and groups of the patients were evaluated through the Student’s t-test for independent samples. Normality of these was tested through the P-P plot and equality of variances through Levene’s test. Associations for years of school that was abnormally distributed and groups of study was evaluated through Mann-Whitney test. Associations between categorical variables were tested by the use of the chi-squared test. The association between sweets consumption and 10-year ACS fatal or non-fatal incidence was evaluated by multiply logistic regression analysis, after controlling for socio-demographic, clinical and lifestyle characteristics of the patients. Results are presented as odds ratio (OR) and their corresponding 95% confidence intervals (95%CI). Time to event was also taken into account and Cox proportional hazards models were applied; but it was decided to keep the results from logistic regression as the models had better correct classification rates for the cases and some time-points were missing. The tested hypothesis was also evaluated after splitting the sample in two groups concerning the financial and educational level. A probability value of 5% was considered as statistically significant. All statistical calculations were performed on the SPSS version 21 software (IBM SPSS Inc, Athens, Greece).
The fatal or non-fatal 10-year ACS incidence was n=811 (37.3%) (38.8% in men and 32.9% in women, p=0.016). Median time of event was not taken into account because of some missing cases with exact event time.
Baseline dietary habits data analysis revealed that n=885 (41%) patients reported never/rare sweet consumption, while n=1,287 (59%) reported at least 1 portion per week (n=876 (68%) patients (1-2 portions/week), n=322 (25%) (3-5 portions/week) and n=89 (7%) (>5 portions/week). In table 1 various clinical, lifestyle and behavioural characteristics of the patients by sweets consumption are presented. Patients reported even 1 portion of sweet/week were more likely to have an MI than unstable angina (p=0.009), were characterized by higher frequency of smoking (p=0.004) and family history of CVD (p=0.001), but lower frequency of diabetes mellitus (p<0.001), whereas no significant association was observed in other factors, as compared with those reported never/rare consumption (Table 1).
In table 2 the association of sweets consumption with the 10 year ACS event rate, after taking into account conventional confounders is presented. As it can be seen, ACS patients with even 1 portion of sweet /week consumption had 23% higher risk of a recurrent ACS event, as compared with the ones who reported never/rare consumption, after various adjustments made (p=0.06). No significant interactions were observed between sweets’ consumption and diabetes history (p=0.887) or age (p=0.435) or sex (p=0.905) of the patients (Table 2).
Among the aims of the present work was also to test the potential moderating effect of socio-economic factors on the tested association. In table 3 various clinical, lifestyle and behavioural characteristics of the patients by educational and financial group are presented. No significant associations were observed as regards sweets consumption and education or financial status of the patients. Moreover, as it can be seen, patients belonging to the low educational status were older (p<0.001), more prone to suffer from hypertension (p=0.001) and diabetes mellitus (p=0.001), more physically inactive (p<0.001) but less likely to smoke (p<0.001), while they did not seem to differ in the dietary habits, as compared with the ones reported moderate/higher education. Concerning the stratified by financial status analysis, patients with low/moderate income were older (p<0.001), with higher prevalence of diabetes (p=0.01) and more physically inactive (p=0.001) whereas they had no difference in the sweets consumption and MedDietScore, as compared with those with better financial level; on the contrary, the latter group was more likely to have a family history of CVD (p=0.001) (Table 3).
The association of sweets consumption with the ACS 10-year survival by education group, and after taking into account various confounders, is presented in table 4. Data analysis revealed that the adverse effect of sweets in the ACS prognosis was significant only in the patients’ group with moderate/high educational status (p for interaction between education and sweets consumption on the outcome=0.007). In table 5, the association of sweets consumption with the ACS 10-year survival by financial group is also presented. After taking into account various potential confounders no significant association of sweets consumption on ACS risk was observed, in both financial status groups (p for interaction=0.56) (Tables 4 and 5).
The findings of the present work suggested that sweets consumption was associated with increased risk of recurrent ACS events, among cardiac patients, after taking into account several potential confounders. Furthermore, the aforementioned relationship remained significant only in patients with moderate/ high educational status, highlighting a social impact on the tested relationship, while no moderation effect of patients’ financial status was revealed. Despite the limitations of the present work due to its observational nature, the reported findings are of major importance for public health, since they highlight not only the aggravating role of sweets consumption in the ACS prognosis but also the fact that the high education level does not seem to be as such protective as usually believed.
At first, it should be discussed why the sole effect of sweets consumption on ACS risk was evaluated, instead of a more holistic dietary approach. It is true that the majority of the studies nowadays follow the approach to evaluate whole dietary patterns, like the western-type or Mediterranean or prudent or “healthy” type of diet, instead of food-specific, as regards CVD risk or other health outcomes. Although dietary pattern approach has many benefits instead of specific food ascertainment on health related outcomes, sweets consumed by a variety of ways, which cannot be identified in the common dietary patterns. Thus, its solely assessment as regards health outcomes should be undertaken, in order to lead to robust conclusions. Although, the relationship between various dietary patterns, as well as a variety of foods, and cardiometabolic outcomes has, extensively been studied, data regarding sweets consumption are sparse in the literature [11,12]. Sweets consumption has been associated with increased risk of diabetes mellitus and obesity, particularly due to high sucrose-concentration, energy density and glycemic load/index [9,10,19].
Recent literature makes some suggestions about the role of sugars as an independent risk factor for CVD, highlighting its aggravating role due to not only the empty calories leading to excess weight, but also the association with other major risk factors, mainly several metabolic perturbations, like abdominal obesity dyslipidemia, hypertension and insulin resistance [20,21]. The present In the AHA guidelines, the recommended added sugar consumption varies from 5 to 9 teaspoons per day, depending on the calorie level; according to a USDA database most of the widely consumed sweets have a sugar content 45g per 100g only by such products [22,23] work is one of the few studies giving light to the association of sweets consumption and re-current cardiac episodes. The mechanisms, through which sweets exert their aggravating role in the pathogenesis of CVD, depend on their nutrient composition. Sucrose is the major dietary determinant in this food group. Under no circumstances, could it be denied the multifactor harmful role of overconsumption of added sugars and it is fructose blamed for it. This causes high uric acid resulting in increasing blood pressure while it is, also, associated with increased triglycerides and low-density cholesterol (LDL-C) versus decreased high-density cholesterol (HDL-C) and abnormal insulin resistance through synthesis of fat in the liver. Another proposed pathway has to do with the association of fructose with inflammation markers which are on the top of the list concerning the pathogenesis of CVD events [20,24]. Apart from sucrose, the majority of sweets contain unhealthy fats; trans fatty acids (TFA) in industrialized products, have adverse effects in the lipid profile and lead to endothelial dysfunction, either before or after replacing the mono/poly unsaturated fatty acids (MUFA/PUFA), according to clinical trials and meta-analysis [25,26]. Concerning saturated fatty acids (SFA), although a recent meta-analysis revealed doubtful non clear effect as regards cardiac risk , a replacement with a higher carbohydrate intake, particularly the refined ones, can exacerbate the atherogenic dyslipidemia, insulin resistance and obesity . In addition, there are some other pathways by which sweets consumption, i.e., the high glycemic load and the energy density, may also lead to high CVD risk [29,30].
Although the above nutritional components are common among the majority of sweets, scientific evidence support that some of them have cardioprotective effects. In particular, recent literature has been oriented towards functional food ingredients which are used in the preparation of several sweets; for example, minor bioactive components of olive oil (i.e., polyphenols) or the resveratrol in grapes and other fruits are dietary components for which there are indications as regards their cardioprotective role, through the improvement of hyperlipidaemia and diabetes even beyond pharmaceutical regimens . Similarly, it has also been supported that dark chocolate consumption may have cardioprotective effects; however, necessary cautions including the high energy density and the possibility of overconsumption cannot be disregarded . Of major importance for public health is the consumption of foods which contain hidden added sugar, which is used as a food-processing method, which results in increased energy intake. Furthermore, according to the Mediterranean dietary pattern sweets are placed in the top vertex of the pyramid, suggesting the occasional consumption or even the avoidance of unhealthy fat rich food (sweet products). Apart from the cardio-protective effect of the Mediterranean diet, it is also suggested that it promotes long-term weight control and consequently prevents weight gain .
Dietary habits have long been associated with people socioeconomic status. Determinants of social status, like education level and financial status, may define the way people eat and behave as regards food choices and may have a detrimental effect in the prognosis of chronic diseases [18,34]. Moreover, financial status is a dominant characteristic of man quality of life; and diet, as a part of quality of life may be affected by peoples’ financial status. In the present work, the potential moderating effect of the socio-economic component was also studied (i.e., years of school and annual income). The tested research hypothesis was evident among ACS patients with moderate/high educational status, whereas among the other socioeconomic sub-group, although in the same harmful direction, failed to reach significance. A potential explanation could be that less educated individuals, who usually live in rural areas and villages, they mostly consume home-prepared sweets, having as basic ingredients honey, olive oil or fruits. Indeed, the majority of patients in the present study, with low education status were living in rural areas (81%). On the contrary, patients with higher educational status, mainly live in urban areas and tend to consume “industrialized” sweet products, full of added sugars and fats [35,36]. However, a critical point, with public health implications, is the inadequate consumer’s knowledge around the food labels, regarding nutrient content of foods, or the misunderstanding of the nutritional guidelines . However, taste and price are major factors influencing purchase decisions and consequently affecting food choices.
Aging and the co-morbidities observed among patients could be, also, regarded as mediating factors in the tested relationship, accounting for social status, since the least educated patients were almost one decade older, physically inactive and more likely to have abnormal blood pressure, glucose levels than those in the higher education group (Table 3) [38,39]. Regarding financial status, no significant results were found, probably, due to the fact that patients’ income may not play a role in the choice of sweets selection to consume. As regards the other factors, smoking habit at baseline examination seems to be protectively associated with the outcome, however, when pack-years were taken into account an oppose association was observed (data not shown in the presented models, since baseline factors were preferred to be presented); this may has the following explanation, ACS patients who reported current smokers at entry examination, quitted smoking during the 10-year follow-up period and therefore, the true observed effect is quitting smoking and not current, on the 10-year ACS incidence.
Despite the strengths of the present work mentioned above, there are some limitations that should be reported for better generalization of the findings. No lifestyle or other sociodemographic information from patients who died during the first days of baseline hospitalization was available; however, the in-hospital mortality rate was low (0.5% to 2% from region to region). Although this study has managed to evaluate the influence of the long term sweets consumption on ACS severity and prognosis, it relied on hypotheses concerning the possible mechanisms through which these effects are exerted, according to major characteristics of this food group, such as the sucrose concentration having bad influence on the lipid profile or inflammation markers. No objective biochemical markers concerning coagulation, oxidative stress or inflammation status were available. Even though sweets consumption was recorded through the FFQ, which may underestimate or overestimate the absolute amounts of dietary carbohydrates intake, it still remains the most valid tool for dietary habits assessment in observational studies. Moreover, compliance changes over time may also cause recall bias in patients’ reporting. Finally, the studied food group consists of an extensive variety of ingredient-combinations resulting in an extra bias as it seems rather difficult to estimate the characteristics of every kind of sweet, in detail, so as to include them in the study with a more extensive and analytical way. Nevertheless, the widely consumed, by the Greek population, sweets have undoubtedly certain common properties, which have been examined and reported in this work.
Increased sweets consumption was associated with high risk of recurrent ACS events, particularly among higher educated cardiac patients. As part of a healthy diet, the kind of sweets, as well as the way of preparation may affect the disease prognosis. Regardless patients’ socioeconomic status, lower consumption of sweets should be one of the main nutritional targets for eliminating adverse disease outcomes.
None to declare
The authors would like to present and thank the field investigators of the “GREECS” study: Yannis Antonoulas, Athanasios Karanasios, Lambros Rizos, Michalis Mparmparoussis, George Kassimatis, George Giannopoulos, Sophia Arapi, Theodoros Gialernios, Constandina Massoura, Skevos Sideris, Nick Daskalopoulos, Eleni Tsompanaki, Moscho Michalopoulou, Eleni Kalli, Margarita Verdi, Nikos Vasileiou, Evangelos Papataxiarchis, Despoina Tzanoglou, Georgia Kouli, Semina Kouroupi, Marifei Balli, Ifigenia Stergiouli, Vivi Tsomboli, as well as Prof Christodoulos Stefanadis, for their support in the clinical evaluation of the patients and Alexandros Chalamandaris for the database management.