Unhealthy Food Choices among Healthcare Shift Workers: A Cross-Sectional Study

Shift healthcare workers are a group particularly exposed to an increased risk of poor eating habits and are affected by many diseases. The aim of the study was to evaluate the dietary patterns (DPs), including the Polish-adapted Mediterranean Diet (Polish-aMED®) score, and dietary fat intake in association with the shift work of healthcare workers. This cross-sectional study involved 445 healthcare workers from the West Pomeranian in Poland. Dietary data were collected using an FFQ-6®. A posteriori DPs were derived with a Principal Component Analysis (PCA). The Polish-aMED® score and the individual’s percentage of energy from dietary fat (Pfat) were calculated. Healthcare shift work compared to the daily work was associated with approximately 2-times higher odds of adherence to the ‘Meat/fats/alcohol/fish’ DP in the upper tertile (OR: 2.38; 95% Cl: 1.27–4.47; p < 0.01) and higher Pfat >35% of total energy intake (OR: 1.73; 95% Cl: 1.06–2.83; p < 0.05). Healthcare shift work compared to the daily work was associated with approximately 50% lower odds of adherence to the ‘Pro-healthy’ DP in the middle tertile (OR: 0.48; 95% Cl: 0.26–0.89; p < 0.05) and a higher level of the Polish-aMED® score (OR: 0.57; 95% Cl: 0.33–0.98; p < 0.05), as well as lower odds of the constants of mealtime (OR: 0.54; 95% Cl: 0.33–0.89; p < 0.05). The obtained findings highlight the unhealthy food choices among shift healthcare workers. Thus, to avoid the negative health consequences, there is a need for nutritional education for healthcare workers, especially those working shifts.


Introduction
The definition of the International Labour Organization describes shift work as an organization of time in which persons succeed one another in the same position. Shift work is characterized by varying shift lengths, which can be anywhere from 12 to 24 h. Notably, in 2020, shift work was performed by approximately 15-20% of employees in Europe, 20% in the United States of America, 6-32% in Asia, as well as 8% in Poland [1]. The healthcare sector is one of the professional groups that perform shift work. The medical professions are represented by paramedics, nurses, and doctors. Paramedics are employed full-time in one, two, or even three facilities, such as in emergency and private company transporting patients. They work on a contract of mandate, a contract for specific work, or a contract or a second employment contract for 1/4, 1 2 , or even more time [2,3].

Ethical Approval
The study was approved by the Bioethics Committee of the Pomeranian Medical University (consent no. KB-0012/144/18). All procedures were carried out according to the Helsinki Declaration, and the study protocol has been approved by the Bioethics Committee. All test results were confidential.

Study Design and Sample Characteristics
The current study was started in November 2019 and completed in April 2021. Initially, it was planned to include only doctors and nurses working in shifts (12 h/12 h or 24 h/24 h) in hospitals in Szczecin (West Pomeranian Voivodeship, Poland). Due to the SARS-CoV2 virus pandemic and its related restriction about entering the wards, data was collected online, and the study group was expanded to include all healthcare workers (i.e., doctors, nurses, paramedics) working in various shift systems (day, 12 h shift/12 h, 24 h shift/24 h, night shift). In total, this study included 445 healthcare workers-doctors and nurses working in 3 hospital departments: gynecology and obstetrics, cardiology and psychiatry, including 193 shift workers and 252 day care workers.

Dietary Patterns Identification
Dietary data were obtained using a validated and self-administered version of the 62-items Food Frequency Questionnaire (FFQ-6 ® ) [19]. The wide scope of application of the FFQ-6 ® questionnaire has been confirmed by its use, among others, in a pilot, controlled, and randomized study in children with celiac disease on a gluten-free diet [27], also in patients with non-alcoholic fatty liver disease [21] and among hemodialysis patients [28].
Data about the frequency of consumption of 62 food groups at least 12 months prior to participation in the study (categories) were obtained. The frequency consumption was expressed as times/day after assigning the values for categories of frequency consumption as follows: 'never or almost never' = 0; 'once a month or less' = 0.025; 'several times a month' = 0.1; 'several times a week' = 0.571; 'daily' = 1; 'several times a day' = 2. The consumption frequency of some food groups was summed up to form 23 food groups. The full description of 23 food groups aggregated was shown in Supplementary Table S1. Next, data on the frequency of consumption of 23 food groups expressed in times/day (input variables) was standardized and included in the Principal Component Analysis (PCA) with varimax rotation to identify dietary patterns (DPs). The main criteria for the PCA-derived DPs identification were the plot of eigenvalues and the total variance explained. DPs were labelled according to the main components with the value of factor loadings > |0.30| [22]. For further analysis, tertile intervals of each DP were calculated. The characteristics of the DPs were given in the Results section.

Polish-Adapted Mediterranean Diet Score
The Polish-adapted Mediterranean Diet (Polish-aMED ® ) score is a Polish version of the Mediterranean diet score developed by Krusinska et al. [21]. The Polish-aMED ® score was calculated based on the qualitative data of the frequency of consumption (times/day) of eight food items: vegetables, fruit, whole grains, fish, legumes, nuts and seeds, processed meat, and the ratio of vegetable oils to animal fat. All dietary data were obtained using the FFQ-6 ® . The difference with the original version of the Polish-aMED ® Score was that the group "processed meat" was included instead of the combined group "red and processed meat" due to the lack of data on red meat consumption [21]. For each of the subjects, 1 or 0 points were given for individual food items depending on the frequency of consumption. The cut-off points were medians of the frequency of consumption for the initial control sample of the case-control study conducted by Krusinska et al. [21]. The Polish-aMED ® score was calculated based on the sum of points assigned to the individual items. Details of the Polish-aMED ® score calculation are shown in Supplementary Table S2. The adherence to the Polish-aMED ® score was expressed in a range from 0 to 8 points and was considered at two levels: lower (0-4 points), and higher (5-8 points).

Dietary Fat Intake Assessment
The Quick Food Scan of the National Cancer Institute and the Percentage Energy from Fat Screener scoring procedures were used to estimate an individual's percentage energy from dietary fat (Pfat) [23,24]. The participants were asked about the frequency of consumption of food items during last 12 months. The frequency of consumption was calculated (8 categories to select) into the number of times consumed per day as follows: 'never' = 0; 'less than once a month' = 0.018; '1-3 times per month' = 0.066; '1-2 times per week' = 0.214; '3-4 times per week' = 0.499; '5-6 times per week' = 0.784; '1 time per day' = 1; '2 or more times per day' = 2. The median age-and gender-specific portion sizes for each food item were multiplied by the frequency of consumption expressed in times/day. Total fat (Tfat) was calculated as the sum of the frequency of consumption of the margarine, butter, or oil added to foods (on bread, rolls, pancakes; vegetables including potatoes; and rice or pasta). Next, the regular fat (Rfat) was calculated as a product of the Tfat and the values assigned to the frequency of reduced-fat margarine use, as follows: 'did not use or almost never' = Tfat *1; 'about 1/4 of the time' = Tfat *0.75; 'about 1/2 of the time' = Tfat *0.50; 'about 3/4 of the time' = Tfat *0.25; 'almost always or always' = Tfat *0. The individual's Pfat was estimated by applying gender-specific regression coefficients to each of the 12 food items (b1-b12) and Rfat (b13) in the following algorithm [23,24]: Percentage of energy from dietary fat = intercept + (b1*cereal) + (b2*skim milk) + (b3*eggs) + (b4*bacon) + (b5*citrus juice) + (b6*fruit) + (b7*hot dogs) + (b8*cheese) + (b9*fries) + (b10*mayonnaise) + (b11*salad dressings) + (b12*rice) + (b13*regular fat) (1) For the Percentage of Energy from Fat Screener scoring procedures, the US Department of Agriculture 1994-1996 was the source of data for portion sizes and regression coefficients (Continuing Survey of Food Intakes by Individuals CSFII) [24,25].

Confounders
The socioeconomic status (SES) was calculated on the basis of three components: place of residence, declared economic situation of the household and the level of education (Table 1). For each respondent, the sum of the numerical values assigned to individual categories of SES components was calculated. In the abovementioned way, the SES index was calculated (which could be in a range from 3 to 13 points). Depending on the value of SES index, there were 3 selected categories SES: low (3-6 points), average (7-10 points), and high (11-13 points). Situation of household (self-declared) We live very poorly-we do not have enough resources even for basic needs (food/clothing/housing fees) 1 We live poorly-we only have enough resources for the cheapest food and clothing, but not for housing fees 2 We live poorly-after paying the housing fees, we only have enough resources for the cheapest food and clothing 3 We live very thriftily-we only have enough resources for basic needs (food/clothing/housing fees) 4 We live thriftily-we have enough resources for everything 5 We live well-we have enough resources for everything, without any special restrictions 6 Scoring-values assigned to the response categories.
The Godin-Shephard Leisure-Time Physical Activity Questionnaire (GLTEQ) was used for the physical activity assessment [26]. Respondents were asked how many times on average they exercise for more than 15 min during their free time in a typical 7-day period (week). There were three categories of exercise: strenuous (heart beats rapidly, e.g., running, jogging, football, vigorous swimming, vigorous long-distance bicycling), moderate (not exhausting, e.g., fast walking, tennis, easy bicycling, easy swimming), and mild activities (minimal effort, e.g., yoga, easy walking, bowling, fishing from a riverbank) (Godin et al. 2011). Then, the total weekly leisure activity score was calculated as a sum of the products of strenuous, moderate, and mild activities expressed in the number of times/week, and the metabolic equivalent of task (MET) values corresponding to these categories: nine, five, and three, respectively, as shown in the following formula: Weekly leisure-time activity score = (9*Strenuous) + (5*Moderate) + (3*Mild) (2) In summary, in reference to the score in units obtained using only moderate and strenuous physical activities, respondents were categorized as: active (24 units or more), moderately active (14-23 units), and insufficiently active (less than 14 units) [26].
Besides socioeconomic status index (SESI) and physical activity (GLTEQ summary in points), the present analyses were taking into account a set of many potential confounders: age (years), gender (men, women), BMI (kg/m 2 ), current smoking status (non-smoker, smoker), alcohol drinking within the last 12 months (no, yes), chronic diseases (no, yes), taking medication > 1 year (no, yes), and vitamin/mineral supplements use within the last 12 months (no, yes).

Statistical Analysis
All continuous data, including food frequency consumption (times/day), Polish-aMED ® score (points), total and regular fat intake (g), and the percentage energy from dietary fat (%) were expressed in means and standard deviations (SD). For categorical data, involving tertiles intervals or levels of DPs or Pfat, the percentage distributions were calculated. Differences in baseline sample characteristics were verified with a Kruskal-Wallis test (continuous data) or Pearson Chi2 test (categorical data). The Pearson's correlation coefficients were calculated for 23 food items in the Polish-aMED ® score (Armitage et al. 2001). The percentage distribution of healthcare workers was compared in tertiles or levels of DPs using the Pearson Chi2 test with Yates' correction as necessary. The logistic regression analysis was used to assess the associations of the adherence to the dietary patterns, percentage of energy from dietary fat, and constants of mealtime by mode of work among healthcare workers. The odds ratios (ORs) and 95% confidence intervals (95% CI) were calculated. The references (OR = 1.00) were the healthcare workers working daily, the bottom tertile or lowest level of each DP, and no constants of mealtime. Two models were created: an unadjusted model, and a model adjusted for the potential confounders mentioned above. The level of significance of the odds ratio was verified with Wald's test (Armitage et al. 2001). The statistical analysis was performed using STATISTICA software (version 13.0 PL; StatSoft Inc., Tulsa, OK, USA; StatSoft, Krakow, Poland). A p-value < 0.05 was considered statistically significant.

Baseline Sample Characteristics
This study included 445 healthcare workers, including 252 daytime workers and 193 shift workers 18-74 years old. Most of the respondents came from large cities with more than 200,000 inhabitants (52.1%) and had higher education (92.8%). Shift workers compared to day workers were younger (34.0 vs. 36.4 years; p = 0.0188), characterized by a lower socio-economic status index (11.2 vs. 11.6 points, p = 0.0045), and they significantly less frequently declared taking medications for more than one year (25.9 vs. 35.7%; p = 0.0272) and the use of vitamin and mineral supplements (65.3 vs. 75.0%; p = 0.0255). There were no significant associations in place of residence, educational level, BMI, some of health and lifestyle data such as physical activity, alcohol drinking, smoking status, or occurrence of chronic diseases depending on lifestyle of healthcare workers. The characteristics of the participants are presented in Table 2.

Dietary Patterns
Three totally different PCA-derived DPs were identified. The total share in explaining the variance was 38.5%. The 'Lacto-ovo-vegetarian' DP was characterized by a lesser frequency of consumption of snacks, sugar and sweetened foods, refined grains, and breakfast cereals, and a higher frequency of consumption of vegetables, fruits, whole grains, legumes, nuts, and seeds than the 'Sweet-salty-snack-dairy' DP (Table 3 and  Supplementary Tables S3 and S4). The 'Sweet-salty-snack-dairy' DP was the opposite of the 'Lacto-ovo-vegetarian' pattern, i.e., it involves more frequent consumption of sweetened milk drinks/flavored cheese, salty snacks, sugar/honey/sweets, refined grains, breakfast cereals, and sweetened beverages, and much less frequent consumption of vegetables, fruits, whole grains, legumes, nuts, and seeds (Tables S3 and S4). The pattern labeled 'Meatfats-alcohol-fish' exceeds the other two patterns not only for fats but also for processed meats, white meat, alcohol, and fish (Tables S3 and S4).  The a priori Polish-aMED ® score was positively correlated with the frequency of consumption of vegetables, fruits, whole grains, legumes, nuts and seeds, fish, and vegetable oils including olive oil, and negatively correlated with the frequency of consumption of processed meats ( Table 3). The frequency of consumption of food groups and the sample characteristics by tertiles of DPs are shown in Supplementary Tables S3 and S4, respectively.

Food Choices and Fat Intake: Associations with Shift Work
Compared to the daily healthcare workers, significantly more shift workers were in the upper tertile of the 'Meat-fats-alcohol-fish' DP (42.0 vs. 27.0%), as well as having a higher total fat intake (4.1 vs. 2.9 g), and dietary fat intake of total energy intake >35% (50.3 vs. 36.1%; Table 4). On the other hand, compared to the daily workers, the number of shift workers was lower in the higher level of the Polish-aMED ® score (26.4% vs. 38.5%) and lower in the middle tertile of the 'Lacto-ovo-vegetarian' DP (26.4% vs. 38.9%). Shift workers compared to daily workers significantly less frequently declared using a special diet or restrictions in food consumption (34.2 vs. 45.8%), including dairy (21.8 vs. 34.7), fruits (5.7 vs. 12.7%), potatoes and cereals (21.8 vs. 36.1%), or meat and meat products (33.2 vs. 50.4%), and less frequently declared constants of mealtime (44.0 vs. 62.7%). There were no significant differences in the number of shifts and daily workers within the tertiles of the 'Sweet-saltysnack-dairy' DP ( Table 4). The mean consumption of food groups by mode of work among healthcare workers is shown in Supplementary Table S5. Associations between dietary patterns and energy from dietary fat are presented in Supplementary Table S6.

Discussion
According to the authors' best knowledge, this is the first study describing the association of dietary patterns, fat intake, and selected dietary habits including constants of mealtime with the mode of work among healthcare workers. The obtained results highlight that the Polish healthcare workers working in shifts did not have regular mealtime consumption, and made unhealthy food choices regarding the frequent consumption of processed meats, fats, and alcohol, which contributed to a high dietary fat intake of >35% of the total energy intake. In regards to the pro-healthy patterns, there was an inverse association of the shifts mode of work with the Polish-aMED ® score and the 'Lacto-ovo-vegetarian' pattern based on the plant food, fish, and non-sweetened dairy.

Mealtime among Healthcare Workers
Healthcare professionals are often obligated as a professional group to provide a model of healthy eating habits. In this study, it was observed that shift workers had approximately 50% less chance of consuming meals regularly compared to workers who worked in the daytime system. The study by Filipczuk et al. [29] showed that all surveyed nurses clearly confirmed that their behavior and eating attitudes may influence the shaping of nutritional and health attitudes of their charges [29]. In another study conducted by Sokołowska et al., it was observed that nurses declared the knowledge about the roles of healthy eating, but did not follow them by eating irregularly [30]. Additionally, half of the surveyed nurses emphasized that the shift work system hinders

Discussion
According to the authors' best knowledge, this is the first study describing the association of dietary patterns, fat intake, and selected dietary habits including constants of mealtime with the mode of work among healthcare workers. The obtained results highlight that the Polish healthcare workers working in shifts did not have regular mealtime consumption, and made unhealthy food choices regarding the frequent consumption of processed meats, fats, and alcohol, which contributed to a high dietary fat intake of >35% of the total energy intake. In regards to the pro-healthy patterns, there was an inverse association of the shifts mode of work with the Polish-aMED ® score and the 'Lacto-ovo-vegetarian' pattern based on the plant food, fish, and non-sweetened dairy.

Mealtime among Healthcare Workers
Healthcare professionals are often obligated as a professional group to provide a model of healthy eating habits. In this study, it was observed that shift workers had approximately 50% less chance of consuming meals regularly compared to workers who worked in the daytime system. The study by Filipczuk et al. [29] showed that all surveyed nurses clearly confirmed that their behavior and eating attitudes may influence the shaping of nutritional and health attitudes of their charges [29]. In another study conducted by Sokołowska et al., it was observed that nurses declared the knowledge about the roles of healthy eating, but did not follow them by eating irregularly [30]. Additionally, half of the surveyed nurses emphasized that the shift work system hinders the regular nutrition of this professional group. It occurs more and more often that shift workers use catering and the box diet. A diverse, absorbing schedule often discourages systematic shopping and daily preparation of wholesome meals for the next days [30].
According to dietary recommendations, adults should consume 4-5 meals per day (polish recommendation [31][32][33]). The role of regular consumption of a smaller amount of food protects against hypoglycemia and consequently, hyperinsulinemia, which occurs after eating a heavy digestion meal eaten once during work [34]. In the conducted study, the consumption of at least 4 meals a day was declared by over 60% of healthcare workers, and the number of meals was not statistically significantly differentiated depending on the work mode of the respondents. Similarly, in the study by Sińska et al. [35] it was shown that, regardless of the nature of their work, 50% of the surveyed nurses eat 4 to 5 meals per day. In the current study, 8% of respondents declared eating 1-2 meals per day. In Bielak's study, it was observed that 30% of nurses eat only 1-2 meals a day, whereas 25% of the nurses left home without eating their breakfast [36]. This model of nutrition may lead to a decrease in blood glucose levels, dizziness, increased fatigue, impaired concentration, decreased strength, endurance, and muscle regeneration. The consumption of two meals that are large in volume lengthens the digestive process. Long breaks between meals also lead to excessive secretion of ghrelin, which can consequently lead to overconsumption.

Unhealthy Food Choices and Fat Intake among Shift Healthcare Workers
In the current study it was noted that shift health workers consumed more processed meat and higher amounts of fat, especially of animal origin, compared to day care workers. This model of nutrition, combined with alcoholic beverages and fish, often in the form of canned fish, referred to as 'Meat-fats-alcohol-fish', was associated with a higher proportion of fat in the diet (over 35%). Consumption of saturated fatty acids (SFAs) is related to the development of cardiovascular disease [37]. SFAs increase the concentration of LDL, reduce the concentration of HDL, and increase the ratio of total cholesterol to HDL [38,39]. In 2015, the World Health Organization (WHO) classified the consumption of processed red meat-referring to compounds present in meat preserved by smoking, curing, salting, or the addition of chemical preservatives, including those contained in processed foods-as potentially carcinogenic [40,41]. Increased consumption of red meat is associated with premature death from various causes, increasing the risk of cancer, cardiovascular disease, and stroke [42][43][44][45]. There is also an increased risk of type 2 diabetes [45]. A high proportion of fat in the diet promotes weight gain. Both overweight and obesity in adulthood are associated with an increased risk of many types of cancers, coronary artery disease, and stroke, and were among the major risk factors influencing disability-adjusted life years in 2015 [46].

Pro-Healthy Dietary Patterns among Shift Healthcare Workers
The Mediterranean diet has been recognized as one of the healthiest models of nutrition. The Mediterranean diet is rich in fresh vegetables, especially greens and fruits, whole grains, pods, olive oil, lean fish and seafood, lean dairy products, herbs and spices, and small amounts of red wine [47]. In the current study, healthcare workers who worked shifts, compared to day workers, had about 50% less chance of achieving a higher level of the Polish-aMED ® score of the research. The Mediterranean diet is especially recommended due to its health benefits. The international project Healthy Aging: a Longitudinal study in Europe (HALE), confirmed the relationship between higher adherence to the Mediterranean diet and a reduced number of cardiovascular events [48]. Other studies confirmed that following the Mediterranean diet was associated with a reduced incidence of overweight and obesity, including abdominal obesity. According to a recent systematic review and meta-analysis, greater adherence to the Mediterranean diet appears to be inversely related to multiple cancer risk and overall cancer mortality [39]. The Mediterranean diet provides not only essentials vitamins and minerals but also large amounts of dietary fiber and polyphenols, which are included in fruits and vegetables, red wine, and olive oil. Phenolic compounds may act as modulators of gene expression and intracellular signaling cascades involved in cell function and protection [40,41].
In the current study, shift workers compared to day workers had about 50% less chance of adherence to the 'Lacto-ovo-vegetarian' pattern in the middle tertile. As in the case of the Mediterranean diet, unfortunately there is no similar research in this area, which makes comparative analyzes impossible. This plant-based pattern is associated with an increased consumption of dietary fiber, which has a number of health-promoting effects. Observational data show a reduction in morbidity and mortality from cardiovascular disease, type 2 diabetes and colorectal cancer in people who follow a diet rich in dietary fiber. In addition, the 'Lacto-ovo-vegetarian' pattern was also positively loaded by fish and non-sweetened dairy, which due to the content of beneficial polyunsaturated fatty acids and biopeptides are involved in the regulation of body weight, blood pressure and blood lipid profile [49].

Strengths and Limitations
This study included 445 participants who represented the healthcare environment. Additionally, it is the first study conducted among healthcare workers that divided them depending on the work system (day/shift). Thanks to this division, we were able to assess the differences in the consumption of selected groups of products and on their basis determine the nutrition patterns and calculate the Polish-aMED index, using two approaches-a priori and a posteriori. It is the first study in which holistic eating patterns, fat consumption, and selected eating habits were assessed in a group of healthcare professionals based on their work system. Three validated questionnaires were used in the study: (i) the 62-item Food Frequency Questionnaire FFQ-6 ® , (ii) the Quick Food Scan of the National Cancer Institute and the Percentage Energy from Fat Screener, and (iii) Godin Leisure Time Exercise Questionnaire, which allowed adequately assessing the diet, dietary fat intake, and level of physical activity of the participants. Notably, the questionnaires, which were used in this study, are widely used in studies assessing nutrition and lifestyle in various age groups, both for healthy people as well as patients with non-alcoholic fatty liver or cancer patients.
A limitation in the studies was the lack of a quantitative assessment of the diet using more precise methods, such as a 24-h interview or a diary of the current consumption. Thus, we cannot precisely assess the amount of calories consumed as well as the macroand micronutrients depending on the type of work. The FFQ questionnaire provides the consumption over the last year, which is not precise. However, it should be emphasized that this is a questionnaire that allows for a retrospective assessment of food consumption over a longer period of time. Moreover, it is a relatively easy method, less labor-intensive for the respondent (it can be used online), and cheap. To increase the precision of the obtained results, the AICR quick fat intake questionnaire (variant FFQ) was used, which confirmed the results obtained with the FFQ-6 questionnaire (high coverage of the 'Meat-fats-alcoholfish' pattern among shift system workers with a high fat consumption >35 %). Another limiting factor for the study was the fact that the data was not collected personally, but online. It was caused by the SARS-CoV2 virus pandemic and strongly limited/blocked the access to the wards. However, all participants were carefully instructed on how to complete the questionnaires correctly. Another limitation of the study was the representativeness of the group-it was a study involving healthcare workers only from Poland, but this group was sufficient to achieve the assumed the aim of study.

Conclusions
The obtained findings suggest the negative influence of shift work on regular mealtime consumption, and pro-healthy dietary patterns, including the Mediterranean diet, based on plant food, fish, and non-sweetened dairy among healthcare workers. Healthcare workers working in shifts more frequently consumed processed meats, animal fats, and alcohol, which contributed to an increased energy intake from dietary fat above 35% of the total energy intake. These unhealthy food choices could result from cheaper and more processed food purchases due to the thrifty living of shift healthcare workers.
Unhealthy dietary habits observed in shift healthcare workers can lead to negative health consequences and reduce the effectiveness of the work, which is associated with high responsibility. These study points indicate the need for nutritional education for healthcare workers, especially those working shifts, of the importance of regular meal consumption during a shift, and that imbalanced meals may have a harmful influence to their functioning at work during night shifts. There is a need to recommend shift healthcare workers to prepare and take pro-healthy meals and healthy snacks to work based on vegetables, fruits, grains and nuts, legumes, vegetable oils, and un-sweetened dairy, and the elimination of the consumption of processed food, including meats, animal fats, and alcohol. As a future recommendation, there should also be efforts to remove some of the institutional limitations of the healthcare system regarding increasing the access to healthy food choices, e.g., by providing healthy "vending machines" or restaurants with cooked meals for healthcare workers, as well as, ensuring more time and space for meal breaks during work time, also in night shifts.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/nu14204327/s1, Table S1: Description of 23 food groups aggregated: data based on the FFQ-6 questionnaire; Table S2: Description of food groups for the Polish-adapted Mediterranean Diet ® score (0-8 points) calculation-data for the initial control sample (n = 412); Table S3: The mean (95%CI) of the frequency of food consumption by dietary patterns among Polish healthcare workers (times/day); Table S4: The sample characteristics by dietary patterns among Polish healthcare workers (%) or mean (SD); Table S5: The mean (SD) of the food consumption by mode of the work among Polish healthcare workers; Table S6: Fat intake, food consumption, and dietary patterns in association with the percentage of energy from dietary fat among Polish healthcare workers (% or mean (SD).