Hostname: page-component-76fb5796d-vfjqv Total loading time: 0 Render date: 2024-04-25T11:46:22.033Z Has data issue: false hasContentIssue false

The association between lunch composition and obesity in Iranian adults

Published online by Cambridge University Press:  08 July 2021

Zahra Akbarzade
Affiliation:
Department of Community Nutrition, School of Nutritional Sciences and Dietetics, Tehran University of Medical Sciences (TUMS), Tehran, Iran
Kurosh Djafarian
Affiliation:
Department of Clinical Nutrition, School of Nutritional Sciences and Dietetics, Tehran University of Medical Sciences, Tehran, Iran
Nasim Nasim Saeidifard
Affiliation:
Department of Community Nutrition, School of Nutritional Sciences and Dietetics, Tehran University of Medical Sciences (TUMS), Tehran, Iran
Shabnam Aliakbari Majd
Affiliation:
Bou Ali Sina Medical Training Center, Qazvin University of Medical Science, Qazvin, Iran
Nazila Garousi
Affiliation:
Food Security Research Center; Department of Clinical Nutrition, Isfahan University of Medical Sciences, Isfahan, Iran
Fatemeh Samadi
Affiliation:
Student Research Committee, Department of Clinical Nutrition and Dietetics, Faculty of Nutrition and Food Technology, National Nutrition and Food Technology Research Institute, Shahid Beheshti University of Medical Sciences, Tehran, Iran
Hanieh Jebraeili
Affiliation:
Department of Community Nutrition, School of Nutritional Sciences and Dietetics, Tehran University of Medical Sciences (TUMS), Tehran, Iran
Maryam Chamari
Affiliation:
Department of Community Nutrition, School of Nutritional Sciences and Dietetics, Tehran University of Medical Sciences (TUMS), Tehran, Iran
Cain C. T. Clark
Affiliation:
Centre for Sport, Exercise, and Life Sciences, Coventry University, Coventry CV15FB, UK
Sakineh Shab-Bidar*
Affiliation:
Department of Community Nutrition, School of Nutritional Sciences and Dietetics, Tehran University of Medical Sciences (TUMS), Tehran, Iran
*
*Corresponding author: Sakineh Shab-Bidar, email s_shabbidar@tums.ac.ir
Rights & Permissions [Opens in a new window]

Abstract

We aimed to assess the dietary composition of lunch meal using a posteriori-derived dietary patterns and to determine the association of lunch composition with obesity in a sample of Iranian adults. This cross-sectional study was conducted on 850 men and women in Tehran (aged 20–59 years). Dietary intakes were assessed using three 24-h dietary recalls, and dietary patterns were identified via principal component factor analysis. For each identified pattern, scores were calculated for each participant and then classified into tertiles. Central obesity was defined WHO criteria. General obesity was defined as a BMI of more than 30 kg/m2. Three major dietary patterns were identified at lunch meal using twelve food groups: ‘Bread, grains and fat’, ‘Western’ and ‘Potato and eggs’. After adjustment for potential confounders, participants at the top tertile of the ‘Bread, grains and fat’ dietary pattern had greater odds for a higher waist-to-hip ratio (WHR), compared with those in the lowest tertile (OR: 1·44, 95 % CI 1·01, 2·07). However, we found no association between ‘Western’ or ‘potato and eggs’ patterns and WHR (OR: 0·89, 95 % CI 0·62, 1·28 and OR: 1·16, 95 % CI 0·69, 1·42, respectively). None of the identified dietary patterns was associated when defining obesity with waist circumference or BMI. In conclusion, participants had a greater chance of central obesity defined based on WHR following a lunchtime pattern with a higher and positive loading factor for ‘Bread, grains and fat’.

Type
Research Article
Copyright
© The Author(s), 2021. Published by Cambridge University Press on behalf of The Nutrition Society

Obesity represents one of the most important metabolic diseases worldwide. In recent decades, the prevalence of overweight and obesity has doubled, globally, while it has been estimated that by the year 2030, the number of obese people will reach 1·21 billion(Reference Antelo, Magdalena and Reboredo1). In addition, in Iran, almost 21·7 % of adults are affected by obesity(Reference Rahmani, Sayehmiri and Asadollahi2). BMI is one of the most common and simple methods used in many epidemiological studies, which has been proposed as a tool for screening and early clinical evaluation of obesity(Reference Cornier, Despres and Davis3). However, BMI has some limitations since it is an indicator of total body fat and does not provide information about localised abdominal fat that is associated with metabolic disease. Some studies showed that the pattern of fat distribution in the body plays a key role in identifying chronic disease(Reference Esmaillzadeh, Mirmiran and Azizi4) and people with fat accumulation in the abdomen are at a higher risk of diabetes, hypertension and CVD(Reference Azizi, Esmaillzadeh and Mirmiran5). In this regard, other anthropometric indicators such as waist circumference (WC) and waist-to-hip ratio (WHR) are superior to BMI for revealing obesity and overweight. WC and WHR are mostly used to express how fat is distributed in the body, and WHR is widely applied to distinguish between the central and peripheral distribution of adipose tissue(Reference Lee, Huxley and Wildman6,Reference Taylor, Brooking and Williams7) .

Socio-demographic determinants of obesity have been of great research interest recently. It is demonstrated that certain health behaviours including smoking and physical activity are associated with obesity(Reference Gruber and Frakes8,Reference Fogelholm and Kukkonen-Harjula9) , and it has been reported differences in obesity prevalence across various sex, age and socio-economic groups(Reference Paeratakul, Lovejoy and Ryan10). Socio-economic status is usually measured by education, occupation, employment, income and wealth. Overall, higher welfare level and higher and higher socio-economic status were observed to be associated with a lower BMI, WC and WHR(Reference McCormack, Friedenreich and McLaren11). This may be due to a healthier diet in subjects with a higher level of education, income and socio-economic status(Reference Ruf, Nagel and Altenburg12), as well as a higher degree of physical activity(Reference McCormack, Friedenreich and McLaren11).

Diet is the fundamental component of a healthy lifestyle which can play a significant role in the prevention of non-communicable diseases(13). Previous studies have tended to focus on individual foods or nutrients; however, given the complexity of human diets, several authors have proposed the analysis of overall dietary patterns(Reference Hu, Rimm and Smith-Warner14,Reference Hu, Rimm and Stampfer15) . Moreover, the relationship between dietary patterns, considering the complexity of diets and the potential interaction between food components, and obesity has been well documented in the literature(Reference Cunha, de Almeida and Sichieri16Reference Shu, Zheng and Zhang18). Findings, such as the association between the Mediterranean diet and low rates of chronic diseases, or the successful treatment of hypertension through changes in dietary patterns, have suggested the investigation of dietary intake patterns(Reference Helsing19,Reference Sacks, Svetkey and Vollmer20) . Apart from well-founded evidence regarding the benefits of vegetables, fruits, fibre, nuts and fish, and the value of reducing or eliminating snacks, people regularly consume combinations of foods in three or more meals per day(13,Reference Hu21) . The main meals are described as foods that are typically consumed in the largest volume(Reference Leech, Worsley and Timperio22,Reference Bellisle, Dalix and Mennen23) . It is important to acknowledge that different nutritional compositions in main meals may have an impact on diet quality and influence diet–disease relationships. There is an association between meal patterns and energy balance and weight status(Reference Mesas, Muñoz-Pareja and López-García24,Reference Szajewska and Ruszczyński25) . Results of a study in Brazilian population showed that the traditional Brazilian lunch pattern is inversely associated with obesity in insufficiently active individuals(Reference de Oliveira Santos, dos Santos Vieira and Miranda26). In another study, breakfasts containing > 25 % of total energy intake and lunches containing > 35 % of total energy intake were associated with an increased likelihood of central obesity(Reference Aparicio, Rodríguez-Rodríguez and Aranceta-Bartrina27). Indeed, previous studies have shown that the lunchtime meal represents the highest proportion of protein, fat and carbohydrate intake(Reference Almoosawi, Winter and Prynne28) and supplies about 30% of the daily energy intake(Reference Benelam and Stanner29). Therefore, this study aimed to, first, identify major dietary patterns at lunch, and, second, to determine their association with obesity among adults living in Tehran.

Subjects and methods

Participants

This cross-sectional study was conducted using a sample of 850 men and women, aged 20–59 years, who have visited the health centre and were been selected from the five regions of Tehran from 2018 to 2019. The following formula was used for sample size calculation: n (pqz2)/E2. Considering the total prevalence of 65 % for overweight and obesity(Reference Ahmad Kiadaliri, Jafari and Vaez Mahdavi30), an error coefficient of d = 0·04 and at a level of 0·05, the sample size of 546 people was obtained. With a design effect of 1·5 and to compensate for the potential exclusion of participants due to under- and over-reporting of total energy intake, or attrition due to other reasons, the final sample size of 850 participants was selected for inclusion. A two-stage cluster sampling was used to recruit participants from health care centres. First, we classified health centres into five districts of the city including North, South, East, West and centre. Next, a list of all health centres that existed in each district was provided. Then, twenty-five health centres (due to budget and time limits) were divided according to the number of health centres in each area. After a randomised selection of health care centres from enlisted health centres in each region, the number of eligible health centres was randomly selected. Subsequently, the total sample (850) was divided by the number of health centres (25), yielding the required number of samples in each health centre. Following this, we entered each of the health centres, and those who willing to participate in our study and those who were members of the health centre and living in Tehran were included and conducted random sampling to ascertain the required number of samples needed and on the other hand, at the end of the interview with the individuals, nutritional information appropriate to each person’s conditions was provided to them. Subjects were considered eligible for inclusion if the following criteria were met: (a) participants within the age range of 20–59 years; (b) apparently healthy individuals who did not report any previous diagnosis of chronic diseases such as diabetes, CVD and chronic kidney, lung and liver diseases by a physician; (c) be willing to take part in the study; (d) being a resident of Tehran and (e) being a member of the health centre. Participants were excluded from the analysis if (a) their daily energy intake was implausibly low or high (<3347·2 kJ/d or >17572·8 kJ/d); and (b) those who did not report any adherence to certain dietary patterns, any special diet or diet therapy such as vegetarian diet. The sample collection was facilitated by the coordination of the Health Bureau of the Municipality of Tehran and the co-operation of the health centres of Tehran.

Outcome measures

Weight and height were measured according to the standard methods(Reference Lohman, Roche and Martorell31). Weight was recorded to the nearest 100 g while minimally clothed and unshod using digital scales. Height was measured in a standing position, unshod, using a tape stadiometer to the nearest 1 mm, while the shoulders were maintained in a normal position. BMI was calculated using the following formula: weight in kg, divided by height in meters squared, expressed as kg/m2. General obesity was defined as BMI ≥ 30 kg/m2(32). WC was recorded to the nearest 0·1 cm at the umbilical level and hip circumference at the maximal point over light clothing, using a non-stretch tape meter, without putting pressure on the body surface. Central obesity was defined as WC ≥ 102 in men and WC ≥ 88 in women, respectively, and a WHR above 0·90 for males and above 0·85 for females, respectively(33). Participants rested for 15 min before blood pressure was measured. Then, a trained assessor measured blood pressure twice, with the participant in a seated position, with a standard mercury sphygmomanometer and the mean of two measurements was calculated.

Dietary assessment

Dietary intake was assessed using three 24-h dietary recalls. The first recall took place during the participants’ first visit to the health centre. The other two 24-h dietary recalls were obtained at random days, including weekends over the phone. All 24-h dietary recalls were performed by trained interviewers. Any food or beverage that the participants consumed during the denominated meal time was considered. Finally, the standard unit size and items reported based on home weighing scales were converted into g using the home scale guideline(Reference Ghaffarpour, Houshiar-Rad and Kianfar34). The data from these questionnaires were entered into a purpose-built excel spreadsheet, where the gram equivalent was obtained for each item and each individual. In addition, we used Nutritionist IV software (First Databank), modified for Iranian foods, to analyse the energy and nutrients of food items. Lunch was predefined as a large meal eaten between 12.00 and 16.00 hours(Reference Kahleova, Lloren and Mashchak35). Foods were grouped according to similar nutritional values, Iranian consumption habits, literary data and experience of the research team in previous studies(Reference Karamati, Jessri and Shariati-Bafghi36Reference Lockheart, Steffen and Rebnord38). Some individual food items that consisted of separate items (e.g. eggs) or that represented special dietary habits (such as potatoes) were retained as a single food. Moreover, it should be noted that some single FFQ items (salt and potatoes) considered as a single group because in the FFQ there were no other food items to be appropriate for combination with these foods into multiple-item food groups. Finally, we created twelve pre-determined dietary groups (bread and grains(Reference Serra-Majem and Bautista-Castaño39,Reference Liu, Willett and Manson40) , dairy products(Reference Wang, Wu and Zhang41,Reference Barba and Russo42) , poultry(Reference Donma and Donma43,Reference Petermann-Rocha, Parra-Soto and Gray44) , eggs(Reference Liu, Zhao and Li45,Reference Saande, Bries and Pritchard46) , fat(Reference Lissner and Heitmann47,Reference Gillis, Kennedy and Gillis48) , potatoes(Reference Borch, Juul-Hindsgaul and Veller49,Reference Schwingshackl, Schwedhelm and Hoffmann50) , processed meat and red meat(Reference Rouhani, Salehi-Abargouei and Surkan51,Reference Dabbagh-Moghadam, Mozaffari-Khosravi and Nasiri52) , soft drinks(Reference Gibson53,Reference James and Kerr54) , vegetables(Reference He, Hu and Colditz55), legumes and nuts(Reference Martínez-González and Bes-Rastrollo56), salt(Reference Ma, He and MacGregor57) and sauces(Reference Appleton58)) which also had an association with obesity. Also, for some foods like potato, eating habits of our population has been considered. As white potato is a good source of carbohydrate, dietary fibre and resistant starch, it is a favourite staple food in several cultures as well as a good source of vitamin C and K, especially in Iran. Moreover, due to the high content of carbohydrate, it is suspected to have a link to obesity, and most common foods with potato in Iran contain more fat energy content than carbohydrate energy content. Then, we decided to consider potato as a separate food group in our study.

Assessment of other variables

Physical activity information was obtained using participants’ oral responses to the international physical activity questionnaire and expressed as the metabolic equivalent h/week (MET-h/week)(Reference Ainsworth, Haskell and Whitt59). We asked the participants to think about all the intense and moderate activities that they engaged in during the past 7 d, considering the time spent on these activities, before completing the questionnaire. Additional covariates, including age (year), BMI (kg/m2), an education level (illiterate, under diploma (Primary School, Secondary School, High School), diploma, University degree), marital status (married or other), occupation (employee or unemployed), medical condition (healthy or underlying disease), smoking status (not smoking, quit smoking, smoker) and lifestyle (living alone, with someone), were obtained using questionnaires. Furthermore, we defined underlying diseases in this study as diabetes, hypertension, dyslipidaemia, CVD, cancer and respiratory disease. According to the self-reporting of participants in the study, if a person had one of the diseases considered in this research, he/she would receive code 1 and otherwise code 0. Finally, if the individual did not have any of the respective diseases, he/she was classified in the healthy group, and those who even had one of the diseases were assigned to the group with underlying disease.

Statistical analyses

Factor analysis (principal component analysis) was used to identify major lunch consumption patterns based on twelve pre-determined dietary groups, and two interpretable factors were retained based on the scree test(Reference Rezazadeh, Rashidkhani and Omidvar60). Then an orthogonal rotation method (varimax rotation) was applied to simplify the factor structure and present it in an interpretable manner. The number of factors retained from each dietary pattern classification method was determined by eigenvalues (>1·10), scree plots and factor interpretability. Higher loadings (≥0·2) show that the food shares more variance with that factor. The derived factors (lunch patterns) were labelled based on our interpretation of the data, as well as on prior literature. The factor score for each pattern was calculated by collecting consumed food groups, weighted by factor loadings, and each participant determined the score for each identified pattern. Pearson correlation coefficients were also computed to identify the association between food groups. We used one-way ANOVA, with Tukey post hoc comparisons where appropriate, for quantitative variables, and χ 2 tests for qualitative variables, to determine significant differences across tertiles of lunch pattern scores. The association between major dietary patterns and general and central obesity was assessed by logistic regression analysis. ANCOVA was used to adjust for covariates as follows: in the first model for age, sex, education, marriage, lifestyle and smoking, while the second model was adjusted for model 1 plus physical activity and total energy intake. Logistic regression analysis for general and central obesity, according to lunch patterns, was used to obtain the OR and 95 % CI, which were adjusted for potential confounders, including age, sex, education, marriage, lifestyle, smoking, physical activity and total energy intake. Confounders were selected based on literature review including age (years), sex (male or female), physical activity level, smoking status (never smoke or former/current smoker) and total energy intake.

All data were analysed using the statistical software package SPSS version 22, and statistical significance was accepted at P < 0·05.

Results

Of the 850 participants who enrolled in the study, sixty were excluded due to a lack of adequate information and lack of co-operation in their recall report; thus, 790 remained in the study for final analysis. In general, the average energetic intake at lunchtime was higher than other meals (Fig. 1).

Fig. 1. The distribution of energy content was consumed across the time of day. , mean energy.

Using the factor analysis method, three major dietary patterns were identified and presented in Table 1. Factor 1, named the ‘Bread, grains and fat’ dietary pattern, had high and positive factor loadings for fats, bread and grains, salt, vegetables, poultry and fish and high negative factor loadings for legumes and nuts and sauces; factor 2, named the ‘Potato and eggs’ dietary pattern, had positive factor loadings for potato and eggs, and high negative factor loadings for intake of bread and grains, nuts and legumes; and finally factor 3, named the ‘Western’ dietary pattern, showed positive factor loadings for red or processed meat, sauces, soft drinks, nuts and legumes and negative factor loadings for poultry and fish and dairy products.

Table 1. Food groups used in the factor analysis and factor loadings for each of the identified lunch patterns*

* Factor loadings of < 0·2 have been removed to simplify the table.

Table 2 provides an overview of the distribution of qualitative and quantitative variables across tertiles of major dietary patterns. No significant difference was found in the distribution of qualitative variables across the tertiles of three major dietary patterns. Adherence to the ‘Potato and eggs’ pattern was associated with an increase in age (P = 0·04) and blood pressure (P = 0·03). Mean intake of vegetables (P < 0·001), dairy products (P = 0·01), grains (P < 0·001) and meat (P = 0·006) was higher at the third tertiles of the ‘Bread, grains and fat’ dietary pattern in comparison with the first tertile. Moreover, adherence to the ‘Eggs and potato’ dietary pattern was associated with the higher intake of vegetable (P < 0·001) and grain (P < 0·001). In addition, the mean intake of dairy products, grains and meat was significantly different across the tertiles of ‘Western’ dietary pattern.

Table 2. Characteristics of the study participants by tertiles (T) of lunch pattern scores

(Numbers and percentages; mean values and standard deviations)

P-values obtained using χ 2 test.

* Diabetes, hypertension, dyslipidaemia, CVD, cancer and respiratory disease.

Heat map shows the Pearson correlation matrix of food groups at lunchtime. Correlation analysis showed that there were strong positive correlations between potato and eggs (P < 0·001), while strong negative correlations were found between nuts and legumes and fish and poultry (P < 0·001), and between meat and processed meat and fish and poultry (P < 0·001) (Fig. 2).

Fig. 2. Heat map showing the Pearson correlation matrix for lunch food intake in g (n 790) by food groups. The colour corresponds to the strength of correlations (red: positive correlation; white: no correlation; blue: negative correlation).

Multivariable-adjusted means for anthropometric measures and indexes across tertiles of dietary patterns are depicted in Table 3. The results showed that higher adherence to the ‘Bread, grains and fat’ dietary pattern was associated with a higher WHR (P = 0·04). In addition, the mean weight significantly increased across the ‘Egg and potato’ dietary pattern (P = 0·02). However, there was no significant relationship after controlling for confounding factors. Also, there was no significant difference in means and standard deviations of other anthropometric measures across major dietary patterns at lunchtime.

Table 3. Multivariable-adjusted means for anthropometric measures and indexes across tertiles (T) of lunch pattern scores

(Means and standard deviations)

aModel 1: adjusted for age (continuous), bmodel 2: additionally adjusted for marital status, education, physical activity, smoking and cmodel 3: further adjustment for dietary intake of fruits, vegetables, dairy, grains, energy intake.

P-values obtained using ANCOVA test.

Unadjusted and adjusted OR for the participants’ general obesity in the tertiles of lunch patterns are presented in Table 4. According to our findings, after control for confounders, there were no significant associations between ‘Bread, grains and fat’ pattern (OR: 0·89, 95 % CI 0·60, 1·33, P-value = 0·59), ‘Eggs and potato’ pattern (OR: 1·07, 95 % CI 0·71, 1·60, P-value = 0·36) and the Western dietary pattern (OR: 1·32, 95 % CI 0·88, 1·99, P-value = 0·17) with general obesity.

Table 4. General obesity (BMI ≥ 30) across tertiles (T) of dietary patterns score

(Odds ratios and 95 % confidence intervals)

aModel 1: unadjusted, bmodel 2: age, sex, education (categorical), marriage, lifestyle, smoking and cmodel 3: model 2 + physical activity, total energy intake.

Table 5 shows the unadjusted and adjusted OR for central obesity across the tertiles of major dietary patterns at lunchtime. In the unadjusted model, no significant association was found between central obesity, based on WC definition, and dietary patterns which remained unchanged even after adjustment for confounders. No significant difference was also observed in the odds of central obesity, when defined based on a WHR, across the tertiles of the ‘Bread, grains and fat’ dietary pattern (OR: 1·38, 95 % CI 0·98, 1·95). However, after adjustment for confounders, the participants in the top tertile of the ‘Bread, grains and fat’ pattern had a 1·44 times higher chance of central obesity (OR: 1·44, 95 % CI 1·01, 2·07) relative to individuals in the first tertile. No significant associations were observed between adherence to the ‘Western’ pattern and ‘Eggs and potato’ patterns with OR of central obesity in our population.

Table 5. Central obesity across tertiles (T) of dietary patterns score

(Odds ratios and 95 % confidence intervals)

Model 1: unadjusted, model 2: age, sex, education (categorical), marriage, lifestyle, smoking and model 3: model 2 + physical activity and total energy intake.

Discussion

The present study, which was conducted on 850 adults in Tehran, showed that 25·4 % of subjects had general obesity. In addition, 46·6 and 54·8 % of subjects had central obesity based on WC and WHR, respectively. The results of this study showed that participants in the top tertile of the ‘Bread, grains and fat’ pattern had a 44 % higher risk of increased WHR compared with the participants in the lowest tertile. There were no significant associations between the ‘Western’ and ‘Eggs and potato’ patterns at lunchtime and central obesity. Moreover, our findings showed that there was no relationship between major dietary patterns at lunchtime and general obesity.

Socio-demographic determinants of obesity have been of great research interest recently. It is demonstrated that certain health behaviours including smoking and physical activity are associated with obesity(Reference Gruber and Frakes8,Reference Fogelholm and Kukkonen-Harjula9) . Diet, total energy intake, physical activity, sedentary lifestyles and other health-risk behaviours are known to act as proximate, intervening variables in the relationship between socio-economic status and obesity(Reference Singh, Siahpush and Hiatt61) In addition, it was found that obesity prevalence varies across various sex, age and socio-economic groups(Reference Paeratakul, Lovejoy and Ryan10). Socio-economic status is usually measured by education, occupation, employment, income and wealth. For example, lower education, lower occupational status and lower incomes have been associated with a higher prevalence of obesity(Reference McLaren62). It has been found that smokers are likely to be obese(Reference Berlin63). It has been shown that smoking increases levels of cortisol and testosterone, whereas the levels of estradiol and progesterone are decreased(Reference Dušková, Hruškovičová and Šimůnková64). Furthermore, it has been reported that smoking changes dietary habits, leading to a lower intake of fibres, fruits and vegetables(Reference Dušková, Hruškovičová and Šimůnková64). Physical activity has also been established as one of the important predictors of weight gain. In that, those who have enough physical activity, especially during leisure time, are less likely to be obese(Reference Fogelholm and Kukkonen-Harjula9).

Given the complex combinations of nutrients involved in the human diet, identifying dietary patterns may represent the best way to highlight the effects of nutrients and specific foods on health(Reference Ocké65,Reference Hu66) , permitting insight into the synergistic outcomes of nutrients and foods. However, most previous studies have investigated diet in general, regardless of the timing of food intake, and consequently, they were unable to identify specific meal characteristics and compositions(Reference Andersson, Nydahl and Gustafsson67). Bellisle et al., reported that the greatest nutrient consumption over an average day occurs around midday, corresponding to lunch(Reference Bellisle, Dalix and Mennen23). Therefore, due to the large proportion of the daily energy intake attributed to lunch, such meals should, ideally, provide sufficient amounts of macro and micronutrients to help achieve and adhere to dietary guidelines.

Dietary patterns were specified based on factor analysis using the principal component analysis method with varimax rotation. The resulting factors were judged based on the eigenvalues (the total squared load factor of food items in terms of their consumption in g per day) of food groups, and any factor with an eigenvalue > 1 was considered as the main dietary pattern. Load factor values ≥0·2 were used to describe the main food groups that make up each food pattern.

One of the possible reasons for the difference between the diet patterns of the present study with other studies is that the analysis of dietary patterns is strongly dependent upon the study population. Therefore, significant differences in dietary patterns of different populations are observed by geographical area, race and culture. In addition, factor analysis is limited by the researcher’s choice of food items’ grouping and the number of factors to be retained, and these decisions can somewhat affect the findings and their interpretation.

In the present study, three dominant lunch consumption patterns were identified among the participants: dietary pattern 1 (‘Bread, grains and fat’) was rich in fats, bread and grains, salt, poultry and fish, dietary pattern 2 (‘Eggs and potatoes’) included a high intake of potato and eggs, and a low intake of bread and grains, nuts and legumes and dietary pattern 3 (‘Western’) included a high intake of red or processed meat, sauces, soft drink, nuts and legumes and a low intake of poultry and fish and dairy products. Santos et al. previously identified five lunch patterns, among which the ‘Western’ dietary pattern (positive factor loadings for soft drinks, alcoholic beverages, sweets, gnocchi/stuffed pasta, sauces/mayonnaise and processed meats) was relatively similar to our ‘Western’ dietary pattern(Reference de Oliveira Santos, dos Santos Vieira and Miranda68). In addition, our ‘Bread, grains and fat’ dietary pattern was similar to Santos’ ‘meat’ dietary pattern (positive factor loadings for eggs, poultry meat and fish/seafood; negative for beef). However, some of the items found in Santos’ ‘meat’ dietary pattern (like eggs) were also found in our other dietary patterns. Similarly, Schwedhelm et al. also identified five lunch meal patterns(Reference Schwedhelm, Iqbal and Knüppel69), among which the ‘Western’ pattern (high intake of potatoes, cabbage, red meat, beer, sauces and condiments and low intake of fresh fruits, milk and dairy products and tea) was comparable to our ‘Western’ dietary pattern, and their ‘traditional’ dietary pattern (high intake of bread, processed meat, butter, sugar, confectionery, cakes and cookies and low intake of water) was similar to our ‘Bread, grains and fat’ dietary pattern. Moreover, Esmaillzadeh et al. (Iran) identified three dietary patterns(Reference Esmaillzadeh and Azadbakht70), among which the ‘Western’ dietary pattern (high in refined grains, red meat, butter, processed meat, high-fat dairy products, sweets and desserts, pizza, potatoes, eggs, hydrogenated fats, and soft drinks and low in other vegetables and low-fat dairy products) was similar to our ‘Western’ dietary pattern, and the ‘Iranian’ foods pattern (high in refined grains, potato, tea, whole grains, hydrogenated fats, legumes, and broth) was similar to our ‘Bread, grains and fat’ dietary pattern.

Contradictory results among studies may be attributed to distinct differences in the culture, tradition and eating habits of the Iranian people compared with other countries. Additionally, differences in dietary patterns may be attributed to the fact that some studies examined habitual dietary patterns, while in our study, these patterns were derived at the meal level. The present study revealed no association between the ‘Western’ pattern, as an unhealthy dietary pattern, and obesity; indeed, some previous studies have reported similar findings and found no significant association between the patterns rich in fat and sugar and overweight and obesity(Reference de Oliveira Santos, dos Santos Vieira and Miranda68,Reference Craig, McNeill and Macdiarmid71,Reference McNaughton, Ball and Mishra72) . However, it has been reported that adherence to a ‘Western’ or unhealthy pattern dietary increases the risk of overweight and obesity(Reference Lera, Olivares and Leyton73,Reference Herber-Gast and Mishra74) . Given that the ‘Western’ dietary pattern is considered as an unhealthy pattern, the lack of a significant association with obesity was somewhat unexpected. However, the inclusion of some food groups, such as legumes and nuts, in this pattern might have hindered the identification of an association. In addition, no association was found between the ‘Egg and potato’ pattern and obesity, which may be considered as a healthy diet, given that it is rich in fibre, vitamins and high-quality protein. Although the healthy dietary patterns found in some previous studies have been inversely associated with obesity and overweight(Reference Newby, Muller and Hallfrisch75), few studies have shown a positive association between healthy dietary patterns and BMI(Reference Nouri, Tarighat-Esfanjani and Ghazizahedi76). However, some studies have also reported no association between healthy eating patterns and weight status(Reference Craig, McNeill and Macdiarmid71,Reference McNaughton, Ball and Mishra72) , which may conceivably be due to measurement error in the variables under investigation.

In the present study, no significant association was also found between the ‘Bread, grains and fat’ lunch pattern and general obesity. However, participants in the highest tertile had greater odds of increased WHR compared with those in the lowest tertile. The complex nature of this dietary pattern makes interpretation difficult. This pattern contains bread and grains, fats and salt; hence, a positive association was expected between this dietary pattern and obesity. However, some healthy food groups, such as poultry and fish included in this dietary pattern, may have interacted with other foods and counteracted the effects on obesity. On the other hand, it can be argued that the positive association between this dietary pattern and increased WHR is justified by the fact that this food pattern is high in fat and carbohydrates. Low-nutrient density foods can lead to central nervous system insulin resistance, which may, in turn, result in leptin resistance and increased pleasurable responses to foods(Reference Isganaitis and Lustig77).

One of the strengths of the present study is the recruitment of a large sample size compared with other studies in Iran. In addition, all districts of Tehran were selected, so that various socio-economic statuses, educational and welfare levels, and other variables affecting the outcome could be included in the study. The present study also had some potential limitations; for instance, no causal inference could be made due to the cross-sectional nature of the study. Another limitation of this study was that the information regarding the economic conditions of the families could not be collected. As a result, the possible relationship between the economic status of a family with dietary patterns could not be assessed. In addition, the data were collected using retrospective questions, which might have yielded information recall bias.

Conclusion

Greater adherence to the ‘Bread, grains and fat’ dietary pattern at lunchtime was associated with greater odds of central obesity when defined based on WHR. However, no significant association was found between identified lunch patterns and obesity defined by WC and BMI.

Acknowledgements

The authors thank all those who participated in this study. We wish to express our sincere thanks to Dr. Khalid Iqbal from Department of Human Nutrition, Institute of Basic Medical Sciences, Khyber Medical University, Peshawar, Pakistan, for his helpful suggestions during the study. The authors also thank Dr. Carolina Schwedhelm, from Department of Epidemiology, German Institute of Human Nutrition Potsdam-Rehbruecke (DIfE), Nuthetal, Germany for her technical support and help in preparing data.

This manuscript has been granted by Tehran University of Medical Sciences (Grant No: 40186).

S. S.-B. contributed to conception/design of the research; Z. A., S. A. M., N. G., F. S., H. J., M. C. contributed to acquisition of data. Z. A. and S. S.-B. participated in analysis and interpretation of the data; Z. A. drafted the manuscript; K. D, S. S.-B. and C. C. T. C. critically revised the manuscript; and S. S.-B. agrees to be fully accountable for ensuring the integrity and accuracy of the work. All authors read and approved the final manuscript.

We wish to confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome.

References

Antelo, M, Magdalena, P & Reboredo, JC (2017) Obesity: a major problem for Spanish minors. Econ Hum Biol 24, 6173.CrossRefGoogle Scholar
Rahmani, A, Sayehmiri, K, Asadollahi, K, et al. (2015) Investigation of the prevalence of obesity in Iran: a systematic review and meta-analysis study. Acta Med Iranica 53, 596607.Google ScholarPubMed
Cornier, M-A, Despres, J-P, Davis, N, et al. (2011) Assessing adiposity: a scientific statement from the American Heart Association. Circulation 124, 19962019.CrossRefGoogle ScholarPubMed
Esmaillzadeh, A, Mirmiran, P & Azizi, F (2004) Waist-to-hip ratio is a better screening measure for cardiovascular risk factors than other anthropometric indicators in Tehranian adult men. Int J Obes 28, 13251332.CrossRefGoogle ScholarPubMed
Azizi, F, Esmaillzadeh, A, Mirmiran, P, et al. (2005) Is there an independent association between waist-to-hip ratio and cardiovascular risk factors in overweight and obese women? Int J Cardiol 101, 3946.CrossRefGoogle ScholarPubMed
Lee, CMY, Huxley, RR, Wildman, RP, et al. (2008) Indices of abdominal obesity are better discriminators of cardiovascular risk factors than BMI: a meta-analysis. J Clin Epidemiol 61, 646653.CrossRefGoogle ScholarPubMed
Taylor, RW, Brooking, L, Williams, SM, et al. (2010) Body mass index and waist circumference cutoffs to define obesity in indigenous New Zealanders. Am J Clin Nutr 92, 390397.CrossRefGoogle ScholarPubMed
Gruber, J & Frakes, M (2006) Does falling smoking lead to rising obesity? J Health Econ 25, 183197.CrossRefGoogle ScholarPubMed
Fogelholm, M & Kukkonen-Harjula, K (2000) Does physical activity prevent weight gain–a systematic review. Obes Rev 1, 95111.CrossRefGoogle ScholarPubMed
Paeratakul, S, Lovejoy, JC, Ryan, DH, et al. (2002) The relation of gender, race and socioeconomic status to obesity and obesity comorbidities in a sample of US adults. Int J Obes 26, 12051210.CrossRefGoogle Scholar
McCormack, GR, Friedenreich, C, McLaren, L, et al. (2017) Interactions between neighbourhood urban form and socioeconomic status and their associations with anthropometric measurements in Canadian adults. J Environ Public Health 2017, 5042614.CrossRefGoogle ScholarPubMed
Ruf, T, Nagel, G, Altenburg, H-P, et al. (2005) Food and nutrient intake, anthropometric measurements and smoking according to alcohol consumption in the EPIC Heidelberg study. Ann Nutr Metab 49, 1625.CrossRefGoogle ScholarPubMed
Organization WH (2009) Global Health Risks: Mortality and Burden of Disease Attributable to Selected Major Risks. Geneva: World Health Organization.Google Scholar
Hu, FB, Rimm, E, Smith-Warner, SA, et al. (1999) Reproducibility and validity of dietary patterns assessed with a food-frequency questionnaire. Am J Clin Nutr 69, 243249.CrossRefGoogle ScholarPubMed
Hu, FB, Rimm, EB, Stampfer, MJ, et al. (2000) Prospective study of major dietary patterns and risk of coronary heart disease in men. Am J Clin Nutr 72, 912921.CrossRefGoogle ScholarPubMed
Cunha, DB, de Almeida, RMVR, Sichieri, R, et al. (2010) Association of dietary patterns with BMI and waist circumference in a low-income neighbourhood in Brazil. Br J Nutr 104, 908913.CrossRefGoogle Scholar
Naja, F, Hwalla, N, Itani, L, et al. (2015) A Western dietary pattern is associated with overweight and obesity in a national sample of Lebanese adolescents (13–19 years): a cross-sectional study. Br J Nutr 114, 19091919.CrossRefGoogle Scholar
Shu, L, Zheng, P-F, Zhang, X-Y, et al. (2015) Association between dietary patterns and the indicators of obesity among Chinese: a cross-sectional study. Nutrients 7, 79958009.CrossRefGoogle ScholarPubMed
Helsing, E (1995) Traditional diets and disease patterns of the Mediterranean, Circa 1960. Am J Clin Nutr 61, 1329S1337S.CrossRefGoogle ScholarPubMed
Sacks, FM, Svetkey, LP, Vollmer, WM, et al. (2001) Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) diet. New Engl J Med 344, 310.CrossRefGoogle ScholarPubMed
Hu, FB (2002) Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opinion Lipidol 13, 39.CrossRefGoogle ScholarPubMed
Leech, RM, Worsley, A, Timperio, A, et al. (2015) Understanding meal patterns: definitions, methodology and impact on nutrient intake and diet quality. Nutr Res Rev 28, 121.CrossRefGoogle ScholarPubMed
Bellisle, F, Dalix, AM, Mennen, L, et al. (2003) Contribution of snacks and meals in the diet of French adults: a diet-diary study. Physiol Behav 79, 183189.CrossRefGoogle ScholarPubMed
Mesas, A, Muñoz-Pareja, M, López-García, E, et al. (2012) Selected eating behaviours and excess body weight: a systematic review. Obes Rev 13, 106135.CrossRefGoogle ScholarPubMed
Szajewska, H & Ruszczyński, M (2010) Systematic review demonstrating that breakfast consumption influences body weight outcomes in children and adolescents in Europe. Crit Rev Food Sci Nutr 50, 113119.CrossRefGoogle ScholarPubMed
de Oliveira Santos, R, dos Santos Vieira, DA, Miranda, AA, et al. (2018) The traditional lunch pattern is inversely correlated with body mass index in a population-based study in Brazil. BMC Public Health 18, 17.CrossRefGoogle Scholar
Aparicio, A, Rodríguez-Rodríguez, EE, Aranceta-Bartrina, J, et al. (2017) Differences in meal patterns and timing with regard to central obesity in the ANIBES (‘Anthropometric data, macronutrients and micronutrients intake, practice of physical activity, socioeconomic data and lifestyles in Spain’) Study. Public Health Nutr 20, 23642373.CrossRefGoogle ScholarPubMed
Almoosawi, S, Winter, J, Prynne, C, et al. (2012) Daily profiles of energy and nutrient intakes: are eating profiles changing over time? Eur J Clin Nutr 66, 678.CrossRefGoogle ScholarPubMed
Benelam, B & Stanner, S (2015) Development of a methodology to assess the nutrient profile of popular UK meals. News Views 40, 315325.Google Scholar
Ahmad Kiadaliri, A, Jafari, M, Vaez Mahdavi, MR, et al. (2015) The prevalence of adulthood overweight and obesity in Tehran: findings from Urban HEART-2 study. Med J Islamic Republic Iran 29, 178.Google ScholarPubMed
Lohman, TG, Roche, AF & Martorell, R (1988) Anthropometric Standardization Reference Manual. Champaign, IL: Human Kinetics Books.Google Scholar
WHO (2019) Obesity. https://www.who.int/topics/obesity/en/ (accessed July 2019).Google Scholar
Organization WH (2011) Waist Circumference and Waist-Hip Ratio: Report of a WHO Expert Consultation. Geneva: WHO.Google Scholar
Ghaffarpour, M, Houshiar-Rad, A & Kianfar, H (1999) The manual for household measures, cooking yields factors and edible portion of foods. Tehran: Nashre Olume Keshavarzy 7, 213.Google Scholar
Kahleova, H, Lloren, JI, Mashchak, A, et al. (2017) Meal frequency and timing are associated with changes in body mass index in Adventist health study 2. J Nutr 147, 17221728.CrossRefGoogle ScholarPubMed
Karamati, M, Jessri, M, Shariati-Bafghi, S-E, et al. (2012) Dietary patterns in relation to bone mineral density among menopausal Iranian women. Calcified Tissue Int 91, 4049.CrossRefGoogle ScholarPubMed
Azadbakht, L & Esmaillzadeh, A (2012) Dietary patterns and attention deficit hyperactivity disorder among Iranian children. Nutrition 28, 242249.CrossRefGoogle ScholarPubMed
Lockheart, MS, Steffen, LM, Rebnord, HM, et al. (2007) Dietary patterns, food groups and myocardial infarction: a case–control study. Br J Nutr 98, 380387.CrossRefGoogle ScholarPubMed
Serra-Majem, L & Bautista-Castaño, I (2015) Relationship between bread and obesity. Br J Nutr 113, S29S35.CrossRefGoogle ScholarPubMed
Liu, S, Willett, WC, Manson, JE, et al. (2003) Relation between changes in intakes of dietary fiber and grain products and changes in weight and development of obesity among middle-aged women. Am J Clin Nutr 78, 920927.CrossRefGoogle ScholarPubMed
Wang, W, Wu, Y & Zhang, D (2016) Association of dairy products consumption with risk of obesity in children and adults: a meta-analysis of mainly cross-sectional studies. Ann Epidemiol 26, 870882 e872.CrossRefGoogle ScholarPubMed
Barba, G & Russo, P (2006) Dairy foods, dietary calcium and obesity: a short review of the evidence. Nutr Metab Cardiovasc Dis 16, 445451.CrossRefGoogle Scholar
Donma, MM & Donma, O (2017) Beneficial effects of poultry meat consumption on cardiovascular health and the prevention of childhood obesity. Med One 2, e170018.Google Scholar
Petermann-Rocha, F, Parra-Soto, S, Gray, S, et al. (2021) Vegetarians, fish, poultry, and meat-eaters: who has higher risk of cardiovascular disease incidence and mortality? A prospective study from UK Biobank. Eur Heart J 42, 11361143.CrossRefGoogle ScholarPubMed
Liu, R, Zhao, Y, Li, Q, et al. (2020) Body fat mass, fat distribution and egg consumption: a population-based study in Chinese adults: egg consumption and body fat in Rural Chinese. J Am Coll Nutr 39, 528536.CrossRefGoogle Scholar
Saande, CJ, Bries, AE, Pritchard, SK, et al. (2020) Whole egg consumption decreases cumulative weight gain in diet-induced obese rats. J Nutr 150, 18181823.CrossRefGoogle ScholarPubMed
Lissner, L & Heitmann, BL (1995) Dietary fat and obesity: evidence from epidemiology. Eur J Clin Nutr 49, 7990.Google ScholarPubMed
Gillis, L, Kennedy, L, Gillis, A, et al. (2002) Relationship between juvenile obesity, dietary energy and fat intake and physical activity. Int J Obes 26, 458463.CrossRefGoogle ScholarPubMed
Borch, D, Juul-Hindsgaul, N, Veller, M, et al. (2016) Potatoes and risk of obesity, type 2 diabetes, and cardiovascular disease in apparently healthy adults: a systematic review of clinical intervention and observational studies. Am J Clin Nutr 104, 489498.CrossRefGoogle ScholarPubMed
Schwingshackl, L, Schwedhelm, C, Hoffmann, G, et al. (2019) Potatoes and risk of chronic disease: a systematic review and dose–response meta-analysis. Eur J Nutr 58, 22432251.CrossRefGoogle ScholarPubMed
Rouhani, M, Salehi-Abargouei, A, Surkan, P, et al. (2014) Is there a relationship between red or processed meat intake and obesity? A systematic review and meta-analysis of observational studies. Obes Rev 15, 740748.CrossRefGoogle ScholarPubMed
Dabbagh-Moghadam, A, Mozaffari-Khosravi, H, Nasiri, M, et al. (2017) Association of white and red meat consumption with general and abdominal obesity: a cross-sectional study among a population of Iranian military families in 2016. Eat Weight Disord-Stud Anorexia Bulimia Obes 22, 717724.CrossRefGoogle ScholarPubMed
Gibson, S (2008) Sugar-sweetened soft drinks and obesity: a systematic review of the evidence from observational studies and interventions. Nutr Res Rev 21, 134147.CrossRefGoogle ScholarPubMed
James, J & Kerr, D (2005) Prevention of childhood obesity by reducing soft drinks. Int J Obes 29, S54S57.CrossRefGoogle ScholarPubMed
He, K, Hu, F, Colditz, G, et al. (2004) Changes in intake of fruits and vegetables in relation to risk of obesity and weight gain among middle-aged women. Int J Obes 28, 15691574.CrossRefGoogle ScholarPubMed
Martínez-González, M & Bes-Rastrollo, M (2011) Nut consumption, weight gain and obesity: epidemiological evidence. Nutr Metab Cardiovasc Dis 21, S40S45.CrossRefGoogle ScholarPubMed
Ma, Y, He, FJ & MacGregor, GA (2015) High salt intake: independent risk factor for obesity? Hypertension 66, 843849.CrossRefGoogle ScholarPubMed
Appleton, KM (2009) Increases in energy, protein and fat intake following the addition of sauce to an older person’s meal. Appetite 52, 161165.CrossRefGoogle Scholar
Ainsworth, BE, Haskell, WL, Whitt, MC, et al. (2000) Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sport Exerc 32, S498S504.CrossRefGoogle ScholarPubMed
Rezazadeh, A, Rashidkhani, B & Omidvar, N (2010) Association of major dietary patterns with socioeconomic and lifestyle factors of adult women living in Tehran, Iran. Nutrition 26, 337341.CrossRefGoogle ScholarPubMed
Singh, GK, Siahpush, M, Hiatt, RA, et al. (2011) Dramatic increases in obesity and overweight prevalence and body mass index among ethnic-immigrant and social class groups in the United States, 1976–2008. J Community Health 36, 94110.CrossRefGoogle ScholarPubMed
McLaren, L (2007) Socioeconomic status and obesity. Epidemiol Rev 29, 2948.CrossRefGoogle ScholarPubMed
Berlin, I (2009) Endocrine and metabolic effects of smoking cessation. Curr Med Res Opin 25, 527534.CrossRefGoogle ScholarPubMed
Dušková, M, Hruškovičová, H, Šimůnková, K, et al. (2014) The effects of smoking on steroid metabolism and fetal programming. J Steroid Biochem Mol Biol 139, 138143.CrossRefGoogle ScholarPubMed
Ocké, MC (2013) Evaluation of methodologies for assessing the overall diet: dietary quality scores and dietary pattern analysis. Proc Nutr Soc 72, 191199.CrossRefGoogle ScholarPubMed
Hu, FB (2002) Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol 13, 39.CrossRefGoogle ScholarPubMed
Andersson, J, Nydahl, M, Gustafsson, K, et al. (2003) Meals and snacks among elderly self-managing and disabled women. Appetite 41, 149160.CrossRefGoogle ScholarPubMed
de Oliveira Santos, R, dos Santos Vieira, DA, Miranda, AA, et al. (2018) The traditional lunch pattern is inversely correlated with body mass index in a population-based study in Brazil. BMC Public Health 18, 33.CrossRefGoogle Scholar
Schwedhelm, C, Iqbal, K, Knüppel, S, et al. (2018) Contribution to the understanding of how principal component analysis–derived dietary patterns emerge from habitual data on food consumption. Am J Clin Nutr 107, 227235.CrossRefGoogle Scholar
Esmaillzadeh, A & Azadbakht, L (2008) Major dietary patterns in relation to general obesity and central adiposity among Iranian women. J Nutr 138, 358363.CrossRefGoogle ScholarPubMed
Craig, LC, McNeill, G, Macdiarmid, JI, et al. (2010) Dietary patterns of school-age children in Scotland: association with socio-economic indicators, physical activity and obesity. Br J Nutr 103, 319334.CrossRefGoogle ScholarPubMed
McNaughton, SA, Ball, K, Mishra, GD, et al. (2008) Dietary patterns of adolescents and risk of obesity and hypertension. J Nutr 138, 364370.CrossRefGoogle ScholarPubMed
Lera, LM, Olivares, SC, Leyton, BD, et al. (2006) Dietary patterns and its relation with overweight and obesity in Chilean girls of medium-high socioeconomic level. Arch Latinoamericanos de Nutr 56, 165170.Google Scholar
Herber-Gast, G-CM & Mishra, GD (2013) Fruit, Mediterranean-style, and high-fat and-sugar diets are associated with the risk of night sweats and hot flushes in midlife: results from a prospective cohort study. Am Clin Nutr 97, 10921099.CrossRefGoogle Scholar
Newby, P, Muller, D, Hallfrisch, J, et al. (2004) Food patterns measured by factor analysis and anthropometric changes in adults.Am J Clin Nutr 80, 504513.CrossRefGoogle ScholarPubMed
Nouri, M, Tarighat-Esfanjani, A & Ghazizahedi, S (2017) The relationships between dietary patterns and energy and nutrient intakes and body mass index in Iranian Adults. J Health 8, 85101.Google Scholar
Isganaitis, E & Lustig, RH (2005) Fast food, central nervous system insulin resistance, and obesity. Arterioscler Thromb Vasc Biol 25, 24512462.CrossRefGoogle ScholarPubMed
Figure 0

Fig. 1. The distribution of energy content was consumed across the time of day. , mean energy.

Figure 1

Table 1. Food groups used in the factor analysis and factor loadings for each of the identified lunch patterns*

Figure 2

Table 2. Characteristics of the study participants by tertiles (T) of lunch pattern scores(Numbers and percentages; mean values and standard deviations)

Figure 3

Fig. 2. Heat map showing the Pearson correlation matrix for lunch food intake in g (n 790) by food groups. The colour corresponds to the strength of correlations (red: positive correlation; white: no correlation; blue: negative correlation).

Figure 4

Table 3. Multivariable-adjusted means for anthropometric measures and indexes across tertiles (T) of lunch pattern scores(Means and standard deviations)

Figure 5

Table 4. General obesity (BMI ≥ 30) across tertiles (T) of dietary patterns score(Odds ratios and 95 % confidence intervals)

Figure 6

Table 5. Central obesity across tertiles (T) of dietary patterns score(Odds ratios and 95 % confidence intervals)