Eur J Nutr DOI 10.1007/s00394-015-0891-4
ORIGINAL CONTRIBUTION
Patterns of dietary habits in relation to obesity in Iranian adults Parvane Saneei1,2,3 · Ahmad Esmaillzadeh1,2 · Ammar Hassanzadeh Keshteli4,5 · Awat Feizi6 · Christine Feinle‑Bisset7 · Peyman Adibi4
Received: 26 August 2014 / Accepted: 24 March 2015 © Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose Findings from few studies that investigated the relation between dietary behaviors and obesity are inconsistent. We aimed to assess the relation between patterns of dietary habits, identified by latent class analysis (LCA), and obesity in a large sample of Iranian adults. Methods In a cross-sectional study on 7958 adults, dietary behaviors were assessed in five domains (meal patterns, eating rate, intra-meal fluid intake, meal-to-sleep interval, and fatty foods intake) using a pretested questionnaire. LCA was applied to identify classes of diet-related
* Ammar Hassanzadeh Keshteli [email protected] Parvane Saneei [email protected] 1
Food Security Research Center, Isfahan University of Medical Sciences, Isfahan, Iran
2
Department of Community Nutrition, School of Nutrition and Food Science, Isfahan University of Medical Sciences, Isfahan, Iran
3
Students’ Research Committee, Isfahan University of Medical Sciences, Isfahan, Iran
4
Integrative Functional Gastroenterology Research Center, Isfahan University of Medical Sciences, Isfahan, Iran
5
CEGIIR‑Division of Gastroenterology, Department of Medicine, University of Alberta, Edmonton, AB T6G 2E1, Canada
6
Department of Epidemiology and Biostatistics, School of Public Health, Isfahan University of Medical Sciences, Isfahan, Iran
7
Centre of Research Excellence in Translating, Nutritional Science to Good Health, University of Adelaide, Adelaide, Australia
practices. Anthropometric measures were assessed through the use of a validated self-reported questionnaire. General and abdominal obesity were defined as a body mass index ≥30 kg/m2, and a waist circumference ≥88 cm for women and ≥102 cm for men. Results General and abdominal obesity were prevalent in 9.7 and 27.7 % of the study population, respectively. We identified three distinct classes of eating rates (moderate, moderate to slow, and moderate to fast), two classes of meal patterns (regular and irregular), two classes of intra-meal fluid intake (moderate and more intra-meal drinking), three classes of meal-to-sleep interval (short, moderate, and long meal-to-sleep interval), and three classes of fatty food intake (low to moderate, moderate to high, and low intake of fatty foods). After adjustment for potential confounders, individuals with ‘irregular meal pattern’ were 21, 24, and 22 % more likely to be overweight/obese, abdominally overweight/obese, and abdominally obese, compared with those who had a ‘regular meal pattern.’ Individuals with ‘more intra-meal drinking’ had greater odds of overweight (OR 1.37; 1.19– 1.458) and obesity (OR 1.51; 1.16–1.97) than those with ‘moderate intra-meal drinking.’ Moderate-to-high intake of fatty foods was inversely associated with abdominally overweight/obese (OR 0.85; 0.73–1.00) and abdominally obesity (OR 0.80; 0.68–0.96) compared with ‘low-tomoderate intake of fatty foods.’ No significant association was observed between eating rate, meal-to-sleep interval, and general or abdominal obesity, after controlling for confounders. Conclusion Irregular meal pattern and more intra-meal drinking were associated with increased odds of general and abdominal obesity, whereas moderate-to-high intake of fatty foods was related to the decreased odds of central obesity among Iranian adults.
13
Keywords Dietary habits · Obesity · Abdominal obesity · Latent class analysis · Eating rate · Fluid intake · Meal regularity · Meal-to-sleep interval · Fatty food intake
Introduction Obesity is a worldwide public health problem affecting both developed and developing nations and is now being described as a global epidemic [1]. The numerous psychological, physical, and economic consequences of obesity are well known. Obesity affects self-esteem and has negative effects on cognitive and social behaviors [2]. Chronic conditions such as type 2 diabetes, hypertension, and cardiovascular diseases are much more common among obese adults [1]. Recent national estimates have reported that more than 50 % of Iranian adult populations are overweight or obese [3]. Obesity is a multifactorial condition in which genetic, psychological, environmental, and lifestyle-related factors, including diet, are involved [4]. So far, most studies on diet–obesity relations have focused on dietary intakes [5]. Summerbell et al. [6] showed that dietary intakes, breast-feeding, food preparation methods, physical inactivity, energy intake, and dietary energy density were among the main dietary culprits of obesity [6]. However, just a few studies have investigated the role of common dietary behaviors in the development of obesity. Although some studies reported the positive relation between eating speed and obesity [7–10], such findings have not been confirmed by others [11]. Eating before bedtime has also been linked to weight gain [12], but not consistently [13]. Frequency and size of meals have also been associated with obesity [11, 14]. It must be noted that prior studies have mostly focused on a single dietary habit, and no information is available about the association between patterns of dietary habits and obesity. Since the individual effects of each dietary habit might be too small to be detected, examining the association between patterns of multiple dietary behaviors and obesity might provide novel insights into the etiology of this condition. Therefore, the use of statistical methods to derive combined dietary habits has recently attracted great attention in nutritional epidemiology. The translation of these combined habits into practice is much easier than the individual behavior because people do not practice the diet-related habits separately; instead, they are behaving with a complexity of practices. Furthermore, the combined effects of dietary habits could be sufficiently large to be detected in assessing diet–health relations. Identification of the favorable patterns of dietary habits could result in the development of preventive measures to control obesity. Adherence to healthy dietary habits is much easier than following a dietary plan to prevent or manage obesity [15]. It
13
Eur J Nutr
is worthy to note that most previous studies on dietary habits and obesity have been conducted in western countries, and currently, there is few related information from developing countries. In the current study, we assessed patterns of dietary habits by the use of latent class analysis (LCA). The advantage of LCA, a person-centered approach, is its ability to estimate unique profiles of probabilities for each mutually exclusive latent class. Similar to cluster analysis, latent class analysis (LCA) [16] is used to classify individuals into mutually exclusive groups such that variation in dietary habits is maximized across different patterns while individuals within each pattern have similar eating habits. Another strength of LCA is the ability to adjust for covariates, quantification of the uncertainty of class membership, and assessment of goodness of fit [16]. However, this method has its own limitations that can be found elsewhere [16]. This study was, therefore, carried out to examine the associations between patterns of common dietary habits and obesity in large sample of Iranian adults.
Materials and methods Participants This cross-sectional study was carried out within the framework of the Study on the Epidemiology of Psychological, Alimentary Health and Nutrition (SEPAHAN) [17], a project that was performed on Iranian general adults working in 50 health centers affiliated to Isfahan University of Medical Sciences (IUMS). A detailed self-administered questionnaire on sociodemographic factors and dietary behaviors was distributed among 10,087 apparently healthy adults, and 8691 individuals (response rate 86.16 %) returned the completed questionnaires. When we compared the demographic data between those who returned the completed questionnaires and those who did not, no significant difference was found. In the current analysis, we used data from 7958 adults who had complete information on body weight and height, as well as 6054 adults with complete waist circumference (WC) values. Participants provided signed informed written consents. The study was ethically approved by the Medical Research Ethics Committee of IUMS, Isfahan, Iran. Assessment of dietary habits Based on earlier studies [8, 9, 13], dietary behaviors were pre-defined and assessed in five domains (meal patterns, eating rate, intra-meal fluid intake, meal-to-sleep interval, and fatty food intake) using a self-administered questionnaire. Detailed information about the questions included in each domain as well as the response categories for each
Eur J Nutr Table 1 Patterns of diet-related practices identified by latent class analysis among participants Type of diet-related practices and their items specific
Regular
Irregular
0.60
0.40
0 0.01 0.99
0.10 0.68 0.22
0.05 0.27 0.48 0.20
0.15 0.41 0.36 0.08
0 0.001 0.02 0.98
0.03 0.08 0.12 0.77
0.01 0.04 0.14 0.81
0.21 0.27 0.20 0.32
0.02 0.07 0.09 0.82
0.18 0.28 0.13 0.41
Moderate
Moderate to slow
Moderate to fast
0.66
0.24
0.10
0.14 0.76 0.10
0.05 0.69 0.26
0.43 0.55 0.02
0 0.0009 0.92 0.08
0.03 0.81 0.16 0
0 0 0.01 0.99
0.06 0.12 0.64 0.18
0.24 0.21 0.49 0.06
0 0 0.02 0.98
Meal pattern Class size How many meals do you eat per day? 1 2 3 How often do you eat your meals regularly? Never Occasionally Often Always How often do you eat breakfast during a week? Never or once per week 2–4 times per week 5–6 times per week Everyday How often do you eat lunch during a week? Never or once per week 2–4 times per week 5–6 times per week Everyday How often do you eat dinner during a week? Never or once week 2–4 times per week 5–6 times per week Everyday
Eating rate Class size How thoroughly do you chew food? Not very well Well Very well How long does it take you to eat lunch? Never eat 20 min How long does it take you to eat dinner? Never eat dinner 20 min
Class-specific response pattern of participants
13
Eur J Nutr
Table 1 continued Moderate intra-meal drinking
More intra-meal drinking
Intra-meal fluid intake Class size 0.70 0.30 How often do you drink fluids (water, dough, cola, other soft drinks) with meals or immediately before and after meals? Never 0.09 0.01 Occasionally 0.49 0.19 Often 0.28 0.32 Always 0.14 0.48 How many glasses of fluids do you drink during meals? ≤1 glass 0.96 0.08 2–3 glasses 0.04 0.79 3–4 glasses 0 0.10 ≥4 glasses 0 0.03 How many glasses of water do you drink every day? 8 glasses 0.02 0.11 Short meal-to-sleep interval
Moderate meal-to-sleep interval
Meal-to-sleep interval Class size 0.62 0.27 How long does it take you to go to bed (or to lie down) after lunch? Never sleep 0.21 0.33
ORIGINAL CONTRIBUTION
Patterns of dietary habits in relation to obesity in Iranian adults Parvane Saneei1,2,3 · Ahmad Esmaillzadeh1,2 · Ammar Hassanzadeh Keshteli4,5 · Awat Feizi6 · Christine Feinle‑Bisset7 · Peyman Adibi4
Received: 26 August 2014 / Accepted: 24 March 2015 © Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose Findings from few studies that investigated the relation between dietary behaviors and obesity are inconsistent. We aimed to assess the relation between patterns of dietary habits, identified by latent class analysis (LCA), and obesity in a large sample of Iranian adults. Methods In a cross-sectional study on 7958 adults, dietary behaviors were assessed in five domains (meal patterns, eating rate, intra-meal fluid intake, meal-to-sleep interval, and fatty foods intake) using a pretested questionnaire. LCA was applied to identify classes of diet-related
* Ammar Hassanzadeh Keshteli [email protected] Parvane Saneei [email protected] 1
Food Security Research Center, Isfahan University of Medical Sciences, Isfahan, Iran
2
Department of Community Nutrition, School of Nutrition and Food Science, Isfahan University of Medical Sciences, Isfahan, Iran
3
Students’ Research Committee, Isfahan University of Medical Sciences, Isfahan, Iran
4
Integrative Functional Gastroenterology Research Center, Isfahan University of Medical Sciences, Isfahan, Iran
5
CEGIIR‑Division of Gastroenterology, Department of Medicine, University of Alberta, Edmonton, AB T6G 2E1, Canada
6
Department of Epidemiology and Biostatistics, School of Public Health, Isfahan University of Medical Sciences, Isfahan, Iran
7
Centre of Research Excellence in Translating, Nutritional Science to Good Health, University of Adelaide, Adelaide, Australia
practices. Anthropometric measures were assessed through the use of a validated self-reported questionnaire. General and abdominal obesity were defined as a body mass index ≥30 kg/m2, and a waist circumference ≥88 cm for women and ≥102 cm for men. Results General and abdominal obesity were prevalent in 9.7 and 27.7 % of the study population, respectively. We identified three distinct classes of eating rates (moderate, moderate to slow, and moderate to fast), two classes of meal patterns (regular and irregular), two classes of intra-meal fluid intake (moderate and more intra-meal drinking), three classes of meal-to-sleep interval (short, moderate, and long meal-to-sleep interval), and three classes of fatty food intake (low to moderate, moderate to high, and low intake of fatty foods). After adjustment for potential confounders, individuals with ‘irregular meal pattern’ were 21, 24, and 22 % more likely to be overweight/obese, abdominally overweight/obese, and abdominally obese, compared with those who had a ‘regular meal pattern.’ Individuals with ‘more intra-meal drinking’ had greater odds of overweight (OR 1.37; 1.19– 1.458) and obesity (OR 1.51; 1.16–1.97) than those with ‘moderate intra-meal drinking.’ Moderate-to-high intake of fatty foods was inversely associated with abdominally overweight/obese (OR 0.85; 0.73–1.00) and abdominally obesity (OR 0.80; 0.68–0.96) compared with ‘low-tomoderate intake of fatty foods.’ No significant association was observed between eating rate, meal-to-sleep interval, and general or abdominal obesity, after controlling for confounders. Conclusion Irregular meal pattern and more intra-meal drinking were associated with increased odds of general and abdominal obesity, whereas moderate-to-high intake of fatty foods was related to the decreased odds of central obesity among Iranian adults.
13
Keywords Dietary habits · Obesity · Abdominal obesity · Latent class analysis · Eating rate · Fluid intake · Meal regularity · Meal-to-sleep interval · Fatty food intake
Introduction Obesity is a worldwide public health problem affecting both developed and developing nations and is now being described as a global epidemic [1]. The numerous psychological, physical, and economic consequences of obesity are well known. Obesity affects self-esteem and has negative effects on cognitive and social behaviors [2]. Chronic conditions such as type 2 diabetes, hypertension, and cardiovascular diseases are much more common among obese adults [1]. Recent national estimates have reported that more than 50 % of Iranian adult populations are overweight or obese [3]. Obesity is a multifactorial condition in which genetic, psychological, environmental, and lifestyle-related factors, including diet, are involved [4]. So far, most studies on diet–obesity relations have focused on dietary intakes [5]. Summerbell et al. [6] showed that dietary intakes, breast-feeding, food preparation methods, physical inactivity, energy intake, and dietary energy density were among the main dietary culprits of obesity [6]. However, just a few studies have investigated the role of common dietary behaviors in the development of obesity. Although some studies reported the positive relation between eating speed and obesity [7–10], such findings have not been confirmed by others [11]. Eating before bedtime has also been linked to weight gain [12], but not consistently [13]. Frequency and size of meals have also been associated with obesity [11, 14]. It must be noted that prior studies have mostly focused on a single dietary habit, and no information is available about the association between patterns of dietary habits and obesity. Since the individual effects of each dietary habit might be too small to be detected, examining the association between patterns of multiple dietary behaviors and obesity might provide novel insights into the etiology of this condition. Therefore, the use of statistical methods to derive combined dietary habits has recently attracted great attention in nutritional epidemiology. The translation of these combined habits into practice is much easier than the individual behavior because people do not practice the diet-related habits separately; instead, they are behaving with a complexity of practices. Furthermore, the combined effects of dietary habits could be sufficiently large to be detected in assessing diet–health relations. Identification of the favorable patterns of dietary habits could result in the development of preventive measures to control obesity. Adherence to healthy dietary habits is much easier than following a dietary plan to prevent or manage obesity [15]. It
13
Eur J Nutr
is worthy to note that most previous studies on dietary habits and obesity have been conducted in western countries, and currently, there is few related information from developing countries. In the current study, we assessed patterns of dietary habits by the use of latent class analysis (LCA). The advantage of LCA, a person-centered approach, is its ability to estimate unique profiles of probabilities for each mutually exclusive latent class. Similar to cluster analysis, latent class analysis (LCA) [16] is used to classify individuals into mutually exclusive groups such that variation in dietary habits is maximized across different patterns while individuals within each pattern have similar eating habits. Another strength of LCA is the ability to adjust for covariates, quantification of the uncertainty of class membership, and assessment of goodness of fit [16]. However, this method has its own limitations that can be found elsewhere [16]. This study was, therefore, carried out to examine the associations between patterns of common dietary habits and obesity in large sample of Iranian adults.
Materials and methods Participants This cross-sectional study was carried out within the framework of the Study on the Epidemiology of Psychological, Alimentary Health and Nutrition (SEPAHAN) [17], a project that was performed on Iranian general adults working in 50 health centers affiliated to Isfahan University of Medical Sciences (IUMS). A detailed self-administered questionnaire on sociodemographic factors and dietary behaviors was distributed among 10,087 apparently healthy adults, and 8691 individuals (response rate 86.16 %) returned the completed questionnaires. When we compared the demographic data between those who returned the completed questionnaires and those who did not, no significant difference was found. In the current analysis, we used data from 7958 adults who had complete information on body weight and height, as well as 6054 adults with complete waist circumference (WC) values. Participants provided signed informed written consents. The study was ethically approved by the Medical Research Ethics Committee of IUMS, Isfahan, Iran. Assessment of dietary habits Based on earlier studies [8, 9, 13], dietary behaviors were pre-defined and assessed in five domains (meal patterns, eating rate, intra-meal fluid intake, meal-to-sleep interval, and fatty food intake) using a self-administered questionnaire. Detailed information about the questions included in each domain as well as the response categories for each
Eur J Nutr Table 1 Patterns of diet-related practices identified by latent class analysis among participants Type of diet-related practices and their items specific
Regular
Irregular
0.60
0.40
0 0.01 0.99
0.10 0.68 0.22
0.05 0.27 0.48 0.20
0.15 0.41 0.36 0.08
0 0.001 0.02 0.98
0.03 0.08 0.12 0.77
0.01 0.04 0.14 0.81
0.21 0.27 0.20 0.32
0.02 0.07 0.09 0.82
0.18 0.28 0.13 0.41
Moderate
Moderate to slow
Moderate to fast
0.66
0.24
0.10
0.14 0.76 0.10
0.05 0.69 0.26
0.43 0.55 0.02
0 0.0009 0.92 0.08
0.03 0.81 0.16 0
0 0 0.01 0.99
0.06 0.12 0.64 0.18
0.24 0.21 0.49 0.06
0 0 0.02 0.98
Meal pattern Class size How many meals do you eat per day? 1 2 3 How often do you eat your meals regularly? Never Occasionally Often Always How often do you eat breakfast during a week? Never or once per week 2–4 times per week 5–6 times per week Everyday How often do you eat lunch during a week? Never or once per week 2–4 times per week 5–6 times per week Everyday How often do you eat dinner during a week? Never or once week 2–4 times per week 5–6 times per week Everyday
Eating rate Class size How thoroughly do you chew food? Not very well Well Very well How long does it take you to eat lunch? Never eat 20 min How long does it take you to eat dinner? Never eat dinner 20 min
Class-specific response pattern of participants
13
Eur J Nutr
Table 1 continued Moderate intra-meal drinking
More intra-meal drinking
Intra-meal fluid intake Class size 0.70 0.30 How often do you drink fluids (water, dough, cola, other soft drinks) with meals or immediately before and after meals? Never 0.09 0.01 Occasionally 0.49 0.19 Often 0.28 0.32 Always 0.14 0.48 How many glasses of fluids do you drink during meals? ≤1 glass 0.96 0.08 2–3 glasses 0.04 0.79 3–4 glasses 0 0.10 ≥4 glasses 0 0.03 How many glasses of water do you drink every day? 8 glasses 0.02 0.11 Short meal-to-sleep interval
Moderate meal-to-sleep interval
Meal-to-sleep interval Class size 0.62 0.27 How long does it take you to go to bed (or to lie down) after lunch? Never sleep 0.21 0.33
