Energy and macronutrient intake over the course of the day of German adults: A DEDIPAC-study

Appetite 114 (2017) 125e136

Contents lists available at ScienceDirect

Appetite journal homepage: www.elsevier.com/locate/appet

Energy and macronutrient intake over the course of the day of German adults: A DEDIPAC-study Friederike Wittig, Eva Hummel, Germaine Wenzler, Thorsten Heuer* Max Rubner-Institut, Federal Research Institute of Nutrition and Food, Department of Nutritional Behavior, Haid-und-Neu-Str.9, 76131 Karlsruhe, Germany

a r t i c l e i n f o

a b s t r a c t

Article history: Received 4 November 2016 Received in revised form 12 February 2017 Accepted 11 March 2017 Available online 16 March 2017

The aim of the study was to analyze the energy and macronutrient intake over the course of the day of selected population groups in Germany defined by sex, age, BMI, SES, and diet quality. The study was based on food consumption data from the German National Nutrition Survey II (2005e2007) assessed by two 4-day dietary weighing records of 662 women and men aged between 18 and 80 years. Energy and macronutrient intake were calculated using the German Nutrient Database 3.02 and summarized for the periods ‘morning’, ‘midday’, ‘afternoon’, ‘evening’, and ‘night’. Generalized estimating equation models were used to examine differences in energy and macronutrient intake. For women and men, a threemain-meal pattern (‘morning’, ‘midday’, and ‘evening’) was observed, indicated as peaks in energy intake at 08:00 to 09:00, 13:00 and 19:00 o'clock. The distributions of carbohydrate, protein, and fat intake mirror the distribution of energy intake over the course of the day. The highest energy intake was found in the ‘evening’ period, especially in young adults, overweight persons, persons with a high SES, and men with a low diet quality. Women of the oldest age group showed a similar energy intake across the three-main-meals in contrast to young adults, who had lower peaks in the ‘morning’ and ’midday’ periods as well as a shift to later meal times. Young adults seem to have a higher variability in energy intake and a less distinct meal pattern, while seniors have a more structured day. Because a high energy intake in the ‘evening’ period is associated with negative health-related factors, the distribution of energy intake should be considered by recommendations for a healthy nutritional behavior. © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Meal pattern Time of energy intake Time of macronutrient intake Adults Germany NVS II

1. Introduction Consumption of food and beverages, and thus energy and macronutrient intake, is usually distributed over several meals and snacks over the course of the day. This can also be described as meal pattern (de Castro, 2004). Currently, there is a controversial discussion about an increasing preference for snacks (examples: Duffey, Pereira, & Popkin, 2013; Mestdag, 2005; Ovaskainen et al., 2006) leading to a disadvantage of main meals (examples: Cutler, Glaeser, & Shapiro, 2003; Lund & Gronow, 2014; Ovaskainen, Tapanainen, & Pakkala, 2010). Wang et al. (2014, p. 255) described that “the timing of energy intake is a modifiable behavior”. Therefore, data on meal patterns over the course of the day are important for the development of recommendations for a healthy diet. Bellisle (2014) showed that snacking may have negative health affects for example as a result of a higher total

* Corresponding author. E-mail address: [email protected] (T. Heuer).

energy intake. However, data on meal patterns over the course of the day are scarce and need to be explored. In many countries, meal patterns are still traditionally structured into the three-main-meals in the morning, midday, and evening (Bellisle et al., 2003; de Graaf, 2000; Huseinovic et al., 2016; Kant & Graubard, 2015; Riou, Lefevre, Parizot, Lhuissier, & Chauvin, 2015; Tani et al., 2015; Winkler, Doring, & Keil, 1999). But few studies compared different population groups regarding their energy intake of main meals or over the course of the day. The results of these studies suggest that the distribution or timing of energy intake over the course of the day differs between sexes €tta €la €, 1997; Tani (examples: Ovaskainen et al., 2006; Roos & Pra et al., 2015), age groups (examples: Ma et al., 2005; StriegelMoore & Franko, 2006; Wang et al., 2014), SES (socioeconomic status) groups (example: Almoosawi, Prynne, Hardy, & Stephen, 2013b; Kant, Schatzkin, & Ballard-Barbash, 1997), and BMI (body us Forslund, Lindroos, mass index) groups (examples: Berte Sjostrom, & Lissner, 2002; Wang et al., 2014), and may influence BMI (examples: de Castro, 2004; Garaulet et al., 2013). For the

http://dx.doi.org/10.1016/j.appet.2017.03.018 0195-6663/© 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

126

F. Wittig et al. / Appetite 114 (2017) 125e136

German population it is unknown whether the three-main-mealpattern still exists and whether there are differences within population groups e.g. defined by different age. The importance of the topic of meal structure is also stressed by current research within the framework of the ‘DEterminants of DIet and Physical ACtivity - Knowledge Hub’ (DEDIPAC KH), the first action of the European Union's joint programming initiative (JPI). One of the objectives of DEDIPAC is to improve the understanding of the determinants of dietary behavior, physical activity, and sedentary behavior (Lakerveld et al., 2014) and to find knowledge gaps in this field. Meal pattern was identified as a determinant of dietary behavior (Stok et al., 2017). Thus, the purpose of the present study, which was undertaken within the framework of the DEDIPAC KH, was to analyze energy and macronutrient intake over the course of day of German adults based on the data of the German National Nutrition Survey (NVS) II. Information on food consumption was given for several days, including different weekdays and the time when foods were consumed. The following questions were approached: How are energy and macronutrient intake distributed over the course of the day in selected population groups defined by sex, age, BMI, SES and dietary quality (measured by a Healthy Eating Index)? Is there still a traditional three-main-meal pattern in Germany? Are there differences in energy intake at and between defined periods?

days with only water, tea, or coffee consumption (and therefore non-energy intake) were excluded from the analysis. As participants completed two 4-day dietary weighing records, they had one or in rare cases more than one double weekday. To achieve an equal distribution of all weekdays, the chronological later day(s) of the doubling day(s) were excluded from the analysis. The later day was excluded under the assumption that fatigue or boredom may in, & de crease with the duration of the recording (van Staveren, Ocke Vries, 2012; Yannakoulia et al., 2012). Furthermore, participants with missing BMI-values could not be included in the analysis. Underreporters were excluded from the analysis, as underreporting deranges the information on energy intake. On the basis of the ratio of energy intake and calculated resting energy expenditure, underreporting was defined by the cut-off
1.053 using the cut-off 2 formula derived by Goldberg et al. (1991) and adopted by Black (2000). Resting energy expenditure was calculated using the formula of Müller et al. (2004), considering age, sex, and individually measured body weight. A total of 201 participants (129 women and 72 men) were identified as underreporters. Altogether, 662 participants and 3997 recorded days, with an overall of 19985 observations, remained for analysis. All analyses described in this paper were performed with this dataset.

2. Material and methods

To investigate energy and macronutrient intake over the course of the day, five periods were defined: ‘morning’ (05:30 to 11:29), ‘midday’ (11:30 to 14:29), ‘afternoon’ (14:30 to 17:29), ‘evening’ (17:30 to 23:29), and ‘night’ (23:30 to 5:29). To take into account the afternoon snack, the time from 11:30 to 17:29 was subdivided into two 3 h-intervals (11:30 to 14:29 and 14:30 to 17:29). The five periods were defined after a first descriptive data analysis and visual examination of the graphs showing energy intake over the course of the day of the study population. The cuts between the periods were set according to the minima in the course of the day curve and were aligned with the meal times in literature conus cerning meal pattern (examples: Bellisle et al., 2003; Berte Forslund et al., 2002; de Castro, 1987, 2004, 2007; Kant & Graubard, 2014; Leech, Worsley, Timperio, & McNaughton, 2015; Lhuissier et al., 2013; Lund & Gronow, 2014; Meyer, 2002; Tani et al., 2015; Winkler et al., 1999).

2.1. Study design The NVS II is a representative study conducted in Germany between November 2005 and January 2007. Dietary intake was assessed with three dietary assessment methods (Heuer, Krems, Moon, Brombach, & Hoffmann, 2015). From a sub-sample of the NVS II, a 4-day dietary weighing record was conducted twice. Food and beverage consumption was protocolled in detail, including time of consumption. The participants were instructed in handling the weighing protocol in the study center and received a digital kitchen scale. Non-consumed residues had to be reweighed in order to obtain most accurate indication of the consumed quantities (Krems et al., 2006). Additionally, an excerpt of the EPIC-SOFT picture book (Slimani et al., 1999) could be used to estimate the consumed amount of foods eaten outside of the home (proportion about 25%). Three weekdays and one weekend day were included in each 4-day dietary weighing record to consider possible different dietary habits on weekdays and weekend days (Krems et al., 2006). Energy and macronutrient intake were calculated with the German nutrient database (BLS) version 3.02 (Hartmann, Heuer, & Hoffmann, 2015). Computer-assisted personal interviews (CAPI) were applied to obtain information on socio-demographic factors like sex, age, and SES. Anthropometric measurements (body height and body weight) were determined in a standardized way. Details of the NVS II have been previously reported (Heuer et al., 2015; MRI, 2008). The NVS II was approved by the German Federal Data Protection Office. Respondents were informed in detail about the study objectives, interview, examination procedures, and the handling of data records and analyses under pseudonymous conditions. It was made clear that participation was voluntary and could be terminated at any time. 2.2. Study population In the NVS II, 914 participants aged 18e80 years completed two 4-day dietary weighing records. Recorded days that were incomplete because time of consumption were not considered. Recorded

2.3. Periods

2.4. Selected population groups Four age groups were formed, pooling together 18e34 years, 35e50 years, 51e64 years, and 65e80 years. BMI was classified into underweight, normal weight, overweight and obese, according to the WHO definition (WHO, 2000). SES was defined as an index based on the “Winkler-Index” (Winkler & Stolzenberg, 1999) and encompassed the monthly net income of the household (nine categories, based on a monthly income from <750 V to 5000 V or higher), employment status of the household's principle earner (eight categories ranging from unskilled worker to executive employee/senior official), and school education level of the participant (five categories ranging from no qualification to baccalaureate; additional points were given for vocational training and university education). By using this index, the participants were classified to the low, medium, or high SES class (MRI, 2008). The combination of these three factors income of household, education level and employment status provides a comprehensive picture of the social status. The overall dietary quality was examined by the Healthy Eating Index of the NVS II (HEI-NVS II) (Wittig & Hoffmann, 2010). The consumed amounts of foods out of nine food groups (fruits, vegetables, bread, milk, fish, meat, eggs, fat, and non-alcoholic beverages) and alcohol as a nutrient were compared

F. Wittig et al. / Appetite 114 (2017) 125e136

with the dietary recommendations of the German Nutrition Society (DGE, 2013; 2014). For this, the ratio of food consumption and recommendations of each single food group/nutrient was converted into a score. Ten points were given if the consumed amount met the recommendation. The exceptions were the food groups fruit/fruit products and vegetables, where a maximum of 15 points were given. A maximum of 110 points calculated by summing the scores of each of the ten components could be achieved. According to the total index sum, women and men were classified in tertiles with a low, medium, and high HEI-NVS II sum. 2.5. Statistical analysis Energy and macronutrient intake were calculated per hour per day. For example, the energy intake at 10 o'clock comprises the energy intake from 9:30 o'clock to 10:29 o'clock as an arithmetic mean of all recorded days. After that, they were summarized for the defined periods ‘morning’, ‘midday’, ‘afternoon’, ‘evening’, and ‘night’. To take into account that the periods ‘morning’, ‘evening’, and ‘night’ encompass 6 h, whereas the periods ‘midday’ and ‘afternoon’ encompass 3 h, energy intake per average hour is also given for the defined periods. Energy and macronutrient intake are not normally distributed; thus, indication as median is appropriate. However, median values are often zero, as energy-providing foods and beverages were not consumed at every hour and at every defined period of the day by every person. Hence, arithmetic mean instead of median and standard error (SE) were chosen to show descriptive results of energy and macronutrient intake in the defined periods. A generalized estimating equation (GEE)-model was used to analyze whether the intake of energy and macronutrients over the course of the day differed between defined periods of selected population groups. Within these selected population groups, differences in energy intake at a specific period were analyzed with the GEE-model. The GEE-model was calculated with the ‘proc glimmix’ procedure by using an empirical variance estimator to request the covariance estimator for fixed effects. To fit the marginal model, the random residual statement, and to structure the covariance matrix, the compound symmetry (cs) statement was used. The dependent variable in the GEE-model was energy or macronutrient intake. Analyses were adjusted for age, BMI, and SES, defined as metric variables. To show differences between periods, the period ‘evening’ was used as the reference period in the presented tables. Because of sex-specific differences in food consumption (Heuer et al., 2015), separate analyses were performed for women and men. Statistically significant differences were defined at a p-level of
0.05. All statistical analyses were performed with the statistical software SAS version 9.3 (SAS Institute, Inc.).

127

Table 1 Characteristics of the study population. Women (n ¼ 372)

Men (n ¼ 290)

Age groups (%) 18e34 years 35e50 years 51e64 years 65e80 years

19.4 34.7 27.4 18.5

12.4 34.1 29.0 24.5

BMI groups (%) Underweight Normal weight Overweight Obese

1.6 52.4 30.4 15.6

0.0 31.0 52.8 16.2

SES groups (%) Low Medium High

14.8 25.8 59.4

15.9 25.2 58.9

HEI-NVS II sum groups (mean ± SE) Low 66.8 ± 0.5 Medium 77.4 ± 0.2 High 87.7 ± 0.4

62.3 ± 0.5 74.5 ± 0.3 84.9 ± 0.5

BMI, body mass index; SES, socioeconomic status; HEI-NVS II, Healthy Eating Index of the National Nutrition Survey II; SE, standard error.

seven days, with a mean of six days. More than half of the participants (59%) recorded seven days. The distribution of the recorded days was nearly equal over the week, with a few more records for Wednesdays (Table 2). 3.3. Energy intake

The study included 662 participants of the NVS II, 372 women (56.2%) and 290 men (43.8%) (Table 1). The age range was 18e80 years, with a mean age of 48.7 years for women and 51.5 years for men. The mean BMI for women was 25.3 kg/m2 and 26.8 kg/m2 for men. More than half of the women and men had a high SES. The mean HEI-NVS II sum was 77.3 points for women and 73.9 points for men.

Mean energy intake over the course of the day (total energy intake) was 1958 kcal for women and 2552 kcal for men (Table 3). Only small differences between women and men are shown in the distribution of energy intake over the course of the day (Fig. 1). Total energy intake as well as energy intake in all five periods (‘night’, ‘morning’, ‘midday’, ‘afternoon’, and ‘evening’) was higher for men than for women (p < 0.001). Three-main-meals, indicated as a peak in energy intake, were observed for women and men. Fig. 1 displays the three peaks at 08:00 to 09:00 o'clock, at 13:00 o'clock, and at 19:00 o'clock. A fourth, lower peak was observed in the ‘afternoon’ at 16:00 o'clock. For both sexes, the highest energy intake over the course of the day was in the ‘evening’, accounting for more than a third of the total energy intake, followed by the periods ‘morning’ and ‘midday’ each accounting for about a quarter, and ‘afternoon’ accounting for 14% of the total energy intake for women and 13% for men (Table 3). The lowest energy intake was during the ‘night’ period, accounting for less than 2% of the total energy intake. These unadjusted, descriptive results were confirmed by the results of the GEE-model (Table 4). On average, 95% of the energy intake was found between 07:00 o'clock and 20:00 o'clock for women and between 07:00 o'clock and 19:00 o'clock for men. For both sexes, 50% of the total energy intake could be seen before 14:00 o'clock, and 70% of the total energy intake could be seen before 18:00 o'clock. In contrast to the results of energy intake per period, the results of energy intake per hour of a period, considering the different length of the periods, showed the highest energy intake per hour in the period ‘midday’, followed by ‘evening’, ‘afternoon’, and ‘morning’. The lowest intake was in the period ‘night’ (Table 3).

3.2. Recorded days

3.4. Macronutrient intake

3. Results 3.1. Study population (general description)

The number of recorded days per person varies from one to

The intake distributions of carbohydrate, protein, and fat (Fig. 2)

128

F. Wittig et al. / Appetite 114 (2017) 125e136

Table 2 Distribution of the recorded days (weighing records).

Total (%) Women (%) Men (%)

Monday

Tuesday

Wednesday

Thursday

Friday

Saturday

Sunday

14.5 14.6 14.3

14.2 14.4 13.9

15.2 15.2 15.3

13.8 13.8 13.7

13.9 13.8 13.9

14.5 14.6 14.4

14.0 13.6 14.4

Table 3 Energy and macronutrient intake over the course of day of women and men; unadjusted, descriptive results. Energy intake (kcal) per period (mean ± SE) Women (n ¼ 372) Night (6 h) 13 ± 3 Morning 493 ± 11 (6 h) Midday 498 ± 11 (3 h) Afternoon 280 ± 8 (3 h) Evening 675 ± 13 (6 h) Entire day 1958 ± 20 (24 h) Men (n ¼ 290) Night (6 h) 38 ± 7 Morning 637 ± 15 (6 h) Midday 607 ± 15 (3 h) Afternoon 338 ± 14 (3 h) Evening 932 ± 19 (6 h) Entire day 2552 ± 29 (24 h)

Proportion of Energy intake (kcal) per hour energy (mean ± SE) intake per period (%)

Carbohydrate intake (g) per period (mean ± SE)

Proportion of carbohydrate intake per period (%)

Protein intake (g) per period (mean ± SE)

Proportion of Fat intake (g) per protein period intake per (mean ± SE) period (%)

Proportion of fat intake per period (%)

Alcohol intake (g) per period (mean ± SE)

Proportion of alcohol intake per period (%)

0.7 24.8

2±0 82 ± 2

2±0 66 ± 2

0.6 30.8

0±0 16 ± 0

0.4 24.5

0±0 17 ± 1

0.4 22.7

0±0 0±0

2.4 1.2

25.1

166 ± 4

51 ± 1

23.3

21 ± 1

29.7

23 ± 1

28.2

1±0

8.4

14.1

93 ± 3

34 ± 1

15.1

8±0

11.8

11 ± 0

14.4

1±0

7.2

34.0

113 ± 2

68 ± 2

30.2

24 ± 1

33.5

28 ± 1

34.3

7±1

80.7

100.0

82 ± 1

221 ± 1

100.0

68 ± 0

100.0

79 ± 0

100.0

8±0

100.0

1.5 25.0

6±1 106 ± 3

4±1 82 ± 2

1.3 30.0

1±0 20 ± 1

1.1 23.9

1±0 24 ± 1

1.1 24.5

1±0 0±0

3.4 2.3

23.8

203 ± 5

60 ± 2

22.1

26 ± 1

28.9

27 ± 1

26.1

2±0

10.3

13.2

113 ± 5

40 ± 2

14.0

10 ± 1

10.5

14 ± 1

12.5

2±0

8.6

36.5

155 ± 3

92 ± 2

32.7

32 ± 1

35.6

38 ± 1

35.8

13 ± 1

75.3

100.0

106 ± 1

276 ± 1

100.0

89 ± 0

100.0

104 ± 0

100.0

17 ± 0

100.0

BMI, body mass index; SES, socioeconomic status; HEI-NVS II, Healthy Eating Index of the National Nutrition Survey II; SE, standard error.

Fig. 1. Distribution of energy intake over the course of day of women and men; unadjusted descriptive results.

F. Wittig et al. / Appetite 114 (2017) 125e136

129

Table 4 Estimated energy intake over the course of day of women and men (GEE-model*). Women (n ¼ 372)

Night (6 h) Morning (6 h) Midday (3 h) Afternoon (3 h) Evening (6 h)

Men (n ¼ 290) Estimated differences in energy intake (kcal) compared to the ‘evening’ (mean ± SE)

pvalue

<0.001 37 ± 7 <0.001 637 ± 16

892 ± 20 292 ± 25

<0.001 <0.001

165 ± 12 385 ± 15

<0.001 610 ± 15 <0.001 329 ± 12

320 ± 24 600 ± 23

<0.001 <0.001

Ref.

e

Ref.

e

Estimated energy intake (kcal) per period (mean ± SE)

Estimated differences in energy intake (kcal) compared to the ‘evening’ (mean ± SE)

pvalue

11 ± 2 492 ± 10

654 ± 12 173 ± 16

500 ± 10 280 ± 8 665 ± 12

Estimated energy intake (kcal) per period (mean ± SE)

929 ± 19

* Adjusted for age, BMI, body mass index, and SES, socioeconomic status. SE, standard error; Ref., Reference period.

mirror the distribution of energy intake over the course of the day. For both women and men, intake of carbohydrates, protein, and fat exhibited peaks in the three-main-meal periods (‘morning’, ‘midday’, and ‘evening’) at the same point of time as the energy intake. The highest carbohydrate intake for women was observed in the ‘morning’ and ‘evening’ periods (p < 0.001). For men, the highest carbohydrate intake was found in the ‘evening’ period (p < 0.010), which was also the case for highest protein and fat intakes for both sexes (p < 0.001). Alcohol was predominantly consumed in the ‘evening’ period, for both sexes. Men had a higher intake of all macronutrients in all periods than women (p < 0.050), with the exception of alcohol intake in the ‘night’ period, in which no differences were seen. 3.5. Energy intake of selected population groups For the population groups defined by age, BMI, SES, and HEI-NVS II, a three-main-meal pattern was observed similar to the findings for the total groups of women and men. There were few differences in the distribution of energy intake over the course of the day between the defined periods and within the selected population groups at a specific period. Larger deviations of energy intake were found between young (18- to 34-years-old) and old adults (65- to 80-years-old) (Fig. 3), indicating that old adults had steeper peaks at the three-main-meal periods in the ‘morning’, ‘midday’, and ‘evening’, whereas young adults had lower peaks with a shift to later meal times. In most population groups of age, BMI, SES, and HEI-NVS II, the highest energy intake was observed in the ‘evening’ period (Table 5 and Table 6). As an exception, women aged 65 to 80 showed equal proportions of total energy intake in the ‘morning’, ‘midday’ and ‘evening’ periods, and women with a low SES had a similar energy intake in the ‘midday’ and ‘evening’ periods, with approximately 28 and 32% of the total energy intake. In all selected population groups, the proportion of total energy intake in the ‘evening’ period ranges from 30 to 37% for women and from 32 to 39% for men. In the ‘morning’ and ‘midday’ periods, a similar range of the proportion of total energy intake was observed of 23e28% (‘morning‘) and 24e29% (‘midday’) for women and 22e27% (‘morning’) and 21e27% (‘midday’) for men. The energy intake in the ‘afternoon’ period was lower and in a narrow range of 12e15% of the total energy intake, whereas in the ‘night’ period, 0e4% of the total energy intake was observed. As described before, the highest energy intake was observed in the ‘evening’ period. Comparing different age groups, the highest energy intake in the ‘evening’ period (36e39% of their total energy intake) was found in younger age groups (18e34 years and 35e50

years) for both sexes while the oldest age group (65e80 years) reached 30e32% (p < 0.001). Furthermore, in the period ‘night’, the oldest age group had the lowest energy intake for both sexes (p < 0.050). For men, no further differences between age groups were found. For women aged 65e80 years, a higher energy intake compared to 35- to 50-year-old women in the ‘midday’ period (p < 0.050) as well as a lower energy intake compared to 18- to 34year-old women in the ‘afternoon’ period (p < 0.050) were found. In the ’morning’ period, no differences were observed. Results in regard to different BMI groups showed that overweight women and men had a higher energy intake in the ‘evening’ period than normal weight participants (p < 0.050). This is also true for obese women, but not obese men. In all other periods, no differences between BMI groups were observed. With regard to the SES groups, the highest energy intake in the ‘evening’ period was observed for women and men with a high SES, and the lowest energy intake was found in participants with a low SES (p < 0.050). In the other periods, no differences between the SES groups could be observed. Comparing different HEI-NVS II groups, men with a low HEI-NVS II sum had a higher energy intake in the ‘evening’ than men with a high HEI-NVS II sum (p < 0.050). For women, no differences were detected between the HEI-NVS II groups in this period. In the ‘morning’ and ‘midday’ periods, differences between the HEI-NVS II groups were observed for women but not for men. Women with a higher HEI-NVS II sum showed a higher energy intake in these periods than women with a lower HEI-NVS II sum (p < 0.050). 4. Discussion In the presented study, intake of energy and macronutrients over the course of the day of German adults was analyzed on the basis of dietary weighing records conducted in the NVS II. For the total group of women and men as well as for the further population groups differentiated by age, BMI, SES, and HEI-NVS II, a threemain-meal pattern (‘morning’, ‘midday’, and ‘evening’) with the highest energy intake in the ‘evening’ period was observed. Women of the oldest age group did not show the highest energy intake in the ‘evening’, but depicted a similar distribution of energy intake across the three-main-meals. In contrast, young adults had lower peaks in the ‘morning’ and ’midday’ periods, with a shift to later meal times. Major differences in energy intake between the selected population groups were also found in comparison to the ‘evening’ period. The ‘morning’, ‘midday’, and ‘evening’ periods are crucial for the intake of energy and macronutrients over the course of the day, for both sexes. Therefore, the traditional three-main-meal pattern

130

F. Wittig et al. / Appetite 114 (2017) 125e136

Fig. 2. Distribution of macronutrient intake over the course of the day for women (a) and men (b); unadjusted, descriptive results.

seems to be valid for German adults at the time point of the NVS II (2005e2007). This result is in agreement with other European studies, such as those from Belgium (Mestdag, 2005), France (Bellisle et al., 2003; Lhuissier et al., 2013; Riou et al., 2015), Spain (Hermengildo et al., 2016), the Netherlands (Kearney, Hulshof, & €tta €la €, Gibney, 2001), Finland (Ovaskainen et al., 2010; Roos & Pra 1997) and other Nordic countries (Kjærnes, 2001; Lund & Gronow, 2014) as well as with previous studies from Germany (Winkler et al., 1999). Nevertheless, authors discuss the breaking up of the traditional meal pattern (examples: Lund & Gronow, 2014; Mestdag, 2005; Ovaskainen et al., 2006; Ovaskainen et al., 2010). A decline of energy intake at the three-main-meals (Kant & Graubard, 2015) and an increasing importance of snacks (Duffey et al., 2013; Kant & Graubard, 2015; Ovaskainen et al., 2010;

Piernas & Popkin, 2010; Popkin & Duffey, 2010; Zizza, Siega-Riz, & Popkin, 2001) were observed. In other studies, a change in meal patterns over the last several years was seen. However, the three-main-meal pattern is still the traditional pattern in these studies (Kant & Graubard, 2015; Mestdag & Glorieux, 2009; Ovaskainen et al., 2010), a fact that was confirmed by the results of the present study for the total group of women and men. In all periods, men had a higher energy intake than women. This was also €tta €la € (1997) and corresponds to men's observed by Roos and Pra higher energy requirements compared to women. The proportion of energy intake delivered by the three-mainmeal pattern differs between studies. In the presented analysis, the highest energy intake for women and men was observed in the ‘evening’ period, which accounted for more than a third of the total

F. Wittig et al. / Appetite 114 (2017) 125e136

131

Fig. 3. Distribution of energy intake over the course of the day for 18- to 34-year-old, 35- to 64-year old and 65- to 80-year-old women (a) and men (b); unadjusted, descriptive results.

energy intake, followed by the ‘morning’ and ‘midday’ periods in equal amounts, which confirms the results from a prior study in South Germany (Winkler et al., 1999). Similar results were found in the Netherlands (Kearney et al., 2001), and the USA (Ma et al., 2005; Wang et al., 2014). However, in Spain (Hermengildo et al., 2016; €ger, Marshall, & Dawson, 2009), Belgium (Mestdag, 2005) and Ja France (Bellisle et al., 2003) it was observed that the most important meal time was in the ‘midday’ period followed by the ‘evening’ period. The ‘morning’ period accounted for the lowest energy intake of the three-main-meal times. Conform to those results, results of the EPIC-study showed pronounced differences in meal patterns across European countries (Huseinovic et al., 2016). When comparing results of those different studies, it should be kept in mind that differences in meal patterns between countries may be explained by different cultural and economic circumstances, such as meal size and meal composition (de Graaf, 2000) as well as the

changing of the meal composition over time (Almoosawi, Winter, €kela €, 2000). In addition, Prynne, Hardy, & Stephen, 2012; Ma methodical reasons, such as different definitions of the periods in the studies, could also influence the results. The ‘evening’ meal dominates the energy intake over the course of the day for most selected population groups. The highest energy intake in the ‘evening’ period was found for young adults, overweight persons, obese women, and participants with a high SES, as well as for men with a low dietary quality, expressed as the HEINVS II sum. A higher energy intake in the ‘evening’ period of young adults compared to old adults was also shown in other studies (examples: Ma et al., 2005; Striegel-Moore & Franko, 2006). The less distinct meal patterns of younger persons, seen in the presented analysis, were also described by Meyer (2002) and Lund and Gronow (2014). To explain this, Lund and Gronow (2014) suggest that younger

132

F. Wittig et al. / Appetite 114 (2017) 125e136

Table 5 Energy intake over the course of day of selected population groups of women; unadjusted, descriptive results. Women (n ¼ 372)

Proportion of total Proportion of total Energy intake Proportion of total Energy intake Proportion of total Energy intake Energy intake (kcal) per period energy intake per (kcal) per period energy intake per (kcal) per period energy intake per (kcal) per period energy intake per period (%) (mean ± SE) period (%) (mean ± SE) period (%) (mean ± SE) period (%) (mean ± SE)

Age groups 18 to 34 years (n ¼ 72) Night (6 h) Morning (6 h) Midday (3 h) Afternoon (3 h) Evening (6 h) Entire day (24 h)

35 to 50 years (n ¼ 129)

51 to 64 years (n ¼ 102)

65 to 80 years (n ¼ 69)

22 ± 9 459 ± 24

1.1 22.9

14 ± 4 478 ± 19

0.7 24.4

13 ± 4 519 ± 18

0.6 26.3

0±0 517 ± 24

0.0 27.6

474 ± 22

23.6

475 ± 18

24.2

510 ± 22

25.8

546 ± 22

29.1

305 ± 19

15.2

284 ± 15

14.5

274 ± 16

13.8

257 ± 19

13.7

747 ± 33

37.2

711 ± 22

36.2

661 ± 23

33.4

552 ± 28

29.5

2007 ± 50

100.0

1962 ± 32

100.0

1977 ± 38

100.0

1873 ± 41

100.0

BMI groupsa

Normal weight (n ¼ 195)

Night (6 h) Morning (6 h) Midday (3 h) Afternoon (3 h) Evening (6 h) Entire day (24 h)

14 ± 4 476 ± 14 472 ± 13 267 ± 10 669 ± 17 1898 ± 27

SES groups

Low (n ¼ 55)

Night (6 h) Morning (6 h) Midday (3 h) Afternoon (3 h) Evening (6 h) Entire day (24 h)

10 ± 5 508 ± 38 533 ± 36 298 ± 23 614 ± 36 1963 ± 51

HEI-NVS II groups

Low (n ¼ 123)

Night (6 h) Morning (6 h) Midday (3 h) Afternoon (3 h) Evening (6 h) Entire day (24 h)

19 ± 4 454 ± 20 468 ± 20 272 ± 15 678 ± 24 1890 ± 33

Overweight (n ¼ 113) 0.8 25.2 25.0 14.2 35.2 100.0

12 ± 4 511 ± 22 522 ± 21 289 ± 17 661 ± 26 1995 ± 36

Obese (n ¼ 58) 0.6 25.9 26.4 14.6 33.1 100.0

Medium (n ¼ 96) 0.5 26.3 27.6 15.4 31.8 100.0

6±2 484 ± 19 502 ± 19 262 ± 15 646 ± 27 1900 ± 36

10 ± 5 479 ± 16 486 ± 17 295 ± 15 691 ± 25 1962 ± 33

0.5 25.0 25.6 14.2 35.1 100.0

High (n ¼ 221) 0.3 25.6 26.5 13.8 34.1 100.0

Medium (n ¼ 124) 1.0 24.0 24.8 14.4 35.9 100.0

10 ± 4 519 ± 24 533 ± 31 296 ± 22 730 ± 36 2087 ± 42

16 ± 4 493 ± 13 487 ± 13 284 ± 11 702 ± 17 1983 ± 26

0.8 25.0 24.7 14.4 35.6 100.0

High (n ¼ 125) 0.5 24.4 24.8 15.0 35.2 100.0

9±3 545 ± 18 539 ± 17 274 ± 14 656 ± 19 2022 ± 35

0.4 27.0 26.7 13.6 32.4 100.0

BMI, body mass index; SES, socioeconomic status; HEI-NVS II, Healthy Eating Index of the National Nutrition Survey II; SE, standard error. a Six women with underweight are not described in this table.

persons like singles have 'unsynchronized' eating patterns and are in a specifically situated life phase. When restrictions associated with influences from family life and work occurs, they pass the phase of 'unsynchronized' eating patterns. The less distinct meal patterns could also be due to a higher snacking frequency of young adults versus old adults, or a change in meal patterns in the last years. Leech, Worsley, Timperio and McNaughton (2017) found that women and men with a “grazing” (snacking) pattern were significantly younger. The grazing pattern was defined by frequent but no distinct peaks in probability of eating occasions. Ovaskainen et al. (2006) showed that for women, but not for men, a snackdominating meal pattern was more prevalent in young women compared to old women. In contrast to young adults, for older adults, meals play an important role for structuring the day (Brombach, 2001; Lund & Gronow, 2014; Meyer, 2002) and older adults still persist in their usual meal pattern (Schlettwein-Gsell, 1992). In the present study, women and men aged 65 to 80 showed distinct peaks at the three-main-meal periods. In particular, women aged 65 to 80 had a similar energy intake in these periods. Regular meal patterns of older persons were also observed in prior studies in Germany and other European countries (examples: Brombach, 2001; Lund & Gronow, 2014; Meyer, 2002; Schlettwein-Gsell, 1992). Overall, the results described above

showed that, in elderly people, the ‘evening’ meal did not dominate the energy intake over the course of the day resulting in a more regular meal pattern. In contrast, young adults showed a more irregular pattern with a very high energy intake in the ‘evening’ period. Studies suggest that the timing of energy intake over the course of the day is associated with total energy intake (examples: de Castro, 2004; Tani et al., 2015) and BMI (examples: Almoosawi, Vingeliene, Karagounis, & Pot, 2016; Bo et al., 2014; Wang et al., 2014). A higher energy intake in the ‘evening’ period is associated with a higher risk of being overweight/obese. These observations are in agreement with the presented results that overweight persons had a higher energy intake in the ‘evening’ period than normal weight persons. However, Yannakoulia et al. (2012) in Greece, €ssner (1996) in Sweden, and Kant et al. (1997) in Andersson and Ro the USA found no differences between normal weight and overweight adults concerning the energy intake at defined periods over the course of the day. de Castro (2004) concludes that obesity may result “… in part from the extension of the active period into the night when satiety mechanisms appear to be weak”. As a consequence, he suggests that large amounts of food in the ‘morning’ period and a restricted amount in the ‘evening’ period might reduce the total energy intake and the risk for obesity. Studies with

F. Wittig et al. / Appetite 114 (2017) 125e136

133

Table 6 Energy intake over the course of the day of selected population groups of men; unadjusted, descriptive results. Men (n ¼ 290)

Energy intake (kcal) per period (mean ± SE)

Proportion of total energy intake per period (%)

Age groups 18 to 34 years (n ¼ 36) Night (6 h) Morning (6 h) Midday (3 h) Afternoon (3 h) Evening (6 h) Entire day (24 h)

Energy intake (kcal) per period (mean ± SE)

Proportion of total energy intake per period (%)

35 to 50 years (n ¼ 99)

Energy intake (kcal) per period (mean ± SE)

Proportion of total energy intake per period (%)

51 to 64 years (n ¼ 84)

Energy intake (kcal) per period (mean ± SE)

Proportion of total energy intake per period (%)

65 to 80 years (n ¼ 71)

107 ± 35 600 ± 51

3.8 21.5

34 ± 10 656 ± 27

1.3 24.7

37 ± 10 624 ± 25

1.5 25.4

8±8 645 ± 31

0.4 26.9

590 ± 47

21.1

614 ± 25

23.1

586 ± 30

23.9

632 ± 26

26.4

404 ± 42

14.4

321 ± 21

12.1

328 ± 28

13.4

339 ± 24

14.2

1097 ± 62

39.2

1032 ± 34

38.8

880 ± 28

35.9

770 ± 31

32.2

2797 ± 97

100.0

2657 ± 53

100.0

2454 ± 46

100.0

2395 ± 50

100.0

BMI groupsa

Normal weight (n ¼ 90)

Night (6 h) Morning (6 h) Midday (3 h) Afternoon (3 h) Evening (6 h) Entire day (24 h)

56 ± 17 632 ± 26 593 ± 25 353 ± 22 917 ± 34 2549 ± 56

SES groups

Low (n ¼ 46)

Night (6 h) Morning (6 h) Midday (3 h) Afternoon (3 h) Evening (6 h) Entire day (24 h)

57 ± 24 610 ± 35 637 ± 34 296 ± 29 802 ± 37 2402 ± 57

HEI-NVS II groups

Low (n ¼ 96)

Night (6 h) Morning (6 h) Midday (3 h) Afternoon (3 h) Evening (6 h) Entire day (24 h)

56 ± 15 625 ± 28 572 ± 25 308 ± 20 987 ± 36 2548 ± 52

Overweight (n ¼ 153) 2.2 24.8 23.3 13.8 36.0 100.0

26 ± 6 657 ± 22 609 ± 21 334 ± 20 956 ± 27 2582 ± 40

Obese (n ¼ 47) 1.0 25.6 23.8 13.1 37.3 100.0

Medium (n ¼ 73) 2.4 25.4 26.5 12.3 33.4 100.0

31 ± 11 676 ± 32 593 ± 31 380 ± 31 904 ± 47 2584 ± 70

2.2 24.5 22.4 12.1 38.3 100.0

22 ± 6 627 ± 27 622 ± 24 343 ± 21 956 ± 33 2570 ± 52

41 ± 16 584 ± 38 629 ± 38 320 ± 30 881 ± 43 2456 ± 62

1.7 23.8 25.6 13.0 35.9 100.0

High (n ¼ 171) 1.2 26.4 23.1 14.8 35.2 100.0

35 ± 8 628 ± 20 606 ± 20 331 ± 17 979 ± 23 2578 ± 36

0.9 24.4 24.2 13.3 37.2 100.0

35 ± 11 660 ± 24 628 ± 27 362 ± 28 853 ± 29 2537 ± 49

Medium (n ¼ 97)

1.4 24.5 23.6 12.9 38.2 100.0

High (n ¼ 97) 1.4 26.0 24.8 14.3 33.6 100.0

BMI, body mass index; SES, socioeconomic status; HEI-NVS II, Healthy Eating Index of the National Nutrition Survey II; SE, standard error. a In the group of men, no person with underweight was identified.

a focus on defined meals instead of energy intake over the course of the day showed similar results. Skipping breakfast or a lower energy intake in the ‘morning’ as well as a higher meal frequency or energy intake in the ‘evening’ are associated with a higher BMI us Forslund et al., 2002; de (examples: Berg et al., 2009; Berte Castro, 2004, 2007; Fricker, Giroux, Fumeron, & Apfelbaum, 1990; Howarth, Huang, Roberts, Lin, & McCrory, 2007; Striegel-Moore & Franko, 2006; Tahara & Shibata, 2013). The presented findings suggest that a higher energy intake in the ‘evening’ period can be one risk factor for being obese. Most studies examining the association between energy intake over the course of the day and dietary quality focus on single food groups (examples: Ovaskainen et al., 2006; Roos & Pr€ att€ al€ a, 1997; Tani et al., 2015; Yannakoulia et al., 2012) or nutrients (example: €tta €la €, 1997). In the present study, a classification based on Roos & Pra the HEI-NVS II was used for comparing population groups with different dietary qualities. The results suggest that a higher energy intake earlier in the day is associated with a higher overall dietary quality. Similar results were found by Aljuraiban et al. (2015). Participants who consumed most of their energy earlier in the day had a higher (better) Nutrient Rich Food Index 9.3 (NRF9.3) value than those who consumed most of their foods later in the day. Other studies found either no association between the Healthy Eating

Index 2010 of the USDA (HEI-2010) and time of eating (Barnes, French, Harnack, Mitchell, & Wolfson, 2015), or the opposite, that persons with a high energy intake in the ‘evening’ period had a better dietary quality, which was also measured with an index, the Mediterranean Diet Adherence Screen (MEDAS) (Hermengildo et al., 2016). These inconsistent results may be explained with different indices, e.g. varying in the components and the point ratings. Therefore, additional studies to estimate the influence of the time of day on food consumption for dietary quality are necessary. Another population group with a high energy intake in the ‘evening’ period was the high SES group compared to the low SES group. This result confirms the result of Kant et al. (1997). Other studies found no association between energy intake in the ‘evening’ period and education (Almoosawi et al., 2013b; Hermengildo et al., 2016; Wang et al., 2014). Further analyses of the energy intake in the ‘evening’ period differentiated by educational level show different to the SES that a higher energy intake in the ‘evening’ period were only found for men with a high education compared to those with a low education level, while for women no differences in the ‘evening’ period was observed (data not presented). In regard to the presented results and the literature, it remains unclear whether SES or its components as well as associated factors like family

134

F. Wittig et al. / Appetite 114 (2017) 125e136

structure or starting time of work are important for energy intake over the course of day. Some authors described associations between the time of food consumption and health related factors. A higher energy intake in the ‘evening’ period was associated with a higher risk of high blood pressure (Almoosawi, Prynne, Hardy, & Stephen, 2013a), being overweight/obese (Almoosawi et al., 2016; Bo et al., 2014; Wang et al., 2014) and developing a metabolic syndrome (Bo et al., 2014). These findings combined with the observations of the present study, which show that energy intake was the highest in the ‘evening’ period, especially in young adults and overweight persons, lead to the conclusion that the ‘evening’ period is of particular interest to derive preventive dietary recommendations. More studies on meal patterns, such as energy intake over the course of the day in regard to health-related factors and how dietary patterns could be beneficial for health, are necessary. The strength of the present study is the use of a populationbased study allowing a detailed analysis of dietary consumption for selected population groups together with time-specific food consumption data. Studies investigating energy intake over the course of the day of different population groups in one study sample are scarce. To avoid undesirable influences of underreporters, these were identified and excluded from the analysis. A limitation of this study is that defined periods instead of meal occasions are used. The defined periods (‘morning’, ‘midday’, and ‘evening’) are not necessarily in accordance with real eating occasions of participants (breakfast, lunch, and dinner). As described by Leech et al. (2015), it is unclear whether the participants would have classified their meals in a similar way. Furthermore, a small sample size of some selected population groups may have limited the ability to detect differences within the population groups. The method of weighing records may have influenced the energy intake over the course of the day. For example, food consumed out of home may be less accurately reported with this method. Also, the recording process of weighing records can lead to change the usual eating habits (van Staveren et al., 2012). Nevertheless, weighing records provide the most detailed information about food consumption at each point in time over the course of the day. The NVS II was conducted in the years 2005e2007. Therefore, the presented results provide information regarding the meal pattern at that time. It might be possible that the described meal pattern in Germany have changed over the last years. A repetition of the analyses in the future with current data can reveal potential changes or stability over time. Further analyses should be conducted for a comparison of meal patterns in Germany to other European countries to depict possible differences between countries. 5. Conclusion The presented analyses provide comprehensive descriptions of meal patterns in regard to the distribution of energy intake over the course of the day of selected population groups in Germany. With few differences within the population groups defined by sex, age, BMI, SES, and HEI-NVS II, the traditional three-main-meal pattern was observed, a result which is also found in other studies. For old adults, meals have an important role for structuring the day as seen in distinct peaks at the three-main-meal periods. In contrast, young adults seem to have a higher variability in energy intake and a less distinct meal pattern. Further, the results show that the highest energy intake was observed in the ‘evening’ period, especially in young adults, overweight persons, and persons with a high SES, as well as men with a low dietary quality (expressed by HEI-NVS II). Because a high energy intake in the ‘evening’ period is associated with health-related factors, such as obesity, higher hypertension

prevalence, and a higher blood pressure in the literature, the distribution of energy intake over the course of the day should be considered by recommendations for the promotion of a healthy nutritional behavior. Funding The funding agency supporting this work is the German Federal Ministry of Education and Research (Funding number: 01EA1372D). The original survey (NVS II) was funded by the German Federal Ministry of Food, Agriculture, and Consumer Protection. The funding ministries were not involved in the study design, collection, analysis and interpretation of data, writing the article, or in the decision to submit the article for publication. Acknowledgment The preparation of this paper was supported by the DEterminants of DIet and Physical ACtivity (DEDIPAC) knowledge hub. This work is supported by the Joint Programming Initiative ‘Healthy Diet for a Healthy Life’. The authors thank the consortium “Populations Germany” in the framework of DEDIPAC for the constructive discussion. We also thank Dr. Alexander Roth from the Department of Physiology and Biochemistry of Nutrition, Max Rubner-Institut (present address: Department of Child and Adolescent Psychiatry and Psychotherapy, University of Zurich, Switzerland), for statistical consulting. We also thank Valentina Breitenstein for support of the literature research. Conflicts of interest: None of the authors has any conflict of interest. References Aljuraiban, G. S., Chan, Q., Oude Griep, L. M., Brown, I. J., Daviglus, M. L., et al. (2015). The impact of eating frequency and time of intake on nutrient quality and body mass index: The INTERMAP Study, a Population-Based Study. Journal of the Academy of Nutrition and Dietetics, 115, 528e536. e521. Almoosawi, S., Prynne, C. J., Hardy, R., & Stephen, A. M. (2013a). Time-of-day and nutrient composition of eating occasions: Prospective association with the metabolic syndrome in the 1946 British birth cohort. International Journal of Obesity, 37, 725e731. Almoosawi, S., Prynne, C. J., Hardy, R., & Stephen, A. M. (2013b). Time-of-day of energy intake: Association with hypertension and blood pressure 10 years later in the 1946 British birth cohort. Journal of Hypertension, 31, 882e892. Almoosawi, S., Vingeliene, S., Karagounis, L. G., & Pot, G. K. (2016). Chrono-nutrition: A review of current evidence from observational studies on global trends in time-of-day of energy intake and its association with obesity. Proceedings of the Nutrition Society, 1e14. Almoosawi, S., Winter, J., Prynne, C., Hardy, R., & Stephen, A. (2012). Daily profiles of energy and nutrient intakes: Are eating profiles changing over time? European Journal of Clinical Nutrition, 66, 678e686. € ssner, S. (1996). Meal patterns in obese and normal weight men: Andersson, I., & Ro The 'Gustaf' study. European Journal of Clinical Nutrition, 50, 639e646. Barnes, T. L., French, S. A., Harnack, L. J., Mitchell, N. R., & Wolfson, J. (2015). Snacking behaviors, diet quality, and body mass index in a community sample of working adults. Journal of the Academy of Nutrition and Dietetics, 115, 1117e1123. Bellisle, F. (2014). Meals and snacking, diet quality and energy balance. Physiology & Behavior, 134, 38e43. Bellisle, F., Dalix, A. M., Mennen, L., Galan, P., Hercberg, S., de Castro, et al. (2003). Contribution of snacks and meals in the diet of French adults: A diet-diary study. Physiology & Behavior, 79, 183e189. Berg, C., Lappas, G., Wolk, A., Strandhagen, E., Toren, K., Rosengren, A., et al. (2009). Eating patterns and portion size associated with obesity in a Swedish population. Appetite, 52, 21e26. us Forslund, H., Lindroos, A. K., Sjostrom, L., & Lissner, L. (2002). Meal patterns Berte and obesity in Swedish women - a simple instrument describing usual meal types, frequency and temporal distribution. European Journal Of Clinical Nutrition, 56, 740e747. Black, A. E. (2000). Critical evaluation of energy intake using the Goldberg cut-off for energy intake:basal metabolic rate. A practical guide to its calculation, use and limitations. International Journal of Obesity and Related Metabolic Disorders: Journal of the International Association for the Study of Obesity, 24, 1119e1130. Bo, S., Musso, G., Beccuti, G., Fadda, M., Fedele, D., Gambino, R., et al. (2014). Consuming more of daily caloric intake at dinner predisposes to obesity. A 6year population-based prospective cohort study. PLoS One, 9, 1e9. Brombach, C. (2001). The EVA-study: Meal patterns of women over 65 years. The

F. Wittig et al. / Appetite 114 (2017) 125e136 Journal of Nutrition, Health & Aging, 5, 263e265. de Castro, J. M. (1987). Circadian rhythms of the spontaneous meal pattern, macronutrient intake, and mood of humans. Physiology & Behavior, 40, 437e446. de Castro, J. M. (2004). The time of day of food intake influences overall intake in humans. Journal of Nutrition, 134, 104e111. de Castro, J. M. (2007). The time of day and the proportions of macronutrients eaten are related to total daily food intake. British Journal of Nutrition, 98, 1077e1083. Cutler, D. M., Glaeser, E. L., & Shapiro, J. M. (2003). Why have Americans become more obese? Journal of Economic Perspectives, 17, 93e118. DGE (German Nutrition Society). (2013). 10 guidelines of the German Nutrition Society (DGE) for a wholesome diet. 9th edition. Internet: https://www.dge.de/ index.php?id¼322 (accessed october 2016). €hrungsempfehDGE (German Nutrition Society). (2014). Lebensmittelbezogene Erna lungen in Deutschland [Food based recommendations in Germany]. Bonn. Internet: https://www.dge-medienservice.de/lebensmittelbezogeneernahrungsempfehlungen-in-deutschland.html. Duffey, K. J., Pereira, R. A., & Popkin, B. M. (2013). Prevalence and energy intake from snacking in Brazil: Analysis of the first nationwide individual survey. European Journal Of Clinical Nutrition, 67, 868e874. Fricker, J., Giroux, S., Fumeron, F., & Apfelbaum, M. (1990). Circadian rhythm of energy intake and corpulence status in adults. International Journal of Obesity, 14, 387e393. mez-Abella n, P., Alburquerque-Be jar, J. J., Lee, Y.-C., Ordova s, J. M., & Garaulet, M., Go Scheer, F. A. (2013). Timing of food intake predicts weight loss effectiveness. International Journal of Obesity : Journal of the International Association for the Study of Obesity, 37, 604e611. Goldberg, G. R., Black, A. E., Jebb, S. A., Cole, T. J., Murgatroyd, P. R., Coward, W. A., et al. (1991). Critical evaluation of energy intake data using fundamental principles of energy physiology: 1. Derivation of cut-off limits to identify underrecording. European Journal of Clinical Nutrition, 45, 569e581. de Graaf, C. (2000). Nutritional definitions of the meal. In H. L. Meiselman (Ed.), Dimensions of the meal: The science, culture, business, and art of eating (pp. 47e59). Gaithersburg: Aspen Publishers, Inc. Hartmann, B. M., Heuer, T., & Hoffmann, I. (2015). The German Nutrient Database: Effect of different versions on the calculated energy and nutrient intake of the German population. Journal of Food Composition and Analysis, 42, 26e29. Hermengildo, Y., Lopez-Garcia, E., Garcia-Esquinas, E., Perez-Tasigchana, R. F., Rodriguez-Artalejo, F., & Guallar-Castillon, P. (2016). Distribution of energy intake throughout the day and weight gain: A population-based cohort study in Spain. The British Journal of Nutrition, 1e8. Heuer, T., Krems, C., Moon, K., Brombach, C., & Hoffmann, I. (2015). Food consumption of adults in Germany: Results of the German national nutrition survey II based on diet history interviews. The British Journal of Nutrition, 113, 1603e1614. Howarth, N. C., Huang, T. T. K., Roberts, S. B., Lin, B. H., & McCrory, M. A. (2007). Eating patterns and dietary composition in relation to BMI in younger and older adults. International Journal of Obesity : Journal of the International Association for the Study Of obesity, 31, 675e684. Huseinovic, E., Winkvist, A., Slimani, N., Park, M. K., Freisling, H., Boeing, H., et al. (2016). Meal patterns across ten European countries - results from the European prospective investigation into cancer and nutrition (EPIC) calibration study. Public Health Nutrition, 19, 2769e2780. €ger, S. R., Marshall, D. W., & Dawson, J. (2009). A quantitative characterisation of Ja meals and their contexts in a sample of 25 to 49-year-old Spanish people. Appetite, 52, 318e327. Kant, A. K., & Graubard, B. I. (2014). Association of self-reported sleep duration with eating behaviors of American adults: NHANES 2005-2010. American Journal of Clinical Nutrition, 100, 938e947. Kant, A. K., & Graubard, B. I. (2015). 40-Year trends in meal and snack eating behaviors of American adults. Journal of the Academy of Nutrition and Dietetics, 115, 50e63. Kant, A. K., Schatzkin, A., & Ballard-Barbash, R. (1997). Evening eating and subsequent long-term weight change in a national cohort. International Journal of Obesity, 21, 407e412. Kearney, J., Hulshof, K., & Gibney, M. (2001). Eating patterns e temporal distribution, converging and diverging foods, meals eaten inside and outside of the home e implications for developing FBDG. Public Health Nutrition, 4, 693e698. Kjærnes, U. (2001). Eating patterns. A day in the lives of nordic people (Vol. 7). Lysaker, Norway: National Institute for Consumer Research. Internet: http:// www.sifo.no/files/file77520_rapport2001-07-web.pdf (02.12.2015). €tz, A., Heuer, T., Hild, A., Mo €seneder, J., et al. (2006). Krems, C., Bauch, A., Go Methoden der Nationalen Verzehrsstudie II [Methods of the National Nutrition €hrungs-Umschau, 53, 44e50. Survey II]. Erna Lakerveld, J., van der Ploeg, H., Kroeze, W., Ahrens, W., Allais, O., Andersen, L., et al. (2014). Towards the integration and development of a cross-European research network and infrastructure: The DEterminants of DIet and physical ACtivity (DEDIPAC) knowledge Hub. The International Journal of Behavioral Nutrition and Physical Activity, 11, 1e10. Leech, R. M., Worsley, A., Timperio, A., & McNaughton, S. A. (2015). Understanding meal patterns: Definitions, methodology and impact on nutrient intake and diet quality. Nutrition Research Reviews, 28, 1e21. Leech, R. M., Worsley, A., Timperio, A., & McNaughton, S. A. (2017). Temporal eating patterns: a latent class analysis approach. The International Journal of Behavioral Nutrition and Physical Activity, 14, 3.

135

Lhuissier, A., Tichit, C., Caillavet, F., Cardon, P., Masullo, A., Martin-Fernandez, J., et al. (2013). Who still eats three meals a day? Findings from a quantitative survey in the Paris area. Appetite, 63, 59e69. Lund, T. B., & Gronow, J. (2014). Destructuration or continuity? The daily rhythm of eating in Denmark, Finland, Norway and Sweden in 1997 and 2012. Appetite, 82, 143e153. Ma, Y., Bertone-Johnson, E. R., Stanek Iii, E. J., Reed, G.w., Herbert, J. R., Cohen, N. L., et al. (2005). Eating patterns in a free-living healthy U. S. Adult population. Ecology of Food and Nutrition, 44, 37e56. €kela €, J. (2000). Cultural definitions of the meal. In H. L. Meiselman (Ed.), DiMa mensions of the meal. The science, culture, business, and art of eating (pp. 7e15). Gaithersburg: Aspen Publishers, Inc. Mestdag, I. (2005). Disappearance of the traditional meal: Temporal, social and spatial destructuration. Appetite, 45, 62e74. Mestdag, I., & Glorieux, I. (2009). Change and stability in commensality patterns: A comparative analysis of Belgian time-use data from 1966, 1999 and 2004. The Sociological Review, 57, 703e726. Meyer, S. (2002). Mahlzeitenmuster in Deutschland [Meal pattern in Germany]]. Dissertation. Germany: Technical Universtity Muinch. MRI (Max Rubner-Institut). (2008). Nationale Verzehrsstudie II. Ergebnisbericht Teil 1 [National Nutrition Survey II. First report]. Karlsruhe. Internet: https://www.mri. bund.de/fileadmin/MRI/Institute/EV/NVS_II_Abschlussbericht_Teil_1_mit_ Ergaenzungsbericht.pdf (accessed october 2016). Müller, M. J., Bosy-Westphal, A., Klaus, S., Kreymann, G., Luhrmann, P. M., Neuhauser-Berthold, M., et al. (2004). World health organization equations have shortcomings for predicting resting energy expenditure in persons from a modern, affluent population: Generation of a new reference standard from a retrospective analysis of a German database of resting energy expenditure. The American Journal of Clinical Nutrition, 80, 1379e1390. Ovaskainen, M. L., Reinivuo, H., Tapanainen, H., Hannila, M. L., Korhonen, T., & Pakkala, H. (2006). Snacks as an element of energy intake and food consumption. European Journal of Clinical Nutrition, 60, 494e501. Ovaskainen, M. L., Tapanainen, H., & Pakkala, H. (2010). Changes in the contribution of snacks to the daily energy intake of Finnish adults. Appetite, 54, 623e626. Piernas, C., & Popkin, B. M. (2010). Snacking increased among U.S. adults between 1977 and 2006. Journal of Nutrition, 140, 325e332. Popkin, B. M., & Duffey, K. J. (2010). Does hunger and satiety drive eating anymore? Increasing eating occasions and decreasing time between eating occasions in the United States. American Journal of Clinical Nutrition, 91, 1342e1347. Riou, J., Lefevre, T., Parizot, I., Lhuissier, A., & Chauvin, P. (2015). Is there still a French eating model? A taxonomy of eating behaviors in adults living in the Paris metropolitan area in 2010 [electronic resource] PLoS One, 10. €la €, R. (1997). Meal pattern and nutrient intake among adult Finns. Roos, E., & Pr€ atta Appetite, 29, 11e24. Schlettwein-Gsell, D. (1992). Nutrition and the quality of life: A measure for the outcome of nutritional intervention.? American Journal of Clinical Nutrition, 55, 1263Se1266S. Slimani, N., Deharveng, G., Charrondiere, R. U., van Kappel, A. L., Ocke, M. C., et al. (1999). Structure of the standardized computerized 24-h diet recall interview used as reference method in the 22 centers participating in the EPIC project. Computer Methods and Programs in Biomedicine, 58, 251e266. , M. C., & de Vries, J. H. M. (2012). Estimation of dietary van Staveren, W. A., Ocke intake. In J. W. Erdman, I. A. Macdonald, & S. H. Zeisel (Eds.), Present knowledge in nutrition (10th ed, pp. 1012e1026). Ames: Wiley-Blackwell. Stok, M. F., Hoffmann, S., Volkert, D., Boeing, H., Ensenauer, R., Stelmach-Mardas, M., et al. (2017). The DONE framework: Creation, evaluation, and updating of an interdisciplinary, dynamic framework 2.0 of determinants of nutrition and eating. PLoS One, 12(2). Striegel-Moore, R., & Franko, D. (2006). Night Eating: Prevalence and demographic correlates. Obesity, 14, 139e174. Tahara, Y., & Shibata, S. (2013). Chronobiology and nutrition. Neuroscience, 253, 78e88. Tani, Y., Asakura, K., Sasaki, S., Hirota, N., Notsu, A., Todoriki, H., et al. (2015). Higher proportion of total and fat energy intake during the morning may reduce absolute intake of energy within the day. An observational study in free-living Japanese adults. Appetite, 92, 66e73. Wang, J. B., Patterson, R. E., Ang, A., Emond, J. A., Shetty, N., & Arab, L. (2014). Timing of energy intake during the day is associated with the risk of obesity in adults. Journal of Human Nutrition and Dietetics: The Official Journal of the British Dietetic Association, 27, 255e262. WHO (World Health Organization). (2000). Obesity: Preventing and managing the global epidemic. Report of a WHO consultation. WHO Technical Report Series 894, Geneva. Internet: http://www.who.int/nutrition/publications/obesity/WHO_ TRS_894/en/index.html (17.01.2012). Winkler, G., Doring, A., & Keil, U. (1999). Meal patterns in middle-aged men in southern Germany: Results from the MONICA Augsburg dietary survey 1984/85. Appetite, 32, 33e37. Winkler, J., & Stolzenberg, H. (1999). Der Sozialschichtindex im Bundes-Gesundheitssurvey [Social Status Scaling in the German National Health Interview and € Examination Survey]. Gesundheitswesen (Bundesverband der Arzte des € Offentlichen Gesundheitsdienstes) [Health Care Sector]., 61, S178eS183. €hrungsmuster von Bio-K€ Wittig, F., & Hoffmann, I. (2010). Erna aufern und Nicht-Bio€ufern [Dietary patterns of buyers and nonbuyers of organic food]. In Ka I. Hoffmann, & A. Spiller (Eds.), Auswertung der Daten der Nationalen Verzehrsstudie II (NVS II): Eine integrierte verhaltens- und lebensbasierte analyse des bio-

136

F. Wittig et al. / Appetite 114 (2017) 125e136

konsums [Data Interpretation Based on the German National Nutrition Survey II (NVS II): An Integrative Analysis of Behavioural and Lifestyle-Related Factors for Organic Food Consumption] (pp. 51e68). Germany: Max Rubner- Institut/Georg€ttingen. Karlsruhe/Go €ttingen. August-University Go Yannakoulia, M., Brussee, S. E., Drichoutis, A. C., Kalea, A. Z., Yiannakouris, N., Matalas, A. L., et al. (2012). Food consumption patterns in Mediterranean

adolescents: Are there differences between overweight and normal-weight adolescents? Journal of Nutrition Education and Behavior, 44, 233e239. Zizza, C., Siega-Riz, A. M., & Popkin, B. M. (2001). Significant increase in young adults' snacking between 1977-1978 and 1994-1996 represents a cause for concern! Preventive Medicine, 32, 303e310.