The association between night eating syndrome and food preferences among Chinese college students: a large-scale cross-sectional study

doi:10.21203/rs.3.rs-5322400/v1

Download PDF

Research Article

The association between night eating syndrome and food preferences among Chinese college students: a large-scale cross-sectional study

https://doi.org/10.21203/rs.3.rs-5322400/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background and aims:

Although the associations between NES and food preferences and the frequency of consumption of three meals a day have been examined among the European population, no studies have yet explored this association specifically among Chinese college students. Therefore, the purpose of this study was to investigate the associations between NES and food preferences and the frequency of consumption of three meals a day in Chinese college students.

Methods

This study used a cross-sectional design and included a sample of 11,856 Chinese university students, with an average age of 18.8 years (SD = 1.2 years). Food preferences were defined as five different food categories: fruits, vegetables, fast food (e.g. pizza, burgers, etc.), snacks (e.g. potato chips, crisps) and sugary beverages (e.g. soft drinks and sugary fruit drinks) and breakfast, lunch and dinner. The participants were invited to express their preferences for five food categories using a 5-point Likert scale, with the anchors "dislike very much" and "like very much". The frequency of consumption of three meals a day (including breakfast, lunch and dinner) was evaluated using a frequency scale ranging from 1 (once) to 7 (7 days per week). To evaluate the presence of Nocturnal Eating Syndrome (NES), we employed the Nocturnal Eating Questionnaire (NEQ) and the following criteria: no NES (total score < 25), mild NES (total score 25–30), and severe NES (total score > 30). Multivariate logistic regression analyses were used to examined the association between the level of NES and each food preference and the frequency of consumption of three meals a day.

Results

After controlling for potential confounding variables including gender, age, annual household income, place of residence, father's education and mother's education, multivariate logistic regression analyses observed a significant positive association between NES and the likelihood of exhibiting 1–6 food behaviors. We further found that, in addition to the fast food and dinner, there were significant positive associations between NES and remained food preferences (including, fruit, vegetable, snack and soft drinks) and two meals (including breakfast and lunch).

Conclusion

The results of this study suggest a strong positive association between NES and food preferences, and meal regularity. This implies that individuals with severe NES may exhibit poor food preferences and irregular meal habits. These findings emphasize the importance of intervening in NES, a condition that can assist individuals to modify their food preferences selection and adopt the habit of three meals a day. Future studies should employ longitudinal prospective or randomized controlled designs to confirm the above associations.

food preferences

night eating syndrome

Chinese

college students

cross-sectional study

In 1955, Stunkard first reported that patients with night eating syndrome (NES) exhibited circadian rhythm dysfunction[1], suggesting an imbalance between eating and sleep patterns[2]. The syndrome is defined as a daily intake of more than 25% of the recommended calorie intake after dinner, accompanied by nocturnal eating at least twice a week for a minimum of 3 months[3]. It is characterized by purging in the morning and hyperphagia in the evening (awakening to eat during the night), along with difficulties in both getting to sleep and maintaining sleep[1]. Furthermore, NES is associated not only with physical conditions such as obesity and diabetes [4, 5] but also with a range of other health issues including insomnia, depression, and anxiety[6]. These comorbidities can significantly affect overall well-being and quality of life. It is crucial to examine the association between NES and lifestyle, particularly dietary habits, to gain insights into the potential risks and consequences of NES.

Individuals with NES exhibit irregular nighttime eating habits, which can lead to the development of eating disorders that affect daytime eating as well and may result in an unbalanced diet. Previous studies based on patient populations and healthy adults have demonstrated a significant correlation between NES and a single dietary preference. For example, a study of obese individuals in Pakistan revealed a positive correlation between nighttime eating disorders and the frequency of daily snacking[7]. Another study in adults revealed a positive correlation between NES and unhealthy dietary habits such as the consumption of sugary beverages and fast food and a negative correlation between NES and healthy eating habits such as the consumption of fruits and vegetables[8]. Furthermore, the association between nighttime eating disorders and breakfast frequency has been explored in studies involving Bangladeshi university students, Turkish pregnant women, and patients with type 2 diabetes in the United States. The results of these studies showed a statistically significant inverse correlation between NES and the frequency of breakfast intake[5, 9, 10]. Furthermore, the association between NES and lunch and dinner consumption was examined in a study involving patients with type 2 diabetes in the United States. The results indicated no significant association between NES and the consumption of lunch or dinner[5].

A significant proportion of university students engage in unhealthy eating patterns at night. Previous research indicates that 57.10% of university students regularly consume late-night snacks, with 6.79% eating late-night snacks every night, 23.15% eating late-night snacks frequently, and 27.16% eating late-night snacks occasionally[11]. The university years represent a critical period in the transition from adolescence to adulthood[12]. It is also a critical period for the development of long-term eating habits and the influence of short-term and chronic diseases. Therefore, it is evident that further exploration of the association between nighttime eating and college students’ eating habits will play an important role in determining the factors involved. Despite several studies investigating the association between the two, no relevant studies on this topic have been conducted among Chinese college students.

There is a scarcity of research in the literature investigating the associations between NES and food preferences and between NES and meal regularity in Chinese populations, particularly college students. Accordingly, this study used a cross-sectional design to examine the association between NES and multiple food preferences, as well as the association between NES and a single dietary preference. The objectives were twofold: first, to investigate the association between NES and food preferences, and second, to develop recommendations for enhancing the dietary habits of college students. This study provides guidance for the identification of appropriate interventions and target populations when implementing healthy diet intervention programs among Chinese college students.

2.1 Participants

This study used an ongoing cohort design with the objective of identifying protective and risk factors for physical fitness levels among a population of Chinese university students. The survey locations for this study were selected based on the physical fitness testing centers of 11 universities in China, including six in Liaoning Province, one in Jilin City, and four in the municipality Chongqing. The selected universities offer a diverse range of academic disciplines, including philosophy, logic, economics, finance, economics and trade, political science, sociology, preschool education, elementary education, Chinese language and literature, and foreign languages and literature.

One and two grades were sampled from each of the 33 universities using a combination of convenience and cluster sampling. Finally, 11,856 university students were included in the study. Of the participants, 3,934 were male (33.2%) and 7,922 were female (66.8%). The mean age was 18.84 ± 1.2 years. The mean age of the male participants was 18.96 ± 1.2 years, while the mean age of female participants was 18.78 ± 1.1 years.

The participants were informed of the purpose, significance, and precautions of the study in detail before providing written informed consent. For those younger than 16 years, parental consent was obtained. This study was approved by the ethics committee of the School of Physical Education and Sports at Southwest University.

2.2 Assessment of food preference

All participants were requested to indicate their preferences for five different food categories: fruits, vegetables, fast food (e.g. pizza, burgers, etc.), snacks (e.g. potato chips, crisps) and sugary beverages (e.g. soft drinks and sugary fruit drinks)[13]. Moreover, the frequency of daily meals consumed (breakfast, lunch, and dinner) was recorded. For the five food preferences (fruit, vegetables, fast food, snacks, and beverages), respondents were required to indicate their level of preference for each item on a five-point Likert scale, with the following response options: (1) dislike very much, (2) dislike, (3) neutral, (4) like, and (5) like very much. For fruits and vegetables, a score of 1 was assigned to participants who selected "dislike very much," "dislike," or "neutral," while a score of 0 was attributed to those who selected "like" or "like very much." For the categories of fast food, snacks, and beverages, respondents who indicated a preference for these items by selecting "like" or "like very much" were assigned a score of 1, while those who selected "dislike very much," "dislike," or "neutral" were assigned a score of 0. The total score, which ranges between 0 and 8, is derived by summing all food preference items. Higher scores indicate a greater prevalence of unhealthy food preferences. Table 1 provides a list of questions and options regarding food preferences and three meals a day.

Table 1

Dietary preference questionnaire.
Do you like to eat fruits༟
very unlikely	unlikely	neutrality	likely	very likely
Do you like to eat vegetables༟
very unlikely	unlikely	neutrality	likely	very likely
Do you like to eat fast food (like pizza,hamburgers, etc.)༟
very unlikely	unlikely	neutrality	likely	very likely
Do you like to eat salty snack food (like potato chips, etc.)༟
very unlikely	unlikely	neutrality	likely	very likely
Do you like to eat soft drinks and sugared fruit drinks༟
very unlikely	unlikely	neutrality	likely	very likely
How many times a week do you have breakfast?
never	1	2	3	4	5	6	7
How many times a week do you have lunch?
never	1	2	3	4	5	6	7
How many times a week do you have dinner?
never	1	2	3	4	5	6	7

2.3 Assessment of night eating syndrome

In this study, the Night Eating Questionnaire (NEQ) was used to assess NES. Previous studies have demonstrated that NEQ exhibits good reliability and validity[14]. The questionnaire assesses the four principal characteristics of the syndrome, including the consumption of food during the night, binge eating, purging behaviour in the morning, mood, and disturbances of the sleeping pattern. The Likert scale is used to assess the severity of NES, with each question assigned a score of 0, 1, 2, 3, or 4. The exclusion of the score for question 13 from the total score calculation results in a total score range of 0 to 52 points. A high total score is indicative of a more severe NES.

According to the prior criteria for distinguishing the severity of NES, this study classified NES as three different groups: normal group (0 ~ 24 points), mild NES group (25 ~ 30 points) and severe NES group (31 ~ 52 points). To evaluate the internal consistency of the NEQ score, the Cronbach alpha coefficient was employed, yielding a value of 0.711, thereby indicating good internal consistency.

2.4 Assessment of covariates

All participants were asked to complete a self-designed questionnaire that assessed several demographic variables, including sex, age, annual household income (divided into three categories: less than RMB20,000, RMB20,000 to RMB35,000, and greater than RMB35,000), place of residence (classified as either urban or rural), and parental education level (divided into five categories: no formal education, completed primary school, completed junior high school, completed high school, or completed college or above).

2.5 Statistical analysis

The data were imported into an Excel database and statistically analyzed using SPSS 23.0. Prior to statistical analysis, all continuous variables were evaluated for normality using the Kolmogorov-Smirnov test. In cases where continuous variables exhibited a normal distribution, count data were expressed as mean ± standard deviation. In contrast, variables exhibiting a non-normal distribution were represented by the median with the corresponding interquartile range, and categorical variables were represented as percentages (%).

To examine the association between NES and dietary preference, NES was categorized as normal, mild, or severe and used as the independent variable, with dietary preference as the dependent variable. Multivariate logistic regression analyses were used to examine the association between the level of NES and each single food preference. The odds ratios (ORs) (95% confidence intervals [CIs]) of the likelihood of dietary preference subgroups was calculated for the mild NES group and severe NES group populations using normal group as the reference group. To explore the effect of confounders, two statistical models were constructed for this study, model 1 being the crude model and model 2 corrected for sex, age, annual household income, place of residence, father’s education level, and mother’s education level. All statistical analyses were performed using two-sided tests, with P < 0.05 indicating statistical significance.

Table 2 illustrates the characteristics of the participants. The study included a total of 11,856 college students, of whom 33.2% were male and the mean age was 18.84 years. In terms of socioeconomic status, more than one-third (34.6%) reported a household income of more than $35,000 per year, while 58.9% resided in rural areas. Approximately one-third of students had parents who had obtained a bachelor's degree or higher. Table 2 also provides a comparison of the various levels of NES categorisation with participant characteristics. The categories exhibiting higher levels of NES were observed to have a relatively higher mean age (P < 0.001) and a greater proportion of males (P < 0.001). In addition, individuals in these categories demonstrated a higher mean annual household income (P = 0.002).

Table 2

Demography characteristics of participants, according to night eating syndrome groups.
	Night eating syndrome groups
	Total(11856)	Normal(10883)	Mild (816)	Severe (157)	p
Age(Mean ± SD)	18.84 ± 1.2	18.81 ± 1.1	19.15 ± 1.3	19.01 ± 1.2	< 0.001
Sex, %					< 0.001
Male	33.3	32.4	42.2	42.2
Female	66.8	67.6	57.8	59.9
Annual family income, %					0.002
<20000	35.5	35.2	38.1	48.4
20000–35000	29.9	30.1	27.1	27.4
>35000	34.6	34.7	34.8	24.2
Place of residence, %					0.195
Town	41.1	40.9	44.1	41.4
Rural area	58.9	59.1	55.9	58.6
Father’s education, %					0.537
primary and below	23.9	24.1	21.7	20.4
Junior high	39.5	39.4	40.6	42.7
High school	21.4	21.4	21.7	19.1
Junior college and above	15.1	15.0	15.0	17.8
Mather’s education, %					0.147
primary and below	31.9	32.0	31.5	28.7
Junior high	36.4	36.6	34.8	31.2
High school	18.6	18.3	21.1	24.8
Junior college and above	13.1	13.1	12.6	15.3

Multivariate logistic regression was used to examine the association between NES and each food preferences, with NES as the independent variable and adherence to each food preferences as the dependent variable.

For fruits preference, the results in Table 3 show a significant association between NES and the likelihood of unhealthy fruits preference. The ORs (95% CIs) of NES compared to the reference normal group for mild NES group and severe NES group respectively were 4.007 (3.467, 4.632) and 2.198 (1.589, 3.039) (P < 0.001). After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to mild NES group and severe NES group respectively in comparison to the reference normal group were 3.718 (3.209, 4.308) and 2.027 (1.458, 2.819) (P < 0.001).

Table 3

Multivariate logistic regression of fruit consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	4.007 (3.467, 4.632) ^c	3.718 (3.209, 4.308)
Group 3 (severe )	157	2.198 (1.589, 3.039)	2.027 (1.458, 2.819)
P for trend	—	< 0.001	< 0.001
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For vegetables preference, the results in Table 4 show a significant association between NES and the likelihood of unhealthy vegetables preference. The ORs (95% CIs) of NES compared to the reference normal group for mild NES group and severe NES group respectively were 2.908 (2.510, 3.368) and 1.596 (1.164, 2.188) (P < 0.001). After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to mild NES group and severe NES group respectively in comparison to the reference normal group were 2.796 (2.411, 3.242) and 1.535 (1.117, 2.108) (P < 0.001).

Table 4

Multivariate logistic regression of vegetable consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	2.908 (2.510, 3.368) ^c	2.796 (2.411, 3.242)
Group 3 (severe )	157	1.596 (1.164, 2.188)	1.535 (1.117, 2.108)
P for trend	—	< 0.001	< 0.001
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For fast food preference, the results in Table 5 show that there was no significant association between NES and the likelihood of unhealthy fast food preference. The ORs (95% CIs) of NES compared to the reference normal group for mild NES group and severe NES group respectively were 0.667 (0.562, 0.791) and 2.086 (1.521, 2.861) (P = 0.723).After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to mild NES group and severe NES group respectively in comparison to the reference normal group were 0.701 (0.589, 0.833) and 2.293 (1.660, 3.167) (P = 0.644).

Table 5

Multivariate logistic regression of fast food consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	0.667 (0.562, 0.791) ^c	0.701 (0.589, 0.833)
Group 3 (severe )	157	2.086 (1.521, 2.861)	2.293 (1.660, 3.167)
P for trend	—	0.723	0.644
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For snack preference, the results in Table 6 show a significant association between NES and the likelihood of unhealthy snack preference. The ORs (95% CIs) of NES compared to the reference normal group for mild NES group and severe NES group respectively were 0.572 (0.483, 0.677) and 1.343 (0.975, 1.851) (P < 0.001).After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to mild NES group and severe NES group respectively in comparison to the reference normal group were 0.610 (0.514, 0.724) and 1.473 (1.062, 2.043) (P = 0.009).

Table 6

Multivariate logistic regression of snack consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	0.572 (0.483, 0.677) ^c	0.610 (0.514, 0.724)
Group 3 (severe )	157	1.343 (0.975, 1.851)	1.473 (1.062, 2.043)
P for trend	—	< 0.001	0.009
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For soft drinks preference, the results in Table 7 show a significant association between NES and the likelihood of unhealthy soft drinks preference. The ORs (95% CIs) of NES compared to the reference normal group for mild NES group and severe NES group respectively were 0.629 (0.539, 0.734) and 1.318 (0.961, 1.806) (P = 0.001).After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to mild NES group and severe NES group respectively in comparison to the reference normal group were 0.647 (0.554, 0.755) and 1.371 (0.999, 1.882) (P = 0.006).

Table 7

Multivariate logistic regression of soft drinks consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	0.629 (0.539, 0.734) ^c	0.647 (0.554, 0.755)
Group 3 (severe )	157	1.318 (0.961, 1.806)	1.371 (0.999, 1.882)
P for trend	—	0.001	0.006
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For breakfast preference, the results in Table 8 show a significant association between NES and the likelihood of irregular breakfast. The ORs (95% CIs) of NES compared to the reference normal group for mild NES group and severe NES group respectively were 1.498 (1.288, 1.742) and 1.217 (0.879, 1.684) (P < 0.001).After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to mild NES group and severe NES group respectively in comparison to the reference normal group were 1.473 (1.265, 1.714) and 1.232 (0.889, 1.708) (P < 0.001).

Table 8

Multivariate logistic regression of breakfast consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	1.498 (1.288, 1.742) ^c	1.473 (1.265, 1.714)
Group 3 (severe )	157	1.217 (0.879, 1.684)	1.232 (0.889, 1.708)
P for trend	—	< 0.001	< 0.001
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For lunch preference, the results in Table 9 show a significant association between NES and the likelihood of irregular lunch. The ORs (95% CIs) of NES compared to the reference normal group for mild NES group and severe NES group respectively were 1.338 (1.133, 1.579) and 1.644 (1.160, 2.331) (P < 0.001).After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to mild NES group and severe NES group respectively in comparison to the reference normal group were 1.377 (1.164, 1.629) and 1.662 (1.169, 2.364) (P < 0.001).

Table 9

Multivariate logistic regression of lunch consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	1.338 (1.133, 1.579) ^c	1.377 (1.164, 1.629)
Group 3 (severe )	157	1.644 (1.160, 2.331)	1.662 (1.169, 2.364)
P for trend	—	< 0.001	< 0.001
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For dinner preference, the results in Table 10 show that there was no significant association between NES and the likelihood of irregular dinner. The ORs (95% CIs) of NES compared to the reference normal group for mild NES group and severe NES group respectively were 1.009 (0.861, 1.183) and 1.000 (0.702, 1.424) (P = 0.935).After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to mild NES group and severe NES group respectively in comparison to the reference normal group were 1.101 (0.936, 1.296) and 1.056 (0.737, 1.515) (P = 0.297).

Table 10

Multivariate logistic regression of dinner consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	1.009 (0.861, 1.183) ^c	1.101 (0.936, 1.296)
Group 3 (severe )	157	1.000 (0.702, 1.424)	1.056 (0.737, 1.515)
P for trend	—	0.935	0.297
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

This study also examined the association between NES and the likelihood of food preferences diversity. For adherence to 1–7 different food preferences diversity, the results showed that there was significant association between NES and adherence to 1–6 food preferences diversity.

For adherence to 1 or more food preferences diversity, the results showed that there was a significant association between the NES and the likelihood of adherence to 1 or more food preferences diversity. In crude model, the ORs (95% CIs) for adherence to the likelihood of 1 or more food preferences diversity, when compared to the reference normal group, were 2.881 (1.990, 4.170) and 3.342 (1.368, 8.164) for mild NES group and severe NES group, respectively (P < 0.001); After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to 1 or more food preferences diversity in comparison to the reference normal group were 2.859 (1.973, 4.142) and 3.450 (1.411, 8.435) for mild NES group and severe NES group, respectively (P < 0.001). (Table 11)

Table 11

Multivariate logistic regression of unhealthy eating behaviors consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Unhealthy eating behaviors (≥ 1 vs < 1)
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	2.881 (1.990, 4.170) ^c	2.859 (1.973, 4.142)
Group 3 (severe )	157	3.342 (1.368, 8.164)	3.450 (1.411, 8.435)
P for trend	—	< 0.001	< 0.001
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For adherence to 2 or more food preferences diversity, the results showed that there was a significant association between the NES and the likelihood of adherence to 2 or more food preferences diversity. In crude model, the ORs (95% CIs) for adherence to the likelihood of 2 or more food preferences diversity, when compared to the reference normal group, were 2.857 (2.270, 3.596) and 3.961 (2.195, 7.149) for mild NES group and severe NES group, respectively (P < 0.001); After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to 1 or more different food preferences in comparison to the reference normal group were 2.878 (2.285, 3.625) and 4.052 (2.243, 7.319) for mild NES group and severe NES group, respectively (P < 0.001).(Table 12)

Table 12

Multivariate logistic regression of unhealthy eating behaviors consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Unhealthy eating behaviors (≥ 2 vs < 2)
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	2.857 (2.270, 3.596) ^c	2.878 (2.285, 3.625)
Group 3 (severe )	157	3.961 (2.195, 7.149)	4.052 (2.243, 7.319)
P for trend	—	< 0.001	< 0.001
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For adherence to 3 or more food preferences diversity, the results showed that there was a significant association between the NES and the likelihood of adherence to 3 or more food preferences diversity. In crude model, the ORs (95% CIs) for adherence to the likelihood of 3 or more food preferences diversity, when compared to the reference normal group, were 1.818 (1.562, 2.116) and 2.354 (1.650, 3.385) for mild NES group and severe NES group, respectively (P < 0.001); After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to 1 or more different food preferences in comparison to the reference normal group were 1.862 (1.597, 2.170) and 2.442 (1.707, 3.493) for mild NES group and severe NES group, respectively (P < 0.001).(Table 13)

Table 13

Multivariate logistic regression of unhealthy eating behaviors consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Unhealthy eating behaviors (≥ 3 vs < 3)
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	1.818 (1.562, 2.116) ^c	1.862 (1.597, 2.170)
Group 3 (severe )	157	2.354 (1.650, 3.385)	2.442 (1.707, 3.493)
P for trend	—	< 0.001	< 0.001
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For adherence to 4 or more food preferences diversity, the results showed that there was a significant association between the NES and the likelihood of adherence to 4 or more food preferences diversity. In crude model, the ORs (95% CIs) for adherence to the likelihood of 4 or more food preferences diversity, when compared to the reference normal group, were 1.370 (1.182, 1.588) and 1.799 (1.310, 2.473) for mild NES group and severe NES group, respectively (P < 0.001); After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to 1 or more different food preferences in comparison to the reference normal group were 1.427 (1.229, 1.658) and 1.882 (1.365, 2.595) for mild NES group and severe NES group, respectively (P < 0.001).(Table 14)

Table 14

Multivariate logistic regression of unhealthy eating behaviors consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Unhealthy eating behaviors (≥ 4 vs < 4)
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	1.370 (1.182, 1.588) ^c	1.427 (1.229, 1.658)
Group 3 (severe )	157	1.799 (1.310, 2.473)	1.882 (1.365, 2.595)
P for trend	—	< 0.001	< 0.001
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For adherence to 5 or more food preferences diversity, the results showed that there was a significant association between the NES and the likelihood of adherence to 5 or more food preferences diversity. In crude model, the ORs (95% CIs) for adherence to the likelihood of 5 or more food preferences diversity, when compared to the reference normal group, were 1.135 (0.932, 1.383) and 1.784 (1.220, 2.607) for mild NES group and severe NES group, respectively (P = 0.004); After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to 1 or more different food preferences in comparison to the reference normal group were 1.185 (0.971, 1.446) and 1.844 (1.257, 2.705) for mild NES group and severe NES group, respectively (P = 0.001).(Table 15)

Table 15

Multivariate logistic regression of unhealthy eating behaviors consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Unhealthy eating behaviors (≥ 5 vs < 5)
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	1.135 (0.932, 1.383) ^c	1.185 (0.971, 1.446)
Group 3 (severe )	157	1.784 (1.220, 2.607)	1.844 (1.257, 2.705)
P for trend	—	0.004	0.001
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For adherence to 6 or more food preferences diversity, the results showed that there was a significant association between the NES and the likelihood of adherence to 6 or more food preferences diversity. In crude model, the ORs (95% CIs) for adherence to the likelihood of 6 or more food preferences diversity, when compared to the reference normal group, were 0.772 (0.520, 1.145) and 2.739 (1.642, 4.570) for mild NES group and severe NES group, respectively (P = 0.085); After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to 1 or more different food preferences in comparison to the reference normal group were 0.823 (0.553, 1.223) and 2.899 (1.730, 4.857) for mild NES group and severe NES group, respectively (P = 0.035).(Table 16)

Table 16

Multivariate logistic regression of unhealthy eating behaviors consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Unhealthy eating behaviors (≥ 6 vs < 6)
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	0.772 (0.520, 1.145) ^c	0.823 (0.553, 1.223)
Group 3 (severe )	157	2.739 (1.642, 4.570)	2.899 (1.730, 4.857)
P for trend	—	0.085	0.035
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

For adherence to 7 or more food preferences diversity, the results showed that there was a significant association between the NES and the likelihood of adherence to 7 or more food preferences diversity. In crude model, the ORs (95% CIs) for adherence to the likelihood of 7 or more food preferences diversity, when compared to the reference normal group, were 0.854 (0.345, 2.115) and 1.787 (0.435, 7.339) for mild NES group and severe NES group, respectively (P = 0.771); After adjusting for potential considerable confounders, the adjusted ORs (95% CIs) for adherence to 1 or more different food preferences in comparison to the reference normal group were 0.934 (0.376, 2.320) and 1.928 (0.468, 7.948) for mild NES group and severe NES group, respectively (P = 0.614).(Table 17)

Table 17

Multivariate logistic regression of unhealthy eating behaviors consumption according to night eating syndrome.
Night eating syndrome	Number of participants	Model 1 ^a	Model 2 ^b
Unhealthy eating behaviors (≥ 7 vs < 7)
Group 1 (normal)	10883	1.000 (reference)	1.000 (reference)
Group 2 (mild)	816	0.854 (0.345, 2.115) ^c	0.934 (0.376, 2.320)
Group 3 (severe )	157	1.787 (0.435, 7.339)	1.928 (0.468, 7.948)
P for trend	—	0.771	0.614
^a Model 1: crude
^b Model 2: Adjusted for sex, age, annual family income (< 20000, 20000–35000, >35000 yuan), place of residence (town or rural), father and mother’s education (primary and below, junior high, high school, or junior college and above).
^c Adjusted data are expressed as odds ratio (95% confidence intervals).

This is the first study to examine the associations between NES and food preferences and NES and meal frequency in college students. The findings indicate that individuals with NES exhibited a significant tendency to select high-calorie foods, such as sugary beverages and snacks, over vegetables and fruits, compared to individuals without NES. Conversely, individuals without NES showed a significant tendency to adhere to regular patterns of breakfast and lunch consumption compared to those with NES.

To date, studies on the negative effect of NES on food preference have primarily focused on investigating the association between NES and a single food preference[7, 8] or the frequency of three meals per day[5, 9, 10]. The findings of the present study support and are consistent with those of previous investigations. In comparison to previous studies, on the one hand, the present study used a representative sample of Chinese university students to comprehensively examine the association between NES and a range of food preferences and meal frequency. For the first time, this study examined the association between NES and the diversity of food preferences and the frequency of consuming three-meals per day.

Regarding the potential mechanisms underlying the association between NES and food preferences, it is plausible that endogenous circadian rhythms mediate this effect. Hunger exhibits an obvious circadian rhythm. The circadian peak in hunger was observed in the biological evening, corresponding to 8 pm, whereas the circadian trough in hunger occurred at 8 am. With respect to food preferences, the influence of circadian rhythms on appetite for specific food groups was more pronounced in the evening. This period is associated with an increased desire for high-energy foods, such as sweets, salty snacks, carbohydrates, meat, poultry, and fruits. In contrast, there is less evidence of a circadian influence on appetite for low-energy foods such as vegetables[15]. In light of the above, it seems reasonable to hypothesize that individuals with NES, characterized by nocturnal sleep, may exhibit a heightened evening appetite driven by the inherent circadian rhythm, which could potentially contribute to the consumption of more food, particularly high-energy food, in the period preceding the fasting required for adequate sleep.

It is also possible that changes in appetite-related hormones such as leptin play a role in the association between NES and food preference. It is widely acknowledged that sleep suppresses appetite by prompting the release of leptin, a hormone produced by adipocytes[16, 17]. Furthermore, this suppression persists during the morning wake state. In contrast, NES is characterized by nighttime and nocturnal hyperphagia, morning purging, and insomnia, accompanied by a reduction in leptin levels during nocturnal sleep periods. This may result in an increased sensation of hunger and subsequent food intake[18], as evidenced by a cross-sectional study that identified a significant positive association between NES and leptin levels[19].

Notably, this study also identified a significant association between severe NES and reduced frequency of breakfast and lunch consumption. The schedules of Chinese university students may be a significant contributing factor to the difficulties experienced by nocturnal eaters regarding the consumption of breakfast and lunch. Breakfast is typically served between 7 am and 9 am. The schedules of the first and second classes (at 8 am and 10 am, respectively) influenced students’ eating habits, particularly those of nocturnal eaters. These students often have no choice but to eat between classes. Furthermore, the proximity of breakfast to lunchtime makes it challenging for nocturnal eaters to reach the desired level of hunger to consume lunch. Indeed, in a study based on Chinese university students (both male and female), approximately 40% consumed breakfast at least 30 minutes after awakening, whereas approximately 20% did not consume breakfast at the optimal time[20, 21].

It must be acknowledged that this study also has some inevitable limitations. Firstly, inherent limitations of cross-sectional studies, which do not prove causality, are inevitable and unavoidable aspects of the research design. Secondly, as this study used a food preference questionnaire, it was not feasible to identify the dietary intake of each food group. Accordingly, it is recommended that future studies use food frequency questionnaires to identify the daily intake of each food group, with a view to acquiring insight into the potential associations between dietary intake and NES. Third, this study solely examined the frequency of the three meals consumption, precluding any determination of the specific food items consumed at each meal. It is recommended that future studies should adopt a more comprehensive approach to the measurement of food intake, including the types of food consumed, at the three main meals in order to establish the potential association between dietary quality at each meal and NES.

Night eating syndrome

NES

Odds ratios

ORs

95% confidence intervals

95% CIs

Declarations

Ethics approval and consent to participate

The studies involving humans were approved by Research Institute of Ethics Committee of the School of Physical Education, Southwest University. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

Consent for publication

Not applicable

Competing interests

The authors declare that the research was conducted in the absence of any commercial or financial associations that could be construed as a potential conflict of interest.

Funding

The author(s) declare that financial support was received for the research, authorship, and/or publication of this article. This study was supported by the Ministry of Education of Humanities and Social Science project, China (Grant No. 21YJCZH125) and Chongqing Municipal Social Science Planning Project - Youth Project (Grant No. 2021NDQN66) and the Science and Technology Research Program of Chongqing Municipal Education Commission (Grant No. KJQN202100216).

Author Contribution

ZH: Writing – review & editing, Writing – original draft, Investigation, Formal analysis, Data curation, Conceptualization.QJ: Writing – review & editing, Writing – original draft, Software, Methodology, Formal analysis, Data curation, Conceptualization. BZ: Writing – review & editing, Writing – original draft, Methodology, Conceptualization. ZR: Writing – review & editing, Writing – original draft, Supervision, Resources, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization.

Acknowledgments

We thank all of the college students who agreed to participate and gave informed consent for analysis of their data. We would also like to thank our staff for their dedicated work.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Stunkard AJ, Grace WJ, Wolff HG. THE NIGHT-EATING SYNDROME - A PATTERN OF FOOD INTAKE AMONG CERTAIN OBESE PATIENTS. Am J Med. 1955;19(1):78–86.
McCuen-Wurst C, Ruggieri M, Allison KC. Disordered eating and obesity: associations between binge-eating disorder, night-eating syndrome, and weight-related comorbidities. Volume 1411. Annals of the New York Academy of Sciences; 2018. pp. 96–105. 1.
Allison KC, et al. Proposed Diagnostic Criteria for Night Eating Syndrome. Int J Eat Disord. 2010;43(3):241–7.
Cleator J, et al. Night eating syndrome: implications for severe obesity. Nutrition & Diabetes; 2012. p. 2.
Hood MM, Reutrakul S, Crowley SJ. Night eating in patients with type 2 diabetes. Associations with glycemic control, eating patterns, sleep, and mood. Appetite. 2014;79:91–6.
Sevincer GM, et al. Night Eating Syndrome Frequency in University Students: Association with Impulsivity, Depression, and Anxiety. Klinik Psikofarmakoloji Bulteni-Bulletin Clin Psychopharmacol. 2016;26(3):238–47.
Kara Y, et al. Night Eating Syndrome According to Obesity Groups and the Related Factors. Jcpsp-Journal Coll Physicians Surg Pakistan. 2020;30(8):833–8.
Lent MR et al. Night eating, weight, and health behaviors in adults participating in the Daily24 study. Eat Behav, 2022. 45.
Khan MSI et al. Skipping breakfast and its association with sociodemographic characteristics, night eating syndrome, and sleep quality among university students in Bangladesh. Bmc Nutr, 2024. 10(1).
Deniz CD, et al. Associations between night eating syndrome and metabolic parameters in pregnant women. Turkish J Obstet Gynecol. 2019;16(2):107–11.
Fan R, et al. Survey on night snack behavior among students in a medical university in Beijing. Volume 33. CHINESE JOURNAL OF HEALTH EDUCATION; 2017. pp. 845–7. 09.
Small M, et al. Changes in Eating and Physical Activity Behaviors Across Seven Semesters of College: Living On or Off Campus Matters. Volume 40. Health Education & Behavior; 2013. pp. 435–41. 4.
Lahteenmaki L. Food Quality and Preference. Food Qual Prefer. 2017;56:231–2.
Allison KC, et al. The Night Eating Questionnaire (NEQ): psychometric properties of a measure of severity of the Night Eating Syndrome. Eat Behav. 2008;9(1):62–72.
Scheer FAJL, Morris CJ, Shea SA. The Internal Circadian Clock Increases Hunger and Appetite in the Evening Independent of Food Intake and Other Behaviors. Obesity. 2013;21(3):421–3.
Gaillard RC et al. Cytokines, leptin, and the hypothalamo-pituitary-adrenal axis. Neuroimmunomodulation: Perspectives at the New Millennium, ed. A. Conti, Vol. 917. 2000. 647–657.
Simon C, et al. Circadian and ultradian variations of leptin in normal man under continuous enteral nutrition: Relationship to sleep and body temperature. J Clin Endocrinol Metabolism. 1998;83(6):1893–9.
Birketvedt GS, et al. Behavioral and neuroendocrine characteristics of the night-eating syndrome. Jama-Journal Am Med Association. 1999;282(7):657–63.
Molzof HE, et al. Nightshift Work and Nighttime Eating Are Associated With Higher Insulin and Leptin Levels in Hospital Nurses. Frontiers in Endocrinology; 2022. p. 13.
Zhang J, et al. Analysis of breakfast situation among college students in a certain university in Hebei Province. Food Nutr China. 2024;30(04):10–4.
Sun X, Sun C, Wang H. Research and analysis on breakfast eating habits and dietary hygiene habits of college students. China Food Saf Magazine, 2023(04): pp. 109–11.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

The association between night eating syndrome and food preferences among Chinese college students: a large-scale cross-sectional study

Status:

Version 1

Abstract

Background and aims:

Methods

Results

Conclusion

1 Background

2 Methods

2.1 Participants

2.2 Assessment of food preference

2.3 Assessment of night eating syndrome

2.4 Assessment of covariates

2.5 Statistical analysis

3 Results

4 Discussion

5 Limitations

6 Conclusion

Abbreviations

Declarations

Declarations

Competing interests

Funding

Author Contribution

Acknowledgments

Availability of data and materials

References

Additional Declarations

Status:

Version 1