Current Developments in Nutrition 7 (2023) 100009

journal homepage: www.journals.elsevier.com/current-developments-in-nutrition

Original Research

Adaptation and Validation of the Malay-Chrononutrition Profile

Questionnaire to Assess Chrononutrition Behavior of Young Adults
in Malaysia
Khairunnisa Fazira Hairudin 1, Nur Islami Mohd Fahmi Teng 1, *, Norsham Juliana 2
1 2
Faculty of Health Sciences, Universiti Teknologi MARA, Bandar Puncak Alam, Malaysia; and; Faculty of Medicine and Health Sciences,
Universiti Sains Islam Malaysia, Nilai, Malaysia

A B S T R A C T

Background: Chrononutrition studies how biological rhythms and nutrition are associated with human health. However, a validated
assessment in Malaysia is still absent.
Objectives: To create a translation of the Chrononutrition Profile Questionnaire (CPQ), test its validity and reliability, and determine the
general chrononutrition behaviors among Malaysian young adults.
Methods: The Malay-CPQ was distributed to respondents through online platforms (n ¼ 110), and data analyses were performed. The data
were analyzed for their validity using content validity index (CVI) and face validity index (FVI), whereas intraclass correlation coefficient
(ICC) was used to determine test–retest reliability.
Results: Our results showed both CVI and FVI of Malay-CPQ were 1, indicating excellent content translation, while the ICC values ranged
from moderate to good (0.50–0.90). The Cronbach α values for all items ranged from moderate to good (0.50–0.90), and the Bland–Altman
analysis showed a P value >0.05, indicating agreement of the item between repeated measurements. The chrononutrition behaviors among
Malaysian young adults presented fair to good scores for all behavior patterns: eating window, breakfast skipping, evening eating, night
eating, and largest meal, except evening latency, being mostly at the poor score (>80% responses).
Conclusions: The Malay-CPQ is a valid and reliable tool to assess the Malaysian chrononutrition profile. However, further testing on Malay-
CPQ should be conducted in a different setting in Malaysia for cross-validation studies.

Keywords: chrononutrition, young adults, Malaysia, Malay-CPQ, validation

Introduction of an individual is optimum. It was reported that poor adherence

to a healthy diet was usually exhibited by individuals with late
Chrononutrition is defined as the study of biological rhythms chronotype [4, 5]. Moreover, this group observed a habit of
and nutrition and its relationship with human health, which delaying the timing of eating, frequently skipping breakfast,
include energy distribution, meal frequency, and regularity [1]. showing a high preference for sweet food and drinks, drinking
It is an emerging study on how food affects the body’s circadian alcohol, and not consuming enough fruit and vegetables. Un-
rhythm [2]. Six distinct behavioral patterns are likely to influ- fortunately, individuals who have a high-calorie intake at a later
ence one’s chrononutrition: 1) eating at night, 2) restricted time time of the day are exposed to the risk of getting obese [6, 7].
feeding, 3) breakfast eating, 4) timing of the largest meal, 5) time This is because the process of lipogenesis and accumulation of
of evening eating, and 6) time between eating and sleep time [3]. adipose tissues tend to occur during the period of the last meal.
The circadian mealtime pattern is essential for food intake Additionally, breakfast skippers will experience a longer period
regulation and body metabolism to ensure the well-being status of lipolysis and lipogenesis owing to their eating habits [8].

Abbreviations used: CPQ, ChrononutritionProfile Questionnaire; CVI, content validity index; FTQ, Food Timing Questionnaire; FTS, Food Timing Screener; FVI, face
validity index; ICC, intraclass correlation coefficient.
* Corresponding author. E-mail address: [email protected] (N.I. Mohd Fahmi Teng).

https://doi.org/10.1016/j.cdnut.2022.100009
Received 13 July 2022; Received in revised form 22 September 2022; Accepted 13 November 2022; Available online 22 December 2022
2475-2991/© 2023 Published by Elsevier Inc. on behalf of American Society for Nutrition. This is an open access article under the CC BY license (http://
creativecommons.org/licenses/by/4.0/).
K.F. Hairudin et al. Current Developments in Nutrition 7 (2023) 100009

These unhealthy eating habits eventually will affect the wellness relevant). 3) pilot testing was conducted with 10 bilingual un-
of an individual, in the aspects of cognitive, behavioural, and dergraduate students who rated and commented on the items. 4)
physical status. Disruption in circadian rhythm will result in the A reliability test through test–retest reliability (n ¼ 110) was
development of morbidities or metabolic diseases such as performed. Respondents were recruited through various online
abdominal obesity, dyslipidemia, dysglycemia, and hyperten- platforms such as WhatsApp, Twitter, and Instagram. On con-
sion. Thus, chrononutrition may affect the quality of life of an sent, the questionnaires were distributed online and were
individual owing to the misalignment interaction between di- administered twice, within 14–21-d intervals. The targeted re-
etary intake and meal timing. A study by Oike et al. [9] high- spondents were Malaysian undergraduate students aged 18–30 y
lighted that many countries with a high prevalence of metabolic who understood the Malay language. In the pilot study, re-
diseases recorded people with disrupted mealtimes and deprived spondents were asked to complete the questionnaire and rate on
sleep. In addition, chronodisruption or loss of circadian pattern a 4-point Likert scale of 1 (item very not clear and understand-
causes a decline in health and exposes one to metabolic abnor- able) to 4 (item is very clear and understandable). Scores 3 and 4
malities. Eating late at night was reported to increase the risk of were reclassified as 1 (clear and understandable) and scores 1
getting obese [10]; whereas, skipping breakfast was found to and 2 as 0 (not clear and understandable). Young adult partici-
affect the concentration of postprandial insulin and might pants was chosen as the target group in this study as little is
disrupt normal glucose homeostasis [4]. known about chrononutrition behavior of the young adults or
The Chrononutrition Profile Questionnaire (CPQ) was intro- college students and feasibility in the subject recruitment.
duced by Veronda et al. [3], which was developed and validated Nonetheless, a recent study in Malaysia on the CPQ validation
in the United States. Generally, CPQ has been designed as a has focused on the adult population (aged 18–65 y). Therefore,
comprehensive assessment of chrononutrition aspects without this study aimed to provide a more specific target group to assess
burdening participants. Consisting of 18 items, CPQ is intended the reliability [11].
to determine patterns of chrononutrition behaviors and chro- The content validity index (CVI) is an index used in quanti-
nonutrition preferences on both weekdays and weekend days. tative evaluation. There are 2 types of CVI: I-CVI for chance
There are 4 domains of the CPQ, which asks about the following: agreement and S-CVI for average (S-CVI/Ave) and universal
1) chrononutrition preferences; 2) frequency of certain chro- agreement (S-CVI/UA). All these have different formulas to
nonutrition behavior; 3) timing of eating events during workday; determine the degree of item appropriateness to be measured in
and 4) timing of eating events during a free day. The first and a questionnaire. A CVI score of >0.83 indicated that the items in
second domains consist of 4 items while the third and fourth 5 the questionnaire were relevant to the domain. A scale of <0.79
items each. Because chrononutrition assessment is still new, indicates that the item requires revision, and if it was <0.70, the
having a specific and validated tool to assess meal timings and item should be eliminated. On the contrary, the face validity
patterns is necessary. index (FVI) is an index to the extent of whether the items mea-
Interventions through chrononutrition profile assessment as a sure what it is intended to measure. Two types of FVI are I-FVI for
recent and different approach for health management shall help chance agreement and S-FVI for average (S-FVI/Ave) and uni-
improve an individual's quality of life. However, a validated tool versal agreement (S-FVI/UA). Similar to CVI, all these have
is absent in Malaysia assessing chrononutrition profiles. Thus, different formulas to measure item relevancy in questionnaires.
this study aimed to create a translation of the CPQ into the Malay The satisfactory level of FVI was taken at 0.83 or more, adopted
language and to test its reliability and validity among young from the CVI value. A scale of <0.79 indicated that the item
adults in Malaysia. This is necessary because the first language in requires revision, and if it was <0.70, the item should be elim-
Malaysia is Malay; hence, a Malay-CPQ is desired to benefit other inated [12].
generations or populations in Malaysia who may not be fluent in Data obtained were analyzed by using IBM SPSS 20 involving
English. several analyses. The first analysis was the intraclass correlation
coefficient (ICC), intended to establish and quantify reproduc-
Materials and Methods ibility and provide an indication of the test–retest reliability of
measurement. The ICCs were calculated by using a 2-way, mixed
This study was a quantitative and cross-sectional study that consistency model, with an acceptable value of 0.50 as moderate
involved 4 stages of validation: 1) Forward and backward reliability, 0.75–0.90 as good reliability, and >0.90 as excellent
translation was conducted by 2 bilingual translators, translating reliability [13]. Next, separated correlation coefficients were
from the English version of CPQ to the Malay language and back computed for weekdays and weekend values for chrononutrition
to the English language by another 2 bilingual translators. All behaviors to ensure the continuous values of items were reliable
translators hold a higher degree in education and have at least 2 because those were the foundation of computed weekly average
years of experience in translation and language teaching. 2) An and chrononutrition behavior cutoff scores.
expert committee review and content validity was performed by The second analysis used Cronbach α to assess the internal
a panel of 6 experts. The panel comprised nutritionists, di- consistency of items with acceptable values ranging from 0.60 to
eticians, researchers, and academics who are proficient in En- 0.95 [14]. A value >0.95 might indicate the presence of redun-
glish and Malay language and understand the construct of CPQ. dant items. Then, the agreement of reported values was
They were asked to rate on a 4-point Likert scale: score 1 (item measured with the Bland–Altman analysis. The P value for each
very not relevant) to 4 (item very relevant), based on the items’ item should be >0.05 to reject the null hypothesis, indicating
relevancy to the Malay-CPQ. Scores 3 and 4 were reclassified as 1 there was no proportional bias for the item between repeated
(relevant), whereas scores 1 and 2 were recategorized as 0 (not measurements. As for chrononutrition behavior, data were

2
K.F. Hairudin et al. Current Developments in Nutrition 7 (2023) 100009

classified according to scoring cutoffs for the chrononutrition moderate correlations. In addition, correlations between evening
profile derived from a table in the study by Engwall [15]. eating, evening latency, and breakfast frequency were good.
This study has obtained approval from the Universiti Tekno- Cronbach α was calculated to determine internal consistency.
logi MARA Research Ethics Committee, with a reference number Reported chrononutrition preferences were relatively high in
REC/06/2021 (UG/MR/589). consistency. In addition, the internal consistency for the
workday was high for lunchtime, last eating event, and bedtime
Results and moderate for wake time and first eating event. Furthermore,
the internal consistency for free days was high for wake time, last
eating event, and bedtime, moderate for lunchtime but low for
Content and face validity indices
first eating event. Weekly averages were computed for breakfast
Overall, the I-CVI, S-CVI/Ave, and S-CVI/UA scales obtained a
frequency, evening eating, after-dinner snack, evening latency,
score of 1, which was considered excellent, indicating that the
and eating window, and the internal consistency was high.
content of translated Malay-CPQ version for each item was well
Table 3 provides detailed information on test–retest reliability
adapted to the local context. A CVI score of >0.83 indicated that
involving ICC, Cronbach α, and Bland–Altman analysis values.
the items in the questionnaire were relevant to the domain. All
The item with a poor reliability coefficient (first eating event
comments given by the expert panels were taken to improve the
during free days) was checked for its extent of agreement with
comprehension of items. Table 1 presents detailed information
the Bland–Altman analysis. The P value of the first eating event
on the CVI.
for free days was >0.05, indicating the item could reject the null
Generally, the I-FVI, S-FVI/Ave, and S-FVI/UA scales ob-
hypothesis and had no proportional bias. Therefore, all items
tained a score of 1, which was considered excellent, indicating
were proven valid and reliable, allowing the final version of
that the face validity of translated Malay-CPQ version for each
Malay-CPQ to consist of 18 items.
item was using clear and understandable sentences. In addition,
an FVI score of >0.83 indicated that the items in the question-
naire were relevant to the domain. In summary, item content and Sociodemographic information
face validity had been proven to be valid, and the validated Most participants in this study were women. Each participant
Malay-CPQ could be used for test–retest reliability purposes. had a consistent demographic profile, with all of them being
Table 2 tabulates the detailed information on the FVI. Malaysian citizens, Malay, and undergraduate students from
various universities in Malaysia. The ages of participants ranged
from 18 to 28 y (mean: 21.50 y; SD: 2.32 y). Most of the par-
Test–retest reliability
ticipants showed normal BMI values (51.8%), followed by un-
Reported chrononutrition preferences (e.g., wake time pref-
derweight (19.1%) and overweight (11.8%) (Table 4).
erence and preferred bedtime) were correlated (Table 3). Cor-
relations for workday values ranged from moderate (wake time,
first eating event, last eating event, and bedtime) to good Chrononutrition and scoring cutoffs for the
(lunchtime), whereas correlations for free day values ranged chrononutrition profile
from moderate (first eating event, lunchtime, and last eating Descriptive statistics were used to explore the characteristics
event) to good (wake time and bedtime). Weekly averages were of the chrononutrition profile. More specifically, this study
computed for eating window and after-dinner snacks, showing examined the range and distributions of scores on individual

Table 1
The clarity and comprehension ratings on the items scale by 6 experts for content validity
Item Expert 1 Expert 2 Expert 3 Expert 4 Expert 5 Expert 6 No. of agreements I-CVI
A1 3 3 3 3 3 3 6 1
A2 3 3 3 3 3 3 6 1
A3 3 3 3 3 3 3 6 1
A4 3 3 3 3 3 3 6 1
B1 4 3 3 3 4 3 6 1
B2 4 3 3 3 4 3 6 1
B3 4 3 3 3 4 3 6 1
B4 4 3 3 3 4 3 6 1
C1 4 3 3 4 4 3 6 1
C2 4 3 3 4 4 3 6 1
C3 4 3 3 4 4 3 6 1
C4 4 3 3 4 4 3 6 1
C5 4 3 3 4 4 3 6 1
D1 4 3 3 4 3 3 6 1
D2 4 3 3 4 3 3 6 1
D3 4 3 3 4 3 3 6 1
D4 4 3 3 4 3 3 6 1
D5 4 3 3 4 3 3 6 1
S-CVI/average 1
Total agreement 18
S-CVI/UA 1

CVI, content validity index; UA, universal agreement.

3
K.F. Hairudin et al. Current Developments in Nutrition 7 (2023) 100009

Table 2
The clarity and comprehension ratings on the items scale by 10 raters for face validity (pilot testing).
Rater 1 Rater 2 Rater 3 Rater 4 Rater 5 Rater 6 Rater 7 Rater 8 Rater 9 Rater 10 No. of agreements I-FVI
Item
A1 4 4 3 4 4 4 4 3 4 4 10 1
A2 4 4 3 4 4 4 4 3 4 4 10 1
A3 4 4 3 4 4 4 4 3 4 4 10 1
A4 4 4 3 4 4 4 4 3 4 4 10 1
B1 3 4 3 4 3 3 4 3 4 3 10 1
B2 3 4 3 4 3 3 4 3 4 3 10 1
B3 3 4 3 4 3 3 4 3 4 3 10 1
B4 3 4 3 4 3 3 4 3 4 3 10 1
C1 4 4 4 4 4 4 4 4 4 4 10 1
C2 4 4 4 4 4 4 4 4 4 4 10 1
C3 4 4 4 4 4 4 4 4 4 4 10 1
C4 4 4 4 4 4 4 4 4 4 4 10 1
C5 4 4 4 4 4 4 4 4 4 4 10 1
D1 4 4 4 4 4 4 4 4 4 4 10 1
D2 4 4 4 4 4 4 4 4 4 4 10 1
D3 4 4 4 4 4 4 4 4 4 4 10 1
D4 4 4 4 4 4 4 4 4 4 4 10 1
D5 4 4 4 4 4 4 4 4 4 4 10 1
S-FVI/Ave 1
Total Agreement 18
S-FVI/UA 1

FYI, face validity index; UA, universal agreement.

Table 3
Values on test–retest reliability involving the ICC, Cronbach α, and Bland–Altman analysis
Item Test (mean  SD) Retest (mean  SD) Cronbach α ICC (average measure) Bland–Altman
Chrononutrition preferences
Wake time 7.72  1.82 7.59  1.71 0.90 0.90 0.255
Morning latency 1.74  1.54 1.61  1.30 0.86 0.86 0.008
Evening latency 2.78  1.68 2.92  1.69 0.79 0.79 0.916
Bedtime 7.18  4.82 8.12  4.62 0.72 0.72 0.600
Chrononutrition behaviors: weekdays
Wake time 6.81  1.13 7.01  1.35 0.60 0.59 0.044
First eating event 8.34  2.35 8.34  2.36 0.64 0.64 0.943
Lunchtime 3.91  4.47 3.65  4.51 0.83 0.83 0.877
Last eating event 9.20  1.38 9.02  1.56 0.73 0.72 0.132
Bedtime 7.06  5.09 7.22  5.06 0.70 0.70 0.946
Chrononutrition behaviors: weekend days
Wake time 8.54  1.94 8.48  2.14 0.76 0.76 0.209
First eating event 8.03  3.56 8.45  3.36 0.50 0.50 0.521
Lunchtime 3.35  3.85 3.59  4.05 0.62 0.62 0.560
Last eating event 9.36  1.62 9.34  1.57 0.70 0.70 0.728
Bedtime 5.76  4.77 6.13  4.93 0.85 0.85 0.638
Computed weekly averages
Breakfast frequency 4.43  2.21 4.74  2.01 0.88 0.87 0.121
Evening eating 9.25  1.37 9.14  1.52 0.75 0.75 0.180
After-dinner snack 3.28  1.95 3.21  1.90 0.71 0.71 0.737
Evening latency 2.56  5.05 2.22  5.11 0.79 0.79 0.867
Eating window 1.00  2.39 0.76  2.50 0.70 0.70 0.594

All correlations were significant at the P < 0.001 level.

ICC, intraclass correlation coefficient.

items and chrononutrition behavior cutoff classifications. All breakfast skipping between the test and retest questionnaire
participants reported responses in the Malay-CPQ test and showed that most of the participants had not or only skipped
retested the version excellently. Hence, this analyses included all breakfast 1 dci, along a week, followed by more than one-third
6 general chrononutrition behaviors. Table 5 summarizes exhibiting poor habit by skipping 4 d/wk, and less than one-
detailed chrononutrition behavior cutoffs for chrononutrition third reporting fair habit with skipping for 2–3 d/wk.
profiles from Malay-CPQ (test) and Malay-CPQ (retest). A poor outcome was reported for evening latency because
The eating window from both test and retest Malay-CPQ were almost 80% of participants reported a poor duration between the
100% matched, and all participants exhibited a good duration last eating event and sleep onset (2:00), followed by <30 of
between the first and last eating events (12:00). The reported them having fair scores (2:01–6:00) for both test and retest

4
K.F. Hairudin et al. Current Developments in Nutrition 7 (2023) 100009

Table 4 skipping, 3) evening latency, 4) evening eating, 5) night eating,

Sociodemographic details (n ¼ 110) and 6) largest meal. Overall, the findings of this study corrobo-
Demographic variable Parameter rate the Malay-CPQ hypothesis.
CPQ was chosen as this study instrument because it consisted
Age (y)
Mean 21.50
of comprehensive items to measure the chrononutrition pattern
SD 2.32 including time of eating event and sleep time that requires a
Sex, n (%) short time for respondents to answer. The items were simple and
Male 11 (10) did not give a significant burden on respondents. The original
Female 99 (90) CPQ was translated to fit Malaysian culture because the first
BMI, n (%)
Underweight 21 (19.1)
language in Malaysia is Malay. Compared with other tools
Normal 57 (51.8) measuring meal timing, such as the Food Timing Questionnaire
Overweight 19 (17.3) (FTQ) and Food Timing Screener (FTS), the CPQ can capture the
Obese 13 (11.8) largest meal of the day and calculate eating misalignment as well
[16]. The FTQ is designed to assess an individual’s usual eating
time and sleep habits throughout the week, whereas FTS assesses
Table 5 the same elements but on separate types of day: school/work
Distribution of chrononutrition behavior cutoffs for Chrononutrition days and free days. FTS seemed to have the same construct as
Profile from Malay-CPQ (test) and Malay-CPQ (retest) (n ¼ 110) CPQ, which allows accurate food intake and mealtime during
Chrononutrition behavior Malay-CPQ (test), Malay-CPQ (retest), school/work days and free days.
cutoffs n (%) n (%) The translated Malay-CPQ version provided valid and reliable
Eating window results when it was applied. This study looked at content
Good (12:00) 110 (100) 110 (100) equivalence and semantic equivalent to ensure that the quality
Fair (12:01–14:00) — — and consistency of the meaning in the translated version
Poor (>14:00) — —
matched the original version. The translation and adaptation
Breakfast skipping
Good (1 d/wk or less) 39 (35.5) 44 (40.0)
processes in this study conformed to the guidelines that the
Fair (2–3 d/wk) 32 (29.0) 32 (29.0) translated instruments need to be evaluated regarding content,
Poor (4 d/wk) 39 (35.5) 34 (30.9) conceptual, context, semantic, and technical equivalency to
Evening latency ensure the instrument is appropriate for use in the new setting.
Good (>6:00) — —
Failure to adhere to the guidelines might lead to significant
Fair (2:01–6:00) 24 (21.8) 25 (22.7)
Poor (2:00) 86 (78.2) 85 (77.3)
contextual and conceptual equivalence issues.
Evening eating In this study, overall, the individual items rated by 6 expert
Good (<20:00) 12 (10.9) 12 (10.9) panels, the I-CVI scale obtained a score of 1 (relevant), which is
Fair (20:00–22:59) 85 (77.3) 86 (78.2) considered excellent, indicating that the content of the translated
Poor (23:00) 13 (11.8) 12 (10.9) Malay-CPQ version is well adapted to the local context. A CVI
Night eating
Good (1 d/wk or less) 94 (85.5) 90 (81.8)
score of >0.8 indicates that the items in the questionnaire are
Fair (2–3 d/wk) 9 (8.2) 14 (12.7) relevant to the domain. The excellent score was supported with
Poor (4 d/wk) 7 (6.4) 6 (5.5) both S-CVI/Ave and S-CVI/UA scales, which obtained a score of
Largest meal 1. A score of 1 was the highest score that can be given to evaluate
Breakfast 1 (0.9) 3 (2.7) items. This score was a combination of scores 3 and 4 given by
Lunch 91 (82.7) 83 (75.5)
Dinner/supper 18 (16.4) 24 (21.8)
expert panels for each item and was classified on an acceptable
scale of 1.
Malay-CPQ, Malay-Chrononutrition Profile Questionnaire. As for the individual items ranged by undergraduate students,
overall, the I-FVI scale obtained a score of 1 (relevant) as well.
Malay-CPQ. The evening eating, in which the risk of eating late This shows that the original CP-Q is well translated into the
in the waking day was reported, showed almost 80% of partici- Malay language using clear and understandable sentences. The
pants having a fair score (20:00–22:59) and <12% for both poor result was supported by both S-FVI/Ave and S-FVI/UA scales,
(23:00) and good (<20:00) scores. The reported frequency of which also obtained a score of 1. The satisfactory level of the FVI
night eating had obtained results with >80% of the participant was taken at 0.8 and above, and it was adopted from the CVI
having good scores (1 d/wk or less), <15% having fair (2–3 d/ values.
wk), and poor (4 d/wk) scores. Lunchtime was reported as the The Malay-CPQ provided strong evidence for test–retest
largest meal by the most of the participants for both test and reliability, with all items significantly correlating over the
retest Malay-CPQ (>75%), dinner/supper as the second-largest 14–21-d period. Two aspects should be noted regarding the
meal (>16%), and breakfast as the least one (>3%). test–retest reliability: 1) weekend values were slightly less
correlated compared with weekdays values, and 2) the correla-
Discussion tion coefficient for the first eating event on weekend days was
particularly low compared with that of the other items. This
This present study aimed to develop and evaluate the Malay- could indicate that the timing of food intake on weekend days
CPQ: a measure derived from the original version, CPQ [15]. The may be more variable over time than the timing of food intake on
Malay-CPQ assessed 6 specific behavioral patterns likely to affect weekdays, and the first eating event may vary more on weekends
one’s chrononutrition profile: 1) eating window, 2) breakfast compared with the timing of other eating events. In addition,

5
K.F. Hairudin et al. Current Developments in Nutrition 7 (2023) 100009

chrononutrition preferences were strongly correlated over the was linked to a lack of fruits and vegetables, increased soft drink
testing period. The data were collected within 2 mo, from consumption, and a failure to avoid fat and cholesterol. Fanelli
September until October of 2021, whereby there were no new et al. [21] supported this study, where they found that adults
changes in the restriction orders due to the COVID-19 pandemic who did not eat breakfast showed a considerably lower overall
announced by the government of Malaysia during the period. diet quality.
Therefore, these strong correlations indicate that the Malay-CPQ An ample time given to the body to digest food is a good habit
is reliable based on the ICC values. to ensure quality sleep. However, almost 80% of our respondents
The internal consistency measured with Cronbach α values reported <2 h gap between the last eating event and sleep onset.
showed good consistency for all items except for the first eating It was reported previously that night-time eating was associated
event from the weekend value. The increased duration of online with additional calorie intakes among college-aged students,
learning was reported to be associated with breakfast skipping which supported other previous studies among shift workers,
and increased sugar-sweetened beverages during the COVID-19 and middle older-aged populations [22]. Excessive calorie intake
restrictions [17]. In addition, the rising use of computers, tab- causes unnecessary weight gain and exposed individuals to
lets, and smartphones for online learning has been linked to poor metabolic disorders. Instead, a study by Goel et al. [23] found
sleep duration and lower quality of life regarding mental health. that evening eating did contribute to weight gain and metabolic
These had affected students’ wake time due to poor quality of dysfunction, regardless of calorie intake. However, a study by
sleeping habits, which were directly associated with the timing Maw and Haga [24] found that in middle-aged and elderly
of their first eating event. Sinha et al. [18] showed that sleep Japanese persons, 2-h or shorter intervals between dinner and
onset–wakeup and breakfast times were significantly delayed night did not affect HbA1C alterations. To sustain stable HbA1C
during the lockdown. Owing to the existing factors, participants patterns, the focus should be on maintaining a normal BMI value
had different and unfixed timing for their first meal. The and refraining from smoking and consuming alcohol in the long
inconsistency of reported timing had affected not only the run. This finding brought new insight into which younger pop-
Cronbach α values but also the ICC values, causing poor reli- ulation who are unable to change their habit of leaving 2 h
ability mainly on the first eating event of weekend days. before bedtime to focus on maintaining BMI in the normal range
For the existing factors that had affected the reported and abstaining from smoking and drinking alcohol instead.
response, the item was later checked for the extent of the The reported frequency of night eating had obtained results
agreement by using the Bland–Altman analysis. Finally, the item with >80% of the respondents having a good (1 d/wk or less)
was plotted on a scatter graph to describe the agreement be- score. Night-time eating is a risk factor for metabolic syndrome
tween 2 quantitative measurements. As presented in the result and obesity. This was supported by a study by Yoshida et al. [25],
part, the first eating event for weekend days with a poor Cron- who found that eating behavior at night was linked to dyslipi-
bach α coefficient value had a P value more than 0.05 for the demia in both men and women. Lunchtime was the largest meal,
Bland–Altman analysis, indicating the null hypothesis was followed by dinner/supper and breakfast. In general, studies by
rejected, an agreement was presented, and no proportional bias Ruddick-Collins et al. [26] and Garaulet et al. [27] suggested
for the item. Hence, no changes in translation or sentence that prioritizing energy intake earlier in the day may encourage
structure were made to the item. Thus, this study was able to increased energy expenditure and morning-loaded energy dis-
keep all items in Malay-CPQ because they were proven to be tribution is a viable weight-control strategy for persons who are
valid and reliable. Therefore, the final version of Malay-CPQ in circadian alignment.
remained with 18 items. In summary, the chrononutrition behaviors of respondents
The reported responses of 110 respondents from both among young adults, mainly the 6 behavior patterns, showed
test–retest Malay-CPQ were analyzed for their chrononutrition that they had fair to good scores (eating window, breakfast
behavior. With only slight differences in the number of reported skipping, evening eating, night eating, and largest meal) except
responses, both questionnaires had similar chrononutrition for evening latency. However, evening latency should not be an
behavior, preferences, and order based on scoring cutoffs for the issue if the aforementioned preventive measures are followed.
chrononutrition profile. All respondents eventually had a good Moreover, the findings revealed a definite pattern in chrononu-
eating window (12:00). An eating window of <12 h is trition behaviors of undergraduate students during the COVID-
considered a good chrononutrition behavior because it allows 19 pandemic, which health care professionals mainly might
the human body to have better digestion and be able to get a apply to improve the well-being of young adults.
good sleep time. Moreover, excessive food intake can be avoided The results of this study may not be representative of
because the time to eat is within 12 h, which helps to avoid being Malaysia because all respondents were from the Malay ethnic
overweight, obese, and cardiac disorders. A study conducted by group, limiting its generalizability. Other limitations include that
Manoogian et al. [19] found that time-restricted feeding/eating, the undergraduates may have the possibility to interpret the
a new meal-timing technique that combines eating and drinking questionnaire differently or have different meal patterns to older
all our daily calories in a regular 8–12-h window, or less, might adults.
help with metabolic and cardiovascular health. In addition, the In conclusion, the translated Malay-CPQ version has a good
study stated that small human studies have shown that daily content validity, face validity, and reliability. Therefore, the
eating lengths of 4–11 h/d can lower blood pressure, improve translated Malay-CPQ version is a reliable tool for determining
blood sugar, and aid with weight, energy levels, sleep, and the chrononutrition profile among young adults in Malaysia.
appetite. However, future studies should apply the Malay-CPQ in different
Most respondents did not have the habit of skipping break- populations and situations in Malaysia for cross-validation
fast. A study by Pengpid et al. [20] found that breakfast skipping because this study focused on undergraduate students during

6
K.F. Hairudin et al. Current Developments in Nutrition 7 (2023) 100009

pandemic COVID-19. Thus, the Malay-CPQ version can become [8] N.A.S. Mohd Azmi, N. Juliana, N.I. Mohd Fahmi Teng, S. Azmani,
one of the valid indicators to measure the chrononutrition profile S. Das, N. Effendy, Consequences of circadian disruption in shift
workers on chrononutrition and their psychosocial well-being, Int J
of the whole Malaysian community, leading to a cost-saving and Environ Res Public Health 17 (6) (2020) 2043.
effective method to assess chrononutrition behaviors for the [9] H. Oike, K. Oishi, M. Kobori, Nutrients, clock genes, and
well-being improvement of Malaysians in general. chrononutrition, Curr Nutr Rep 3 (3) (2014) 204–212.
[10] S. Bo, G. Musso, G. Beccuti, M. Fadda, D. Fedele, R. Gambino, et al.,
Consuming more of daily caloric intake at dinner predisposes to obesity.
Data Availability A 6-year population-based prospective cohort study, PloS One 9 (9)
(2014), e108467.
[11] G.Y. Chong, S. Kaur, R. Abd Talib, S.L. Loy, W.Y. Siah, H.Y. Tan, et al.,
The data described in the manuscript, code book, and analytic Translation, validation, and reliability of the Malay version
code will be made available on request pending application. chrononutrition profile-questionnaire (CPQ-M) in Malaysian adult
populations, Biol Rhythm Res (2022) 1–22.
[12] I.B. Rodrigues, J.D. Adachi, K.A. Beattie, J.C. MacDermid, Development
Funding and validation of a new tool to measure the facilitators, barriers and
preferences to exercise in people with osteoporosis, BMC Musculoskelet
Disord 18 (1) (2017) 1–9.
This work was supported by Universiti Teknologi MARA, 600- [13] T.K. Koo, M.Y. Li, A guideline of selecting and reporting intraclass
RMC/YTR/5/3 (001/2022). correlation coefficients for reliability research, J Chiropract Med 15 (2)
(2016) 155–163.
[14] P. Hinton, I. McMurray, C. Brownlow, SPSS explained, Routledge,
Author disclosures London, United Kingdom, 2014.
[15] A.C. Engwall, Development, Validation and Reliability of the
Chrononutrition Profile, Available from: https://hdl.handle.net/10
All authors report no conflicts of interest.
365/29276, 2019.
[16] P. Chakradeo, H.E. Rasmussen, G.R. Swanson, B. Swanson, L.F. Fogg,
Acknowledgments F. Bishehsari, et al., Psychometric testing of a food timing questionnaire
and food timing screener, Curr Dev Nutr 6 (2) (2022) nzab148.
[17] C. Chusak, M. Tangmongkhonsuk, J. Sudjapokinon, S. Adisakwattana, The
The authors’ responsibilities were as follows – KFH: per- association between online learning and food consumption and lifestyle
formed data collection and data analyses and wrote the first draft behaviors and quality of life in terms of mental health of undergraduate
students during COVID-19 restrictions, Nutrients 14 (4) (2022) 890.
of the manuscript; NIMFT: designed the research, wrote the first [18] M. Sinha, B. Pande, R. Sinha, Impact of COVID-19 lockdown on sleep-
draft, and reviewed the final content of the manuscript; NJ: wake schedule and associated lifestyle related behavior: a national
reviewed the final content of the manuscript; and all authors: survey, J Public Health Res 9 (3) (2020) 1826.
read and approved the final manuscript. [19] E.N.C. Manoogian, A. Chaix, S. Panda, When to eat: the importance of
eating patterns in health and disease, J Biol Rhythms 34 (6) (2019)
579–581.
References [20] S. Pengpid, K. Peltzer, Skipping breakfast and its association with health
risk behaviour and mental health among university students in 28
countries, Diabetes Metab Syndr Obes 13 (2020) 2889–2897.
[1] A. Flanagan, D.A. Bechtold, G.K. Pot, J.D. Johnston, Chrono-nutrition:
[21] S. Fanelli, C. Walls, C. Taylor, Skipping breakfast is associated with
from molecular and neuronal mechanisms to human epidemiology and
nutrient gaps and poorer diet quality among adults in the United States,
timed feeding patterns, J Neurochem 157 (1) (2021) 53–72.
Proc Nutr Soc 80 (OCE1) (2021) E48.
[2] R. Adafer, W. Messaadi, M. Meddahi, A. Patey, A. Haderbache, S. Bayen,
[22] M.C. Ruginski, Associations between nighttime eating and total caloric
et al., Food timing, circadian rhythm and chrononutrition: a systematic
intake in college-aged students. Master's thesis, University of Rhode
review of time-restricted eating’s effects on human health, Nutrients 12
Island, South Kingstown, RI, 2017.
(12) (2020) 3770.
[23] N. Goel, C. Hopkins, M. Ruggieri, Z. Zhang, D.M. Taylor, K.C. Allison,
[3] A.C. Veronda, K.C. Allison, R.D. Crosby, L.A. Irish, Development,
The impact of nighttime eating: A randomized controlled trial of
validation and reliability of the Chrononutrition Profile-Questionnaire,
daytime vs. delayed eating on weight and metabolism in adults of
Chronobiol Int 37 (3) (2020) 375–394.
normal weight, Sleep 42 (Suppl_1) (2019) A15.
[4] F.H. Mazri, Z.A. Manaf, S. Shahar, A.F. Mat Ludin, The association
[24] S.S. Maw, C. Haga, Effect of a 2-hour interval between dinner and
between chronotype and dietary pattern among adults: a scoping
bedtime on glycated haemoglobin levels in middle-aged and elderly
review, Int J Environ Res Public Health 17 (1) (2019) 68.
Japanese people: a longitudinal analysis of 3-year health check-up data,
[5] S. Almoosawi, S. Vingeliene, F. Gachon, T. Voortman, L. Palla,
BMJ Nutr Prev Health 2 (1) (2019) 1–10.
J.D. Johnston, et al., Chronotype: implications for epidemiologic studies
[25] J. Yoshida, E. Eguchi, K. Nagaoka, T. Ito, K. Ogino, Association of night
on chrono-nutrition and cardiometabolic health, Adv Nutr 10 (1) (2019)
eating habits with metabolic syndrome and its components: a
30–42.
longitudinal study, BMC Public Health 18 (1) (2018) 1–12.
[6] T. Arora, S. Taheri, Associations among late chronotype, body mass
[26] L.C. Ruddick-Collins, J.D. Johnston, P.J. Morgan, A.M. Johnstone, The
index and dietary behaviors in young adolescents, Int J Obes 39 (1)
Big Breakfast study: chrono-nutrition influence on energy expenditure
(2015) 39–44.
and bodyweight, Nutr Bull 43 (2) (2018) 174–183.
[7] G.P. Teixeira, K.C. Guimar~aes, A.G.N. Soares, E.C. Marqueze,
[27] M. Garaulet, P. G omez-Abellan, J.J. Alburquerque-Bejar, Y.-C. Lee,
C.R. Moreno, M.C. Mota, et al., Role of chronotype in dietary intake,
J.M. Ordovas, F.A. Scheer, Timing of food intake predicts weight loss
meal timing, and obesity: a systematic review, Nutr Rev (2022).
effectiveness, Int J Obes 37 (4) (2013) 604–611.
Available from: https://doi.org/10.1093/nutrit/nuac044, nuac044.