Timing-of-Energy-Intake-and-Ultra-Processed-Food-Consumpt_2026_Journal-of-th.pdf
RESEARCH
Research Paper
Timing of Energy Intake and Ultra-Processed
Food Consumption Are Associated With Obesity
in Adults in the United Kingdom: A Pooled
Cross-Sectional Analysis of the National Diet
and Nutrition Survey (2008-2019)
Patrícia de Fragas Hinnig, PhD; Fernanda Rauber, PhD; Kiara Chang, PhD; Eszter P. Vamos, PhD; Renata Bertazzi Levy, PhD
ARTICLE INFORMATION
Article history:
Submitted 12 June 2024
Accepted 8 December 2025
Keywords:
Circadian rhythm
Chrono-nutrition
Meals
Ultra-processed food
Obesity
Supplementary materials:
Tables 2, 4, 6, and 8 are available at www.
jandonline.org
2212-2672/Copyright ª 2025 by the Academy of
Nutrition and Dietetics.
https://doi.org/10.1016/j.jand.2025.156261
ABSTRACT
Background The distribution of energy intake and types of food consumed
throughout the day may influence health.
Objective This study investigated the association of energy intake and ultra-processed
food (UPF) consumption in the morning, at midday, and in the evening with indicators
of obesity.
Design A pooled cross-sectional analysis of the National Diet and Nutrition Survey
(2008-2019) of adults residing in England, Wales, Scotland, and Northern Ireland was
conducted. Food consumption was assessed using 4-day food diaries. UPFs intake was
assessed based on the Nova food classification. The 3 time periods were classified as
morning (5:00 AM to 10:59 AM), midday (11:00 AM to 4:59 PM), and evening (5:00 PM to
4:59 am).
Participants/Setting This study included 5749 adults aged 19 to 64 years residing in
private households in the United Kingdom.
Main Outcome Measures The main outcome measures were overweight, obesity, and
abdominal obesity. Body mass index (calculated as kg / m 2 ) values ≥25 and ≥30 were
used to classify overweight (including obesity) and obesity, respectively. Abdominal
obesity was defined as a waist circumference ≥102 cm for men and ≥88 cm for
women.
Statistical Analyses Performed Multivariable logistic regression assessed the asso-
ciations between the percentage of energy intake and the percentage of energy intake
from UPFs in the morning, at midday, and in the evening, and indicators of obesity.
Models were adjusted for sociodemographic variables (ie, sex, age, ethnicity, region,
year of the survey, and social class occupation), behavioral variables (ie, physical ac-
tivity, smoking status, frequency of alcohol consumption, sleep duration, and weight-
loss diet), and variables related to food consumption.
Results The highest percentage of energy intake in the morning (third tertile) was
associated with a 19% reduction in the odds of being overweight (95% CI, 0.68 to 0.96),
and the highest percentage in the evening was associated with a 21% higher odds of
being overweight (95% CI, 1.01 to 1.44). A 10% increase in the percentage of energy
intake from UPFs at midday and in the evening was associated with an 11% (95% CI, 1.02
to 1.20) and 10% (95% CI, 1.02 to 1.18) higher odds of overweight, respectively. The
highest percentage of energy intake from UPFs in the morning, at midday, and in the
evening (third tertile) was associated with a 26%, 25%, and 45% higher odds of obesity,
respectively. A 10% increase in the percentage of energy intake from UPFs in the
evening was associated with a 12% (95% CI, 1.03 to 1.22) higher odds of abdominal
obesity.
Conclusions The timing of energy intake was associated with overweight, and higher
consumption of UPFs at all times of day was associated with obesity. The strength of
the association between UPFs consumption and indicators of obesity was greater in the
evening compared with the morning and midday.
J Acad Nutr Diet. 2026;126(3):156261.
ª 2025 by the Academy of Nutrition and Dietetics.
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
1
Conclusions The timing of energy intake was associated with overweight, and lower
O
BESITY AFFECTS ALL COUNTRIES, AND THE NUM-
ber of people with overweight or obesity is ex-
pected to continue rising over the current decade. 1
According to the 2023 World Atlas of Obesity,
projections estimate that by 2035, approximately 24% of the
population 5 years and older will be living with overweight
or obesity. 1
In high-income countries, one of the primary factors
contributing to the obesity epidemic is believed to be the
increased availability and affordability of ultra-processed
foods (UPFs). 2 Findings from the National Diet and Nutri-
tion Survey (NDNS) in the United Kingdom revealed that
56.8% of dietary calories came from UPFs, and higher con-
sumption of these foods was associated with higher body
mass index (BMI), increased waist circumference (WC), and a
greater likelihood of obesity in adults and older adults. 3,4
Previous researchers have also reported that higher con-
sumption of UPFs was associated with obesity indicators .5-8
This indicates a strong association (odds ratio [OR] 1.55;
95% CI, 1.36 to 1.77) between higher consumption of UPFs
and an increased risk of adverse health outcomes, particu-
larly obesity. 8 These findings emphasized the need to
develop and evaluate public health strategies aimed at
reducing dietary exposures to UPFs, which may contribute to
improved population health outcomes.
Emerging research underscores the potential impact of the
timing of food consumption on obesity risk alongside other
characteristics of foods. 9 Recent studies suggested that the
timing and distribution of energy intake throughout the day
may play a role in influencing weight control. 10,11 The timing
of food consumption may influence physiological functions,
as it is associated with circadian rhythms, which regulate
various metabolic processes, including energy metabolism,
appetite, digestion, and absorption. 12,13
The impact of the timing of food consumption on obesity
remains poorly understood, and emerging epidemiologic
evidence suggests that irregular and concentrated eating
patterns, particularly at the end of the day, may adversely
affect weight control. 14 Studies among adults have suggested
that a higher percentage of energy intake at night was
associated with an increased likelihood of being overweight
or obese. 10,15 However, further studies are needed to
examine this hypothesis. A systematic review with meta-
analysis found no significant association between BMI and
evening energy intake, although the authors noted signifi-
cant heterogeneity among studies and a high risk of bias as
limitations. 16 Understanding these associations may help
generate hypotheses for future studies and provide insights
that could eventually support the refinement of dietary
guidance. 17
To the best of our knowledge, no previous studies have
provided information on the distribution of UPFs con-
sumption throughout the day and its association with
adiposity indicators. Therefore, the aim of this study was to
assess the association of energy intake and UPFs con-
sumption at different times of the day with indicators of
obesity in the adult population in the United Kingdom
(2008-2019). The study hypothesis was that higher energy
intake and UPFs consumption in the evening, compared
with lower intake during this time of day, are associated
with higher odds of being overweight, obese, and having
abdominal obesity.
METHODS
This was a cross-sectional analysis using data from the NDNS
spanning the years 2008 to 2019. The NDNS is a continuous
program of cross-sectional surveys designed to evaluate the
diet, nutrient intake, and overall nutritional status of in-
dividuals aged 1.5 years and older residing in private
households across England, Wales, Scotland, and Northern
Ireland. 18-21
Data across the years 2008 to 2019 were aggregated in this
study to maximize the sample size. Households were
randomly selected from a comprehensive list of all residential
addresses in the United Kingdom, based on UK postcode data.
To improve cost-effectiveness, addresses were first grouped
into primary sampling units—small geographical areas based
on postcode sectors—randomly drawn from across the United
Kingdom. Within each primary sampling unit, a list of ad-
dresses was randomly selected. At each selected address, in-
terviewers enumerated the households, and if more than 1
was present, randomly chose 1 to participate. From the cho-
sen household, the interviewer randomly selected up to 1
adult (aged 19 years or older) and 1 child (aged 1.5-18 years)
to participate in the survey. Details of the sample and sam-
pling are described in previous publications.
18-21
For this study, we included all participants aged 19-64
years (n = 5767) who had complete data on weight and
height and at least 3 food records. It has been suggested that
aging is accompanied by attenuated circadian rhythms and
altered sleep patterns. For this reason, we excluded older
adults from the sample (65 years and older). 22 Individuals
with implausible energy intakes (<1st or >99th percentiles
across diary days 23 ; n = 13) and pregnant or breastfeeding
women (n = 5) were also excluded. The final analytical
sample included 5749 adults with valid weight and height
data, which were used to define overweight and obesity
status. Of these, >98% completed all 4 food diary days. A
subsample of 4244 adults with valid WC measurements was
used to assess abdominal obesity.
Data files for the present study were obtained under li-
cense from the UK Data Archive. Ethical approval for the
study was obtained from the Oxfordshire A Research Ethics
Committee and from the Cambridge South National Research
RESEARCH SNAPSHOT
Research Question: What is the association between energy
intake and ultra-processed food consumption at different
times of the day (morning, midday, and evening) and
indicators of obesity (overweight, obesity, and abdominal
obesity)?
Key Findings: In a pooled cross-sectional analysis (2008-
2019) of 5749 adults in the United Kingdom, higher energy
intake in the morning (5:00 AM to 10:59 AM) was significantly
associated with lower odds of being overweight, and higher
energy consumption in the evening (5:00 PM to 4:59 AM), was
significantly associated with higher odds of overweight. The
highest percentage of energy intake from ultra-processed
food in the morning, midday, and evening was significantly
associated with 26%, 25%, and 45% higher odds of obesity,
respectively.
RESEARCH
2
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
■■ 2026 Volume 126 Number 3
To the best of our knowledge, some previous studies have
Ethics Service Committee thereafter. All adults aged 16 years
and older completed and signed the adult consent form.
The NDNS aimed to maintain consistency in data collec-
tion and analysis methods over time. The processing and
analytical procedures remained largely unchanged; however,
when modifications were necessary, calibration studies were
conducted to ensure the results remained comparable over
the years. 20
Dietary Assessment
To collect food consumption data, participants were
instructed to complete 4 consecutive food diaries within
each 24-hour period. The study provided adults with a diary
in paper format.
The survey was designed so that all days of the week were
evenly represented. The food diary could start on any day of
the week and run for 4 consecutive days. The participants
were asked to record all food and drink consumed both in-
side and outside the home, including the timing of food
intake. Portion sizes were estimated using a food photog-
raphy atlas illustrating a range of portion sizes for commonly
consumed foods. To validate the estimates of energy intake
from food diaries, the NDNS Rolling Programme included a
sub-study with participants aged 4 years and older using
doubly labeled water. 24
The completed diaries were reviewed by interviewers with
the respondents, and any entries lacking details regarding
the quantity or food preparation method were clarified or
completed during the interview. Although the completed
diaries were reviewed with respondents, some details could
not be retrieved because participants were unable to recall
exact portion sizes, brand names, or preparation methods.
Consequently, during dietary data processing, missing in-
formation was handled using standardized procedures, such
as assigning default codes or estimating portion sizes based
on other diary days, age-appropriate averages, or similar
foods. Where reliable information was not available for some
nutrients, values for such foods were obtained by extrapo-
lating from data for similar foods. For homemade dishes and
manufactured products, nutrients were calculated from their
constituents using the Diet In, Nutrients Out (DINO) system,
version 2005, 25 which allows adjustments to be made for
weight and vitamin losses during cooking. 26-29
Dietary data were processed using the DINO system, 25
with nutrient intakes estimated based on the NDNS
Nutrient Databank, updated for each survey year. 30,31 The
DINO system’s time-related database updates analytic data
while maintaining nutrient profiles consistent with the
version used for data collection. Details on the annual up-
dates to the Nutrient Databank and on missing data handling
are published for each NDNS survey. 26-29 For example, in
year 10 (2017-2018), the reviews of manufacturers’ and label
data focused on soft drinks, breakfast cereals, biscuits, buns,
cakes, pastries, fruit pies, and dairy desserts. In addition,
sodium values were updated for products such as bread,
soups, baked beans, and crisps. 29 Food composition values
are checked for accuracy and quality, including the validity of
the analytical method used, before being incorporated into
the DINO system. Further details of the system were pub-
lished by Fitt and colleagues. 25 The DINO system contains 54
nutrient values and energy for a wide range of almost 6000
foods 25 and enables the detailed input of individual foods
and complex recipes. For manufactured products not present
in DINO, nutrient information was obtained from super-
markets, manufacturer websites, or direct contact with
producers before creating new food codes and portion
weights. 29
All recorded food items were categorized by trained re-
searchers according to the Nova food classification sys-
tem, 32,33 which categorizes foods based on the nature,
extent, and purpose of their industrial processing. This
classification system 32 comprises 4 groups: (1) unprocessed
or minimally processed foods; (2) processed culinary in-
gredients; (3) processed foods; and (4) UPFs—the primary
focus of this study. UPFs are products resulting from the
processing of various foods, made predominantly from food-
derived substances and cosmetic additives, with minimal or
no whole foods. 33 The processing methods used for these
foods typically involve salting, the addition of sugar, and
industrial techniques such as extrusion, molding, and pre-
frying. In addition, additives are incorporated to enhance the
palatability or hyperpalatability of the final product. 33 These
foods are often packaged in synthetic materials and designed
to be durable, affordable, convenient, and visually
appealing. 33 Examples of UPFs include soft drinks, packaged
snacks, biscuits, instant noodles, sausages, fast food, and
ready-made or preprepared frozen meals. 33 Further details
regarding food categorization methods can be found in
Rauber and colleagues. 4,34 The categorization was followed
by a review of these classifications by other experts within
the research group. These classifications were then exten-
sively discussed in periodic meetings with the team, as well
as with a group of British researcher-partners—who are co-
authors of this study—with expertise in local foods.
The dietary data were divided into time periods: morning
(5:00 AM to 10:59 AM), midday (11:00 AM to 4:59 PM), and
evening (5:00 PM to 4:59 AM). Then, the mean percentage of
energy intake and the percentage of energy intake from UPFs
(as a percentage of total daily energy intake) from the 4 food
diaries for each time period and for each individual were
calculated. These time slots were chosen based on circadian
patterns of food intake, appetite, digestion, and metabolism,
which are influenced by the light-dark cycle. 12,35 The eve-
ning was defined as starting at 5:00 PM, aligning with the end
of the workday for most adults in the United Kingdom,
where the main meal (dinner or supper) is typically
consumed in the early evening. 36 This timeframe also con-
siders individuals who may have late-night snacks and
meals. Previous researchers have also defined the start of
evening eating habits as 05:00 PM. 15,37,38
Outcomes Assessments
Weight and height were measured on 1 occasion by trained
researchers using a portable stadiometer and scales. For
weight and height measurements, participants stood facing
forward, with their feet together, arms at their sides, and
heads positioned according to the Frankfort plane. 39 WC was
measured at the midpoint between the iliac crest and the
last rib. 40 The BMI was calculated by dividing weight by
height squared (kg / m 2 ), and values ≥25 and ≥30 were
classified as overweight and obesity, respectively, according
to World Health Organization criteria. 41 Abdominal obesity
RESEARCH
■■ 2026 Volume 126 Number 3
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
3
was defined as a WC ≥ 102 cm for males and ≥ 88 cm for
females. 42
Covariates
Participants completed interviews that involved recording
sociodemographic variables and characteristics related to
health behaviors.
Sociodemographic characteristics included in analyses
were: sex (male and female), age (19-34 years, 35-49 years,
50-64 years), ethnicity (White, Black, Asian, mixed
[including White and Black Caribbean, White and Black Af-
rican, White and Asian], and other ethnic groups, which
referred to ethnicities not included in the listed categories),
region (England North, England Central/Midlands, England
South [including London], Scotland, Wales, Northern
Ireland), year of the survey (2008-2019) and social class
occupation. The social class occupation was reported ac-
cording to the National Statistics Socio-Economic Classifi-
cation 43 (routine and manual occupations, intermediate
occupations, lower managerial and professional occupations,
higher managerial and professional occupations). The health
behaviors were: physical activity assessed by the Recent
Physical Activity Questionnaire 44 (hours per day of moderate
or vigorous physical activity 45 classified by quartile, where
the first quartile representing the least physically active in-
dividuals and the fourth quartile representing the most
physically active), smoking status (nonsmoker, smoker),
frequency of alcohol consumption (≥3 times/wk, 1-2 times/
wk, <1 time/wk, never), sleep duration (hours per day,
classified as <7 h/d, 7-8 h/d, >8 h/d), weight-loss diet (yes,
no), and total energy intake (kilocalories per day) derived
from the food diaries.
The selection of covariates was based on a previous study
conducted using the same sample (2008-2016), which
aimed to evaluate the association between overall daily
UPFs consumption and obesity. Alcohol consumption was
included as a covariate due to its association with late
chronotype and related unhealthy behaviors reported in
previous studies. 46
Multiple imputation by chained equations 47 was per-
formed with 10 imputed copies of data for social class
occupation (2.7% missing), ethnicity (0.1%), physical activity
(10.7%), frequency of alcohol consumption (8.6%), sleep
duration (3.9%), smoking status (0.3%), and weight-loss diet
(16.6%). Sensitivity analysis was conducted comparing find-
ings from imputed data and complete case analysis.
Exposure variables included the percentage of energy
intake and the percentage of energy intake from UPFs (as a
percentage of total energy intake) at each time period of the
day. The outcomes assessed were overweight, obesity, and
abdominal obesity (yes, no). A separate model was run for
each time period of the day.
Statistical Analysis
Descriptive Analyses. Categorical variables were summa-
rized as relative frequencies with SEs, and quantitative var-
iables as means with SE. Differences in the prevalence of
overweight, obesity, and abdominal obesity according to
sociodemographic and behavioral variables were assessed
using the χ 2 test.
Differences in the mean percentage of energy intake and
energy from UPFs during morning, midday, and evening
according to sociodemographic and behavioral variables
were examined using univariate linear regression.
Although prevalence, mean values, and SEs are presented
for descriptive purposes, P values were derived from
regression models accounting for the complex survey
design.
Primary Analyses. Percentages of total energy intake and
energy from UPFs within each time slot were categorized
into tertiles (low, medium, high). Multivariable logistic
regression estimated the odds of overweight, obesity, or
abdominal obesity in the highest vs lowest tertile. Linear
trend was tested by modeling tertiles as ordinal variables.
In additional analyses, exposures were modeled as contin-
uous variables, representing a 10% increase in the per-
centage of energy or UPFs intake per time period.
Collinearity among the independent variables was
assessed using tolerance and variance inflation factor values.
All tolerance values were >0.2, and all variance inflation
factors
were
<2.5,
indicating
no
evidence
of
multicollinearity. 48
Multivariable Modeling Strategy
Models for Percentage of Total Energy Intake. A for-
ward selection strategy was used to build the multivariable
models in conceptual blocks:
•
Model 1: sociodemographic variables (sex, age,
ethnicity, region, year, social class occupation)
•
Model 2: model 1 plus behavioral variables (physical
activity, smoking, alcohol frequency, sleep duration,
weight-loss diet)
•
Model 3 (final): model 2 plus total daily energy intake
Model 3 was considered the final model of interest
because it best represents these associations (including all
the adjustment variables) and also to verify this relationship
independent of the total energy intake consumed, as carried
out by other studies. 10,49,50
Sensitivity analyses were also conducted:
•
Model 4: model 2 plus the percentage of total energy
from UPFs throughout the entire day
•
Model 5: model 3 plus the percentage of total energy
from UPFs throughout the entire day
•
Model 6: model 3 plus percentage energy from sugar,
percentage energy from saturated fat, and daily fruit/
vegetable intake (in grams).
Models for Percentage of Energy From UPFs. For the
analyses with percentage of energy from UPFs as exposure,
models followed a similar sequence:
•
Model 1 and 2: as above
•
Model 3 (final): Model 2 plus percentage of UPFs en-
ergy from other time periods (eg, for morning, UPFs
percentage from midday plus evening)
Model 3 was defined as the final model of interest because
it best represents these associations and allows for verifi-
cation of the relationship independent of the percentage of
RESEARCH
4
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
■■ 2026 Volume 126 Number 3
Table 1. Prevalence of overweight, obesity, and abdominal obesity according to sociodemographic and behavioral variables of
the UK population (aged 19 to 64 years), NDNS a 2008-2019 (n = 5749)
Characteristic
Total sample,
unweighted n
All, %
Prevalence of
overweight
Prevalence of
obesity
Prevalence of
abdominal obesity
←̅̅̅̅̅̅̅̅̅̅̅̅̅weighted % b (SE)̅̅̅̅̅̅̅̅̅̅̅̅̅→
Total
—
100
61.3 (0.8)
24.7 (0.7)
35.9 (0.9)
Sex
Male
2371
49.9
66.8 (1.1)***
23.8 (1.0)
31.1 (1.3)***
Female
3378
50.1
55.7 (1.1)
25.7 (0.9)
40.7 (1.2)
Age group
19-34 y
1684
34.8
46.6 (1.5)***
16.9 (1.1)***
21.2 (1.4)***
35-49 y
2176
34.6
65.9 (1.3)
24.8 (1.1)
34.1 (1.4)
50-64 y
1889
30.5
72.8 (1.3)
33.6 (1.4)
52.2 (1.5)
Ethnicity
White
5228
86.8
61.4 (0.9)
25.0 (0.7)
36.4 (1.0)
Black
137
3.3
62.5 (4.9)
32.3 (4.5)
34.7 (5.7)
Asian
235
6.6
63.1 (3.4)
20.9 (3.1)
33.5 (3.8)
Mixed c
56
1.3
50.0 (8.1)
21.9 (5.5)
22.6 (5.8)
Any other group d
86
2.0
53.9 (6.2)
15.8 (4.2)
29.1 (6.7)
Missing
7
0.1
96.5 (4.2)
11.7 (11.1)
0
Region
England North
1000
22.8
65.0 (1.6)***
27.0 (1.5)***
39.8 (1.8)***
England Central/Midlands
738
16.3
66.8 (2.0)
28.9 (2.0)
37.3 (2.5)
England South (including London)
1742
45.1
56.9 (1.3)
20.8 (1.1)
31.0 (1.3)
Scotland
826
8.5
60.6 (2.5)
28.7 (2.2)
43.3 (2.9)
Wales
657
4.6
65.3 (2.5)
29.7 (2.2)
43.1 (2.5)
Northern Ireland
786
2.8
63.2 (1.9)
25.8 (1.8)
42.7 (2.8)
Social class occupation e
Higher managerial and professional
occupations
962
18.9
55.6 (1.9)**
18.4 (1.5)***
29.9 (1.8)***
Routine and manual occupations
1983
32.1
63.1 (1.5)
28.5 (1.3)
40.4 (1.7)
Intermediate occupations
1159
19.6
63.1 (1.7)
25.7 (1.6)
37.9 (1.9)
Lower managerial and professional
occupations
1476
26.8
62.8 (1.6)
23.8 (1.4)
33.7 (1.7)
Missing
169
2.7
50.4 (4.8)
25.6 (4.4)
34.9 (5.9)
Physical activity f
First quartile
1423
22.4
66.6 (1.7)**
32.5 (1.5)
47.1 (2.0)***
Second quartile
1338
22.3
57.9 (1.7)
23.8 (1.5)
35.3 (1.7)
Third quartile
1183
22.3
57.9 (1.8)
20.1 (1.5)
31.7 (1.7)
Fourth quartile
1138
22.3
62.9 (1.7)
21.9 (1.4)
30.2 (1.8)
Missing
667
10.7
60.6 (2.7)
25.9 (2.3)
36.5 (3.0)
Smoking status
Nonsmoker g
4304
77.5
62.7 (0.9)**
25.2 (0.8)
36.2 (1.0)
Smoker
1434
22.2
56.6 (1.7)
23.1 (1.4)
34.9 (1.8)
Missing
11
0.3
38.6 (1.6)
14.9 (13.5)
13.7 (12.5)
(continued on next page)
RESEARCH
■■ 2026 Volume 126 Number 3
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
5
energy intake from UPFs consumed during other periods of
the day.
Sensitivity analyses also conducted:
•
Model 4: model 2 plus total daily energy intake
•
Model 5: model 3 plus total daily energy intake
•
Model 6: model 3 plus percentage energy from sugar,
percentage energy from saturated fat, and daily fruit
and vegetable intake
Temporal Interactions. Interaction terms between expo-
sures (both energy and UPF) and survey year tested potential
temporal variation.
Additional Analyses. To investigate whether the timing of
energy intake is a critical factor associated with obesity in-
dicators, we performed additional multivariable logistic
regression analyses. We calculated the ratio of energy
consumed in the evening to that consumed in the morning
and midday, categorized as <1.0 (reference), 1.00 to 1.49, 1.50
to 1.99, and ≥2.0. The regression was adjusted for socio-
demographic variables, behavioral variables, and daily en-
ergy intake. The same approach was applied using UPFs’
energy ratios.
All statistical analyses were performed using Stata soft-
ware, version 16.1. 51 Survey weights provided by NDNS were
applied in all analyses to account for differential probabili-
ties, ensuring nationally representative estimates. P value <
.05 was considered statistically significant.
Because standard goodness-of-fit tests (eg, Hos-
mer—Lemeshow) are limited for multiply imputed complex
survey data, model adequacy was assessed based on the
stability of ORs across imputations. The consistency of these
estimates supports the robustness of the models.
RESULTS
Approximately one-half of the sample were women (50.1%),
aged 19 to 34 years (34.8%), White (86.8%), from the South
of England (45.1%), and working in routine and manual
occupations (32.1%). Regarding behavioral habits, physical
activity levels were fairly evenly distributed across quar-
tiles, with a slightly higher proportion in the first quartile
(22.4%); most participants were nonsmokers (77.5%),
consumed alcohol once or twice per week (29.3%), slept 7 to
8 hours/d (47.4%), and did not follow a weight-loss diet
(67.4%) (Table 1).
Table 1. Prevalence of overweight, obesity, and abdominal obesity according to sociodemographic and behavioral variables of
the UK population (aged 19 to 64 years), NDNS a 2008-2019 (n = 5749) (continued)
Characteristic
Total sample,
unweighted n
All, %
Prevalence of
overweight
Prevalence of
obesity
Prevalence of
abdominal obesity
Frequency of alcohol consumption
≥3 times/wk
1217
22.3
61.9 (1.7)
20.3 (1.4)***
33.6 (1.8)***
1-2 times/wk
1706
29.3
59.2 (1.6)
22.8 (1.3)
31.8 (1.6)
<1 time/wk
1781
28.6
62.9 (1.5)
30.1 (1.4)
40.6 (1.7)
Never
637
11.2
64.0 (2.4)
28.5 (2.4)
44.4 (2.9)
Missing
408
8.6
57.2 (2.5)
20.0 (2.1)
30.6 (2.9)
Sleep duration
<7 h/d
1871
31.8
68.3 (1.3)***
30.8 (1.3)***
42.6 (1.6)***
7-8 h/d
2689
47.4
59.8 (1.2)
21.8 (0.9)
32.6 (1.2)
>8 h/d
980
16.9
52.5 (2.1)
21.9 (1.6)
33.4 (2.1)
Missing
209
3.9
59.8 (4.2)
24.1 (3.7)
30.1 (5.0)
Weight-loss diet
No
3735
67.4
56.0 (1.0)***
20.2 (0.8)***
31.2 (1.1)***
Yes
937
16.0
85.4 (1.4)
44.3 (2.0)
56.7 (2.3)
Missing
1077
16.6
59.5 (1.8)
24.2 (1.8)
35.2 (2.1)
a NDNS = National Diet and Nutrition Survey.
b All percentages are weighted using NDNS survey weights to account for sampling design.
c White and Black Caribbean, White and Black African, White and Asian.
d Ethnicities not included in the listed categories.
e Routine and manual occupations: jobs that involve repetitive tasks and are often regulated by a basic labor contract; intermediate occupations: are considered to be more difficult
conceptually than routine manual and nonmanual jobs, but less difficult than managerial or professional jobs (clerical, sales, services); lower managerial and professional occupations:
positions that have an attenuated form of service relationship. Employees in these groups generally plan and supervise operations on behalf of the employer under the direction of senior
managers; higher managerial and professional occupations: positions that involve general planning and supervision of operations, and that require specialized educational training. 43
f First quartile represents the least physically active individuals, and the fourth quartile represents the most physically active.
g Former cigarette smoker and never regular cigarette smoker.
**χ 2 test, P < .010.
***χ 2 test. P < .001.
RESEARCH
6
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
■■ 2026 Volume 126 Number 3
UK adults had mean (SD) daily energy intake of 1856 (9.5)
kcal, and 54.7% (0.3%) of energy came from UPFs (data not
shown). The prevalence of overweight, obesity, and abdom-
inal obesity was 61.3%, 24.7%, and 35.9%, respectively
(Table 1).
The prevalence of overweight was higher in men, older
individuals, individuals of lower social class, those who had
lower levels of physical activity, nonsmokers, and adults
following a weight-loss diet. Obesity was more prevalent
among those who were older and followed a weight-loss
diet. The prevalence of abdominal obesity was higher in
women, older adults, those with lower levels of physical
activity, and adults following a weight-loss diet (Table 1).
The percentage of energy intake and energy intake from
UPFs were highest in the evening, followed by midday and
morning (Figure). This information, according to character-
istics of the sample, is presented in Table 2 (available at
www.jandonline.org).
For overweight (Table 3), the highest percentage of energy
intake in the morning was associated with lower odds of
overweight and showed a significant linear trend (OR 0.81;
95% CI, 0.68 to 0.96). A 10% increase in the percentage of
energy intake in the morning was associated with a reduc-
tion of 11% in the odds of overweight (OR 0.89; 95% CI, 0.82
to 0.96). For midday, there was no evidence of a significant
association between energy intake and overweight. In eve-
ning consumption, the highest percentage of energy intake
was associated with greater odds of overweight, and a
significant linear trend was observed (OR 1.21; 95% CI, 1.01 to
1.44). The results for overweight remained similar in the
sensitivity analysis (models 4, 5, and 6) (Table 4, available at
www.jandonline.org).
Regarding UPFs and overweight (Table 3), a 10% increase in
the percentage of energy intake from UPFs at midday was
associated with 11% higher odds of overweight (OR 1.11; 95%
CI, 1.02 to 1.20). The highest percentage of energy from UPFs
in the evening was associated with a higher odds of over-
weight, and a significant linear trend was also observed (OR
1.31; 95% CI, 1.08 to 1.60. The relationship remained signifi-
cant when the variable was analyzed as continuous (OR 1.10;
95% CI, 1.02 to 1.18) (Table 3). The results for overweight
persisted in the sensitivity analysis (models 4, 5, and 6)
(Table 4).
For obesity (Table 5), no evidence of significant association
was found for the percentage of energy intake during the 3
periods of the day. In the sensitivity analysis (models 4, 5,
and 6), the results were consistent with model 3 (Table 6,
available at www.jandonline.org). Considering the UPFs
(Table 5), a significant linear trend was observed for the
association between tertile of energy consumption from
UPFs in the morning and midday and obesity in model 3 (OR
1.26; 95% CI, 1.04 to 1.53; OR 1.25; 95% CI, 1.04 to 1.51,
respectively). For these models, a similar result was observed
when the variable was treated as continuous. Considering
the evening period, the highest percentage of energy intake
from UPFs was associated with higher odds of having obesity
50
47.6f
45
Morning
40
Midday
35
30
Evening
25
20
15
10
5
0
% of total energy intake
% of total energy intake from UPFs
% of total energy intake
Figure. Mean proportions of energy intake and ultra-processed intake (% of total daily energy intake) consumed in three periods
of the day a (morning, midday and evening) in the UK b population aged 19 to 64 years (NDNS c , 2008—2019) (n = 5 749).
a Morning (5:00 am to 10:59 am), Midday (11:00 am to 4:59 pm), Evening (5:00 pm to 4:59 am) b United Kingdom; c National Diet
and Nutrition Survey d Morning 95% CI (16.9; 17.4%) e Midday 95% CI (34.8; 35.6%) f Evening 95% CI (47.2; 48.1) g Morning 95% CI
(10.2; 10.6%) h Midday 95% CI (20.6; 21.2%) i Evening 95% CI (22.7; 23.6%)
RESEARCH
■■ 2026 Volume 126 Number 3
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
7
UK adults had mean (SD) daily energy intake of 1856 (9.5) negligible linear trend was observed (OR 1.21; 95% CI, 1.01 to
Table 3. OR a (95% CI) of the association between tertiles and a 10% increase in energy intake and energy intake from
ultra-processed foods across daytime periods, b and overweight status c in UK adults (aged 19-64 years), NDNS d 2008-2019
(n = 5749)
Variable
Energy Consumption (% of Total Energy) in Each Period of the Day
Tertiles e
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First
Second
Third
←̅̅̅̅̅̅̅̅̅̅̅̅OR (95% CI)̅̅̅̅̅̅̅̅̅̅̅̅→
Period of the day
Morning
Unadjusted model
1
0.94 (0.80 to 1.11)
1.00 (0.86 to 1.17)
.948
0.99 (0.92 to 1.06)
Model 1 g,h
1
0.87 (0.74 to 1.04)
0.89 (0.76 to 1.04)
.154
0.93 (0.86 to 1.00)
Model 2 g,i
1
0.85 (0.71 to 1.01)
0.81 (0.68 to 0.96)
.014
0.89 (0.82 to 0.96)
Model 3 g,j
1
0.87 (0.72 to 1.04)
0.81 (0.68 to 0.96)
.015
0.89 (0.82 to 0.96)
Midday
Unadjusted model
1
0.70 (0.60 to 0.83)
0.76 (0.65 to 0.90)
.001
0.94 (0.89 to 0.99)
Model 1 g,h
1
0.83 (0.70 to 0.98)
0.92 (0.78 to 1.09)
.372
0.99 (0.94 to 1.06)
Model 2 g,i
1
0.85 (0.71 to 1.01)
0.95 (0.80 to 1.13)
.614
1.00 (0.94 to 1.07)
Model 3 g,j
1
0.87 (0.73 to 1.03)
0.96 (0.81 to 1.15)
.693
1.00 (0.94 to 1.07)
Evening
Unadjusted model
1
1.13 (0.96 to 1.33)
1.22 (1.04 to 1.43)
.014
1.06 (1.00 to 1.12)
Model 1 g,h
1
1.13 (0.95 to 1.36)
1.15 (0.97 to 1.36)
.099
1.04 (0.98 to 1.10)
Model 2 g,i
1
1.20 (0.99 to 1.44)
1.21 (1.01 to 1.45)
.034
1.06 (0.99 to 1.13)
Model 3 g,j
1
1.21 (1.01 to 1.46)
1.21 (1.01 to 1.44)
.037
1.06 (0.99 to 1.13)
Variable
Consumption of Ultra-Processed Foods (% of Total Energy)
Tertiles k
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First
Second
Third
←̅̅̅̅̅̅̅̅̅̅̅̅OR (95% CI)̅̅̅̅̅̅̅̅̅̅̅̅→
Period of the day
Morning
Unadjusted model
1
0.89 (0.76 to 1.05)
1.02 (0.86 to 1.20)
.832
1.02 (0.92 to 1.12)
Model 1 l,h
1
0.87 (0.74 to 1.02)
0.93 (0.79 to 1.10)
.403
0.95 (0.86 to 1.06)
Model 2 l,i
1
0.82 (0.69 to 0.98)
0.87 (0.73 to 1.03)
.111
0.91 (0.82 to 1.02)
Model 3 l,m
1
0.85 (0.72 to 1.01)
0.93 (0.77 to 1.12)
.444
0.96 (0.86 to 1.08)
Midday
Unadjusted model
1
0.78 (0.67 to 0.93)
0.94 (0.79 to 1.11)
.452
1.00 (0.94 to 1.07)
Model 1 l,h
1
0.86 (0.72 to 1.02)
1.12 (0.94 to 1.34)
.198
1.08 (1.00 1.16)
Model 2 l,i
1
0.89 (0.74 to 1.07)
1.17 (0.97 to 1.41)
.094
1.10 (1.01 to 1.19)
Model 3 l,n
1
0.89 (0.74 to 1.07)
1.19 (0.98 to 1.44)
.068
1.11 (1.02 to 1.20)
Evening
Unadjusted model
1
0.99 (0.83 to 1.16)
1.17 (0.99 to 1.39)
.069
1.05 (0.99 to 1.11)
Model 1 l,h
1
1.01 (0.85 to 1.20)
1.28 (1.07 to 1.53)
.008
1.08 (1.02 to 1.15)
(continued on next page)
RESEARCH
8
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
■■ 2026 Volume 126 Number 3
(OR 1.45; 95% CI, 1.20 to 1.76), and a significant linear trend
was observed. Furthermore, a 10% increase in the percentage
of energy intake from UPFs in the evening was associated
with 16% higher odds of overweight (OR 1.16; 95% CI, 1.07 to
1.26) (Table 5).
The magnitude of the effect of the highest percentage of
energy intake from UPFs on obesity was greatest in the
evening compared with midday and morning, with corre-
sponding increases in the odds of obesity of 45% (OR 1.45;
95% CI, 1.20 to 1.76), 25% (OR 1.25; 95% CI, 1.04 to 1.51), and
26% (OR 1.26; 95% CI, 1.04 to 1.53), respectively (Table 5).
In sensitivity analysis (Table 6), the association between
percentage energy intake from UPFs and obesity remained
significant only for the evening period after adjusting for
total energy in model 4.
In relation to abdominal obesity (Table 7), no association
was observed for the percentage of energy intake during the
periods of the day analyzed. In the sensitivity analysis, the
results were consistent with model 3 (Table 8, available at
www.jandonline.org). When considering the percentage
energy intake from UPFs (Table 7), a statistically significant
association was observed for the evening period only (OR
1.31; 95% CI, 1.07 to 1.61), with a significant linear trend
observed. A 10% increase in the percentage of energy intake
from UPFs in the evening was associated with 12% higher
odds of abdominal obesity (OR 1.12; 95% CI, 1.03 to 1.22)
(Table 7). The results for abdominal obesity remained similar
in the sensitivity analysis (Table 8).
After testing for interaction between the percentage of
energy intake and from UPFs during the periods of the day
and the year of data collection, no significant interaction was
observed (data not shown).
Sensitivity analysis performed, including complete cases
only, indicated that the results of the multiple imputations
did not differ significantly from the complete case analysis
(data not shown).
A positive association was observed between the highest
category of energy intake timing ratio and overweight (OR
1.34; 95% CI, 1.09 to 1.66). Similarly, regarding UPFs intake,
participants with an evening-to-morning/midday energy
ratio from UPFs >2.0 had higher odds of both obesity (OR
1.22; 95% CI, 1.03 to 1.47) and overweight (OR 1.22; 95% CI,
1.03 to 1.46) (data not shown in tables).
DISCUSSION
We examined the association between the dietary percent-
age of energy intake and the percentage of energy intake
from UPFs during the morning, midday, and evening with
indicators of adiposity. Three notable findings emerged from
the analyses: (1) higher UPFs intake at any mealtime is
associated with higher odds of obesity; (2) higher UPFs
intake in the evening was linked to greater odds of obesity
compared with other times, and it was also linked to over-
weight and abdominal obesity; (3) energy intake in the
morning, regardless of the level of food processing, was
associated with a lower likelihood of overweight, and the
opposite was observed for the evening period.
These findings suggest that higher energy intake in the
morning is associated with a reduced likelihood of overweight.
A clinical trial investigating the effect of redistributing total
energy intake on weight loss reported greater weight loss in
women with overweight and obesity when more energy was
consumed in the morning compared with the evening. 52
Table 3. OR a (95% CI) of the association between tertiles and a 10% increase in energy intake and energy intake from
ultra-processed foods across daytime periods, b and overweight status c in UK adults (aged 19-64 years), NDNS d 2008-2019
(n = 5749) (continued)
Variable
Consumption of Ultra-Processed Foods (% of Total Energy)
Tertiles k
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First
Second
Third
Model 2 l,i
1
1.04 (0.87 to 1.25)
1.29 (1.06 to 1.56)
.011
1.09 (1.01 to 1.16)
Model 3 l,o
1
1.04 (0.87 to 1.25)
1.31 (1.08 to 1.60)
.008
1.10 (1.02 to 1.18)
a OR = odds ratio.
b Morning: 5:00 AM to 10:59 AM, midday: 11:00 AM to 4:59 PM, evening: 5:00 PM to 4:59 AM.
c Overweight: body mass index ≥25.
d NDNS = National Diet and Nutrition Survey.
e The tertiles were categorized based on cutoff points for energy consumption: Morning: first tertile: ≤13.1% of total energy intake; second tertile: 13.2%-20.6% of total energy intake; third
tertile: ≥20.7% of total energy intake. Midday: first tertile: ≤30.7% of total energy intake; second tertile: 30.8%-39.6% of total energy intake; third tertile: ≥39.7% of total energy intake.
Evening: first tertile: ≤42.8% of total energy intake; second tertile: 42.9%-52.1% of total energy intake; third tertile: ≥52.2% of total energy intake.
f P value for linear trend across tertiles of dietary contribution.
g Adjustment model for energy consumption.
h Model 1: adjusted for sex, age, region, survey year, and social class occupation.
i Model 2: Model 1 and physical activity, smoking, sleep duration, following a weight-loss diet, and frequency of alcohol consumption.
j Model 3: Model 2 and total energy intake.
k For ultra-processed foods intake: Morning: first tertile: ≤7.2% of total energy intake; second tertile: 7.3%-12.6% of total energy intake; third tertile: ≥12.7% of total energy intake. Midday:
first tertile: ≤16.3% of total energy intake; second tertile: 16.4%-24.5% of total energy intake; third tertile: ≥24.6% of total energy intake. Evening: first tertile: ≤17.1% of total energy
intake; second tertile: 17.2%-27.1% of total energy intake; third tertile: ≥27.2% of total energy intake.
l Adjustment model for ultra-processed foods consumption.
m Model 3: Model 2 and percentage of energy from ultra-processed foods in the midday and evening.
n Model 3: Model 2 and percentage of energy from ultra-processed foods in the morning and evening.
o Model 3: Model 2 and percentage of energy from ultra-processed foods in the morning and midday.
RESEARCH
■■ 2026 Volume 126 Number 3
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
9
Table 5. OR a (95% CI) of the association between tertiles and a 10% increase in energy intake and energy intake from ultra-
processed foods, across daytime periods, b and obesity status c in UK adults (aged 19-64 years), NDNS d 2008-2019 (n = 5749)
Variable
Energy Consumption (% of Total Energy) in Each Period of the Day
Tertiles e
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First
Second
Third
←̅̅̅̅̅̅̅̅̅̅̅̅OR (95% CI)̅̅̅̅̅̅̅̅̅̅̅̅→
Period of the day
Morning
Unadjusted model
1
0.97 (0.82 to 1.16)
1.17 (0.98 to 1.40)
.082
1.06 (0.98 to 1.14)
Model 1 g,h
1
0.91 (0.76 to 1.08)
1.07 (0.90 to 1.29)
.427
1.01 (0.93 to 1.10)
Model 2 g,i
1
0.88 (0.73 to 1.06)
1.00 (0.82 to 1.21)
.984
0.98 (0.90 to 1.06)
Model 3 g,j
1
0.89 (0.74 to 1.07)
1.00 (0.82 to 1.21)
.978
0.98 (0.90 to 1.06)
Midday
Unadjusted model
1
0.78 (0.65 to 0.93)
0.77 (0.64 to 0.93)
.006
0.95 (0.89 to 1.03)
Model 1 g,h
1
0.85 (0.71 to 1.02)
0.87 (0.72 to 1.05)
.149
0.99 (0.91 to 1.06)
Model 2 g,i
1
0.87 (0.72 to 1.05)
0.86 (0.70 to 1.05)
.125
0.97 (0.90 to 1.05)
Model 3 g,j
1
0.88 (0.73 to 1.06)
0.86 (0.70 to 1.05)
.138
0.98 (0.91 to 1.05)
Evening
Unadjusted model
1
1.01 (0.84 to 1.21)
1.12 (0.93 to 1.35)
.220
1.01 (0.95 to 1.08)
Model 1 g,h
1
0.99 (0.83 to 1.21)
1.09 (0.90 to 1.31)
.382
1.01 (0.94 to 1.08)
Model 2 g,i
1
1.08 (0.89 to 1.32)
1.17 (0.97 to 1.43)
.107
1.03 (0.97 to 1.11)
Model 3 g,j
1
1.09 (0.89 to 1.32)
1.17 (0.96 to 1.43)
.115
1.03 (0.97 to 1.11)
Variable
Consumption of ultra-processed foods (% of total energy)
Tertiles k
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First
Second
Third
←̅̅̅̅̅̅̅̅̅̅̅̅OR (95% CI)̅̅̅̅̅̅̅̅̅̅̅̅→
Period of the day
Morning
Unadjusted model
1
1.05 (0.89 to 1.25)
1.25 (1.06 to 1.49)
.009
1.16 (1.05 to 1.28)
Model 1 l,h
1
1.02 (0.86 to 1.22)
1.19 (0.86 to 1.22)
.049
1.12 (1.01 to 1.24)
Model 2 l,i
1
0.99 (0.83 to 1.19)
1.14 (0.94 to 1.37)
.173
1.09 (0.98 to 1.21)
Model 3 l,m
1
1.05 (0.87 to 1.25)
1.26 (1.04 to 1.53)
.021
1.17 (1.05 to 1.31)
Midday
Unadjusted model
1
0.85 (0.71 to 1.02)
1.06 (0.89 to 1.27)
.514
1.03 (0.96 to 1.12)
Model 1 l,h
1
0.88 (0.73 to 1.07)
1.19 (0.99 to 1.43)
.061
1.08 (1.00 to 1.17)
Model 2 l,i
1
0.90 (0.74 to 1.09)
1.18 (0.98 to 1.43)
.082
1.08 (0.99 to 1.17)
Model 3 l,n
1
0.91 (0.75 to 1.11)
1.25 (1.04 to 1.51)
.020
1.11 (1.03 to 1.20)
Evening
Unadjusted model
1
1.03 (0.86 to 1.23)
1.34 (1.12 to 1.60)
.001
1.10 (1.03 to 1.18)
Model 1 l,h
1
1.04 (0.87 to 1.25)
1.44 (1.19 to 1.73)
<.001
1.14 (1.06 to 1.23)
(continued on next page)
RESEARCH
10
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
■■ 2026 Volume 126 Number 3
In line with the hypotheses, we found that a higher pro-
portion of energy intake in the evening was associated with
overweight. However, no association was observed between
the percentage of energy intake assessed during the morning
and evening periods and obesity or abdominal obesity.
Conversely, other researchers reported an association with
BMI or overweight and obesity classified based on BMI. 10,49 In
a cohort study, during a 6-year follow-up period, participants
who consumed ≥48% of their total energy intake at dinner
were twice as likely to develop obesity compared with par-
ticipants who consumed <33% of their total energy intake at
dinner. 53 A previous study among Korean adults found that
morning eating was associated with a decreased prevalence
of abdominal obesity among women, but not in men. 50
By contrast, Longo Silva and colleagues 54 found no statis-
tically significant differences between early and late dinner
eaters concerning obesity in Brazilian adults. A systematic
review 16 aimed at investigating the association between
evening energy intake and weight in adults reported that only
4 of 10 studies found a significant association with obesity.
The authors noted significant heterogeneity among the
studies, and many trials had an unknown or high risk of bias.
16
We found that the percentage of energy intake from UPFs in
the 3 periods of the day was associated with obesity. In addi-
tion, there was an association between the percentage of en-
ergy intake from UPFs at midday and in the evening with
overweight, and the evening percentage was also associated
with abdominal obesity. To date, according to the best of our
knowledge, no published articles have analyzed this associa-
tion stratified by periods of the day, and therefore, the present
findings cannot be compared with previous studies. UPFs are
known to be high in energy, saturated fats, sugars, and free
sugars, and higher consumption of these foods is associated
with an increased risk of adiposity. 3,4,55,56 A previous study
using data from the UK NDNS found that higher consumption
of UPFs was associated with a 1.66 higher mean BMI, a 3.56-cm
larger mean WC, and a 90% higher odds of obesity compared
with lower consumption. 4
Possible explanations for the observed results are related
to satiety hormones such as leptin or ghrelin. Higher con-
sumption in the morning may reduce hunger, cravings for fat
and sweets, and also decrease postprandial ghrelin concen-
trations, which may prevent susceptibility to weight
gain. 57,58 Studies have shown that earlier intakes are asso-
ciated with higher thermogenesis, increased total energy
expenditure,
and
lower
glycemic
responses. 11,53,59
Conversely, later intake is linked to reduced insulin sensi-
tivity, lower efficiency of the thermic effect of food, and
decreased glucose tolerance. 60-62
The higher mean proportion of UPFs intake in the evening
(23.2%) compared with 20.9% and 10.4% at midday and in the
morning, respectively, may partially account for the observed
association. Furthermore, the magnitude of the association
between the percentage of energy intake from UPFs and
obesity was greater in the evening compared with other pe-
riods of the day. These results might be attributed to the po-
tential association between individuals with evening
chronotypes and a greater appetite for high-energy-dense
foods, including processed foods and UPFs. 63 Chronic sleep
deprivation is common among individuals with evening
chronotypes, as they tend to delay their routines to later hours,
while still meeting social and professional demands earlier in
the day. 64,65 Consequently, this sleep deprivation may lead to
increased ghrelin secretion and decreased leptin secretion,
Table 5. OR a (95% CI) of the association between tertiles and a 10% increase in energy intake and energy intake from ultra-
processed foods, across daytime periods, b and obesity status c in UK adults (aged 19-64 years), NDNS d 2008-2019 (n = 5749)
(continued)
Variable
Consumption of ultra-processed foods (% of total energy)
Tertiles k
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First
Second
Third
Model 2 l,i
1
1.07 (0.89 to 1.29)
1.40 (1.15 to 1.69)
.001
1.13 (1.05 to 1.22)
Model 3 l,o
1
1.07 (0.89 to 1.29)
1.45 (1.20 to 1.76)
<.001
1.16 (1.07 to 1.26)
a OR = odds ratio.
b Morning: 5:00 AM to 10:59 AM, midday: 11:00 AM to 4:59 PM, evening: 5:00 PM to 4:59 AM.
c Obesity: body mass index ≥30.
d NDNS = National Diet and Nutrition Survey.
e The tertiles were categorized based on cutoff points for energy consumption: Morning: first tertile: ≤13.1% of total energy intake; second tertile: 13.2%-20.6% of total energy intake; third
tertile: ≥20.7% of total energy intake. Midday: first tertile: ≤30.7% of total energy intake; second tertile: 30.8%-39.6% of total energy intake; third tertile: ≥39.7% of total energy intake.
Evening: first tertile: ≤42.8% of total energy intake; second tertile: 42.9%-52.1% of total energy intake; third tertile: ≥52.2% of total energy intake.
f P value for linear trend across tertiles of dietary contribution.
g Adjustment models for energy consumption.
h Model 1: adjusted for sex, age, region, survey year, and social class occupation.
i Model 2: Model 1 and physical activity, smoking, sleep duration, following a weight-loss diet, and frequency of alcohol consumption.
j Model 3: Model 2 and total energy intake.
k For ultra-processed foods intake: Morning: first tertile: ≤7.2% of total energy intake; second tertile: 7.3%-12.6% of total energy intake; third tertile: ≥12.7% of total energy intake. Midday:
first tertile: ≤16.3% of total energy intake; second tertile: 16.4%-24.5% of total energy intake; third tertile: ≥24.6% of total energy intake. Evening: first tertile: ≤17.1% of total energy
intake; second tertile: 17.2%-27.1% of total energy intake; third tertile: ≥27.2% of total energy intake.
l Adjustment models for ultra-processed foods consumption.
m Model 3: Model 2 and percentage of energy from ultra-processed foods in the midday and evening.
n Model 3: Model 2 and percentage of energy from ultra-processed foods in the morning and evening.
o Model 3: Model 2 and percentage of energy from ultra-processed foods in the morning and midday.
RESEARCH
■■ 2026 Volume 126 Number 3
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
11
Table 7. OR a (95% CI) of the association between tertiles and a 10% increase in total energy intake and energy intake from
ultra-processed foods, across daytime periods, b and abdominal obesity status c in UK adults (aged 19-64 years), NDNS d 2008-
2019 (n = 4244)
Variable
Energy Consumption (% of Total Energy) in Each Period of the Day
Tertiles e
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First
Second
Third
←̅̅̅̅̅̅̅̅̅̅̅̅OR (95% CI)̅̅̅̅̅̅̅̅̅̅̅̅→
Period of the day
Morning
Unadjusted model
1
1.00 (0.84 to 1.20)
1.17 (0.98 to 1.40)
.081
1.05 (0.97 to 1.14)
Model 1 g,h
1
0.88 (0.73 to 1.07)
1.00 (0.83 to 1.20)
.996
0.97 (0.89 to 1.06)
Model 2 g,i
1
0.87 (0.71 to 1.06)
0.95 (0.79 to 1.16)
.669
0.95 (0.87 to 1.04)
Model 3 g,j
1
0.88 (0.72 to 1.07)
0.96 (0.79 to 1.16)
.676
0.95 (0.87 to 1.04)
Midday
Unadjusted model
1
0.78 (0.65 to 0.93)
0.83 (0.64 to 0.93)
.048
0.95 (0.89 to 1.03)
Model 1 g,h
1
0.87 (0.73 to 1.05)
0.96 (0.79 to 1.16)
.228
0.99 (0.92 to 1.07)
Model 2 g,i
1
0.88 (0.73 to 1.07)
0.95 (0.78 to 1.16)
.592
0.98 (0.91 to 1.06)
Model 3 g,j
1
0.89 (0.73 to 1.08)
0.95 (0.78 to 1.16)
.613
0.98 (0.91 to 1.06)
Evening
Unadjusted model
0.98 (0.82 to 1.19)
1.11 (0.92 to 1.34)
.275
1.01 (0.95 to 1.08)
Model 1 g,h
1
0.99 (0.82 to 1.19)
1.12 (0.93 to 1.36)
.228
1.03 (0.96 to 1.10)
Model 2 g,i
1
1.03 (0.85 to 1.25)
1.16 (0.95 to 1.42)
.147
1.04 (0.97 to 1.12)
Model 3 g,j
1
1.04 (0.86 to 1.26)
1.16 (0.95 to 1.42)
.153
1.04 (0.97 to 1.12)
Variable
Consumption of Ultra-Processed Foods (% of Total Energy)
Tertiles k
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First
Second
Third
←̅̅̅̅̅̅̅̅̅̅̅̅OR (95% CI)̅̅̅̅̅̅̅̅̅̅̅̅→
Period of the day
Morning
Unadjusted model
1
0.97 (0.82 to 1.15)
1.14 (0.96 to 1.35)
.137
1.07 (0.96 to 1.19)
Model 1 l,h
1
0.92 (0.77 to 1.11)
1.07 (0.89 to 1.28)
.474
1.02 (0.91 to 1.14)
Model 2 l,i
1
0.92 (0.76 to 1.11)
1.05 (0.87 to 1.27)
.602
1.01 (0.90 to 1.14)
Model 3 l,m
1
0.94 (0.78 to 1.14)
1.12 (0.92 to 1.36)
.255
1.06 (0.94 to 1.20)
Midday
Unadjusted model
1
0.73 (0.61 to 0.87)
0.89 (0.74 to 1.07)
.200
0.97 (0.89 to 1.05)
Model 1 l,h
1
0.78 (0.64 to 0.94)
1.09 (0.90 to 1.33)
.423
1.05 (0.96 to 1.14)
Model 2 l,i
1
0.77 (0.64 to 0.94)
1.09 (0.89 to 1.34)
.449
1.05 (0.96 to 1.15)
Model 3 l,n
1
0.78 (0.64 to 0.95)
1.13 (0.92 to 1.38)
.283
1.07 (0.98; 1.17)
Evening
Unadjusted model
1
0.96 (0.80 to 1.14)
1.16 (0.96 to 1.40)
.126
1.05 (0.98 to 1.13)
Model 1 l,h
1
0.99 (0.83 to 1.19)
1.38 (1.14 to 1.68)
.001
1.13 (1.05 to 1.23)
(continued on next page)
RESEARCH
12
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
■■ 2026 Volume 126 Number 3
resulting in higher energy intake. 66 A study of Italian adults
found that individuals classified as late eaters reported higher
consumption of UPFs, suggesting that the degree of food pro-
cessing could be one of the factors explaining the relationship
between meal timing and impaired cardiometabolic out-
comes. 67 Food processing affects both the nutritional compo-
sition and the architecture of food. UPFs are often unstructured
(their natural physical form is heavily altered during
manufacturing, resulting in uniform, refined products), frac-
tionated, and frequently enriched with free glucose and su-
crose, thereby increasing the availability of glucose for
absorption. 68,69 However, during the nighttime, there is
reduced glucose tolerance, which could explain the increased
likelihood of obesity indicators during this period of the day. 70
A study limitation included the utilization of cross-sectional
data for the analyses, which precludes causal inference. The
possibility of under-reporting food consumption cannot be
disregarded, as it is a common issue across all dietary
assessment methods and tends to increase with higher BMI. 71
In addition, self-reported dietary intake may be influenced by
social desirability or the desire for social approval, potentially
impacting risk estimates in epidemiologic studies. 72 Although
4 consecutive 24-hour food diaries were used, this approach
has important limitations for estimating usual dietary intake.
As highlighted by Kirkpatrick and colleagues, 73 characterizing
usual intake is more complex when short-term assessments
are collected over consecutive days, given that intake on
consecutive days tends to be more correlated, potentially
reducing day-to-day variability. Furthermore, despite inter-
viewer review of food diaries, some information on portion
sizes and food details remained missing and required impu-
tation. Although standardized procedures were applied to
minimize bias, these imputations may have introduced a
degree of measurement error.
The influence of chronotype and shift work data, which
could affect the timing of food consumption, 50,74 could not
be examined due to the lack of information. It is plausible
that categorizing food consumption into morning, midday,
and evening may not accurately reflect long-term dietary
habits subject to seasonal variations. The NDNS collects
limited information on food processing, potentially leading
to misclassification of food items. 3,34 Although the percent-
age of energy from UPFs does not fully capture artificially
sweetened beverages, it was used as it better reflects the
energy density of UPF. Other studies assessing the associa-
tion between the consumption of UPFs and obesity also use
the percentage energy from UPF. 4,5 Artificially sweetened
beverages accounted for 3.9% of total food grams (data not
shown), but non-nutritive sweeteners in other items may
not be fully captured. Future studies should consider
assessing UPFs in grams and their daily distribution in
relation to obesity risk. A broader measure of overall dietary
quality, such as the Healthy Eating Index, was not accounted
for in the present study. Future research should explore
whether total UPFs consumption and the timing of intake are
linked to obesity risk independently of overall diet quality. 75
The study boasts several strengths, such as the use of a large,
nationally representative sample of UK adults. Extensive ad-
justments for sociodemographic and behavioral factors, total
energy intake, and energy from UPFs were included. The
findings contribute to enhancing the understanding of the
mechanisms underlying the relationship between meal timing
and obesity indicators, which supports the need for further
investigation into the timing of UPFs consumption and its
Table 7. OR a (95% CI) of the association between tertiles and a 10% increase in total energy intake and energy intake from
ultra-processed foods, across daytime periods, b and abdominal obesity status c in UK adults (aged 19-64 years), NDNS d 2008-
2019 (n = 4244) (continued)
Variable
Consumption of Ultra-Processed Foods (% of Total Energy)
Tertiles k
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First
Second
Third
Model 2 l,i
1
0.99 (0.83 to 1.20)
1.29 (1.05 to 1.58)
.016
1.11 (1.02 to 1.20)
Model 3 l,o
1
0.99 (0.83 to 1.20)
1.31 (1.07 to 1.61)
.010
1.12 (1.03 to 1.22)
a OR = odds ratio.
b Morning: 5:00 AM to 10:59 AM, midday: 11:00 AM to 4:59 PM, evening: 5:00 PM to 4:59 AM.
c Abdominal obesity: waist circumference ≥102 cm for men and ≥ 88 cm for women.
d NDNS = National Diet and Nutrition Survey.
e The tertiles were categorized based on cutoff points for energy consumption: Morning: first tertile: ≤13.1% of total energy intake; second tertile: 13.2%-20.6% of total energy intake; third
tertile: ≥20.7% of total energy intake. Midday: first tertile: ≤30.7% of total energy intake; second tertile: 30.8%-39.6% of total energy intake; third tertile: ≥39.7% of total energy intake.
Evening: first tertile: ≤42.8% of total energy intake; second tertile: 42.9%-52.1% of total energy intake; third tertile: ≥52.2% of total energy intake.
f P value for linear trend across tertiles of dietary contribution.
g Adjustment model for energy consumption.
h Model 1: adjusted for sex, age, region, survey year, and social class occupation.
i Model 2: Model 1 and physical activity, smoking, sleep duration, following a weight-loss diet, and frequency of alcohol consumption.
j Model 3: Model 2 and total energy intake.
k For ultra-processed foods intake: Morning: first tertile: ≤7.2% of total energy intake; second tertile: 7.3%-12.6% of total energy intake; third tertile: ≥12.7% of total energy intake. Midday:
first tertile: ≤16.3% of total energy intake; second tertile: 16.4%-24.5% of total energy intake; third tertile: ≥24.6% of total energy intake. Evening: first tertile: ≤17.1% of total energy
intake; second tertile: 17.2%-27.1% of total energy intake; third tertile: ≥27.2% of total energy intake.
l Adjustment model for ultra-processed foods consumption.
m Model 3: Model 2 and percentage of energy from ultra-processed foods in the midday and evening.
n Model 3: Model 2 and percentage of energy from ultra-processed foods in the morning and evening.
o Model 3: Model 2 and percentage of energy from ultra-processed foods in the morning and midday.
RESEARCH
■■ 2026 Volume 126 Number 3
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
13
potential association with health outcomes. This represents a
critical and clearly defined primary objective for public health,
as it addresses both the prevailing epidemic of our era (obesity)
and the primary driver of health care expenditures, affecting
individuals at increasingly younger ages. 76 These may include
the adoption of fiscal measures, regulation of advertising for
these foods, implementation of front-of-package warning la-
bels, and control of food environments related to access to such
products. 77
CONCLUSIONS
The findings indicate an association between a higher per-
centage of energy intake in the morning and a lower odds of
overweight, and consuming more of the day’s total energy
intake in the evening is linked to a higher likelihood of
overweight. The strength of the association between UPFs
consumption and indicators of obesity was notably higher in
the evening compared with other times of the day. Future
longitudinal studies, incorporating individuals’ chronotypes,
should be conducted to further assess causality.
References
1. Lobstein T, Jackson-Leach R, Powis J, Brinsden H, Gray M. World
Obesity Atlas 2023: Global, Regional and National Estimates for the
Prevalence of Overweight and Obesity up to 2035 for Adults, Children
and Adolescents. World Obesity Federation. 2023. Accessed February
2024. https://www.worldobesityday.org/assets/downloads/World_
Obesity_Atlas_2023_Report.pdf
2. Shekar M, Popkin B. Obesity Health and Economic Consequences of an
Impending Global Challenge. World Bank; 2020. https://doi.org/10.
1596/978-1-4648-1491-4
3. Rauber F, da Costa Louzada ML, Steele EM, Millett C, Monteiro CA,
Levy RB. Ultra-processed food consumption and chronic non-
communicable diseases-related dietary nutrient profile in the UK
(2008—2014). Nutrients. 2018;10(5):587. https://doi.org/10.3390/
nu10050587
4. Rauber F, Steele EM, da Costa Louzada ML, Millett C, Monteiro CA,
Levy RB. Ultra-processed food consumption and indicators of obesity
in the United Kingdom population (2008-2016). PLoS One.
2020;15(5):e0232676. https://doi.org/10.1371/journal.pone.0232676
5. Liu J, Steele EM, Li Y, Yi SS, Monteiro CA, Mozaffarian D. Con-
sumption of ultraprocessed foods and body fat distribution among
U.S. adults. Am J Prev Med. 2023;65(3):427-438. https://doi.org/10.
1016/j.amepre.2023.03.012
6. Julia C, Martinez L, Allès B, et al. Contribution of ultra-processed
foods in the diet of adults from the French NutriNet-Santé study.
Public Health Nutr. 2018;21(1):27-37. https://doi.org/10.1017/
S1368980017001367
7. Moradi S, Entezari MH, Mohammadi H, et al. Ultra-processed food
consumption and adult obesity risk: a systematic review and dose-
response meta-analysis. Crit Rev Food Sci Nutr. 2023;63(2):249-260.
https://doi.org/10.1080/10408398.2021.1946005
8. Lane MM, Gamage E, Du S, et al. Ultra-processed food exposure and
adverse health outcomes: umbrella review of epidemiological
meta-analyses. BMJ. 2024;384:e077310. https://doi.org/10.1136/
bmj-2023-077310
9. Davis R, Rogers M, Coates AM, Leung GKW, Bonham MP. The impact
of meal timing on risk of weight gain and development of obesity: a
review of the current evidence and opportunities for dietary inter-
vention. Curr Diab Rep. 2022;22(4):147-155. https://doi.org/10.
1007/s11892-022-01457-0
10. Xiao Q, Garaulet M, Scheer FAJL. Meal timing and obesity: interactions
with macronutrient intake and chronotype. Int J Obes. 2019;43(9):
1701-1711. https://doi.org/10.1038/s41366-018-0284-x
11. Young IE, Poobalan A, Steinbeck K, O’Connor HT, Parker HM. Dis-
tribution of energy intake across the day and weight loss: a sys-
tematic review and meta-analysis. Obes Rev. 2023;24(3):e13537.
https://doi.org/10.1111/obr.13537
12. Waterhouse J, Minors D, Atkinson G, Benton D. Chronobiology and
meal times: internal and external factors. Br J Nutr. 1997;77(suppl
1):S29-S38. https://doi.org/10.1079/BJN19970102
13. Hutchison AT, Heilbronn LK. Metabolic impacts of altering meal
frequency and timing—does when we eat matter? Biochimie.
2016;124:187-197. https://doi.org/10.1016/j.biochi.2015.07.025
14. McCrory MA, Shaw AC, Lee JA. Energy and nutrient timing for
weight control: does timing of ingestion matter? Endocrinol Metab
Clin North Am. 2016;45(3):689-718. https://doi.org/10.1016/j.ecl.
2016.04.017
15. Wang JB, Patterson RE, Ang A, Emond JA, Shetty N, Arab L. Timing of
energy intake during the day is associated with the risk of obesity in
adults. J Hum Nutr Diet. 2014;27(suppl 2):255-262. https://doi.org/
10.1111/jhn.12141
16. Fong M, Caterson ID, Madigan CD. Are large dinners associated
with excess weight, and does eating a smaller dinner
achieve greater weight loss? A systematic review and meta-anal-
ysis. Br J Nutr. 2017;118(8):616-628. https://doi.org/10.1017/
S0007114517002550
17. Leech RM, Worsley A, Timperio A, McNaughton SA. Characterizing
eating patterns: a comparison of eating occasion definitions. Am J
Clin Nutr. 2015;102(5):1229-1237. https://doi.org/10.3945/ajcn.115.
114660
18. Public Health England. Public Health England. National Diet and
Nutrition Survey Rolling Programme (NDNS RP) Years 1-4 User Guide.
Public Health England; 2014.
19. Public Health England. Public Health England. National Diet and
Nutrition Survey Rolling Programme (NDNS RP) Years 5-6 (2012/13-
2013/14): User Guide. Public Health England; 2016.
20. Public Health England. Public Health England. National Diet and
Nutrition Survey Rolling Programme (NDNS RP) Years 9-11 (2016/17-
2018/19): User Guide. Public Health England; 2018.
21. Public Health England. Public Health England. National Diet and
Nutrition Survey Rolling Programme (NDNS RP) Years 7-8 (2014/15-
2015/16): User Guide. Public Health England; 2020.
22. Stowe TA, McClung CA. How does chronobiology contribute to the
development of diseases in later life. Clin Interv Aging. 2023;18:655-
666. https://doi.org/10.2147/CIA.S380436
23. Nielsen SJ, Adair L. An alternative to dietary data exclusions. J Am Diet
Assoc. 2007;107(5):792-799. https://doi.org/10.1016/j.jada.2007.02.
003
24. Public Health England. National Diet and Nutrition Survey Years 1 to 9 of
the Rolling Programme: Appendix X — Misreporting in the NDNS RP:
Summary of Results and Their Interpretation. Publich Health England;
2008.
25. Fitt E, Cole D, Ziauddeen N, et al. DINO (Diet In Nutrients Out)—an
integrated dietary assessment system. Public Health Nutr.
2015;18(2):234-241. https://doi.org/10.1017/S1368980014000342
26. Food Standards Agency, Public Health England. NDNS: results from
Years 5 and 6 (combined). Results of the National Diet and Nutrition
Survey (NDNS) rolling programme for 2012 to 2013 and 2013 to
2014. Gov.UK. Accessed November 1, 2024. https://www.gov.uk/
government/statistics/ndns-results-from-years-5-and-6-combined
27. Food Standards Agency, Public Health England. NDNS: results from
years 7 and 8 (combined). Results of the National Diet and Nutrition
Survey (NDNS) rolling programme for 2014 to 2015 and 2015 to
2016. Gov.UK. Accessed February 12, 2020. https://www.gov.uk/
government/statistics/ndns-results-from-years-7-and-8-combined
28. Food Standards Agency, Public Health England. NDNS: results from
Years 1 to 4 (combined). Results of the National Diet and Nutrition
Survey (NDNS) rolling programme for 2008 and 2009 to 2011 and
2012. Gov.UK. Accessed November 1, 2024. https://www.gov.uk/
government/statistics/national-diet-and-nutrition-survey-results-
from-years-1-to-4-combined-of-the-rolling-programme-for-2008-
and-2009-to-2011-and-2012
29. Food Standards Agency, Public Health England. NDNS: results from
years 9 to 11 (combined) — statistical summary. Gov.UK. Accessed
November 1, 2024. https://www.gov.uk/government/statistics/
ndns-results-from-years-9-to-11-2016-to-2017-and-2018-to-2019/
ndns-results-from-years-9-to-11-combined-statistical-summary
30. Roe M, Pinchen H, Church S, Finglas P. McCance and Widdowson’s The
Composition of Foods Seventh Summary Edition and updated
Composition of Foods Integrated Dataset. Nutr Bull. 2015;40(1):36-
39. https://doi.org/10.1111/nbu.12124
RESEARCH
14
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
■■ 2026 Volume 126 Number 3
31. Wrieden WL, Barton KL, Cochrane L, Adamson AJ. Final Technical
Report to the Food Standards Agency. Calculation and Collation of
Typical Food Portion Sizes for Adults Aged 19—64 and Older People
Aged 65 and Over. Food Standards Agency; 2006.
32. Monteiro CA, Levy RB, Claro RM, de Castro IRR, Cannon G. A new
classification of foods based on the extent and purpose of their
processing. Cad Saude Publica. 2010;26(11):2039-2049. https://doi.
org/10.1590/S0102-311X2010001100005
33. Monteiro CA, Cannon G, Levy RB, et al. Ultra-processed foods: what
they are and how to identify them. Public Health Nutr. 2019;22(5):
936-941. https://doi.org/10.1017/S1368980018003762
34. Rauber F, Louzada MLDC, Martinez Steele E, et al. Ultra-processed
foods and excessive free sugar intake in the UK: a nationally
representative cross-sectional study. BMJ Open. 2019;9(10):
e027546. https://doi.org/10.1136/bmjopen-2018-027546
35. Almoosawi S, Vingeliene S, Karagounis LG, Pot GK. Chrono-nutri-
tion: a review of current evidence from observational studies on
global trends in time-of-day of energy intake and its association
with obesity In: Proceedings of the Nutrition Society. Vol 75. Cam-
bridge University Press; 2016:487-500. https://doi.org/10.1017/
S0029665116000306
36. Huseinovic E, Winkvist A, Freisling H, et al. Timing of eating across
ten European countries - results from the European Prospective
Investigation into Cancer and Nutrition (EPIC) calibration study.
Public Health Nutr. 2019;22(2):324-335. https://doi.org/10.1017/
S1368980018002288
37. Truman SC, Wirth MD, Arp Adams S, Turner-McGrievy GM, Reiss KE,
Hébert JR. Meal timing, distribution of macronutrients, and
inflammation among African-American women: a cross-sectional
study. Chronobiol Int. 2022;39(7):976-983. https://doi.org/10.1080/
07420528.2022.2053702
38. Kahleova H, Lloren JI, Mashchak A, Hill M, Fraser GE. Meal frequency
and timing are associatedwith changes in body mass index in
Adventist Health Study 2. J Nutr. 2017;147(9):1722-1728. https://
doi.org/10.3945/jn.116.244749
39. Lohman T, Roche A, Martorell R. Anthropometric Standardization
Reference Manual. Human Kinetics; 1998.
40. World Health Organization. WHO STEPwise Approach to Surveillance
(STEPS). World Health Organization; 2008.
41. World Health Organization. Body mass index (BMI) classifications.
Accessed February 27, 2024. https://www.who.int/europe/news-
room/fact-sheets/item/a-healthy-lifestyle—who-recommendations
42. World Health Organization. Waist Circumference and Waist—Hip
Ratio: report of a WHO Expert Consultation. World Health Organiza-
tion; 2011.
43. The National Statistics Socio-Economic Classification (NS-SEC). Of-
fice for National Statistics. Accessed November 1, 2024, https://
www.ons.gov.uk/methodology/classificationsandstandards/other
classifications/thenationalstatisticssocioeconomicclassification
nssecrebasedonsoc2010
44. Golubic R, May AM, Benjaminsen Borch K, et al. Validity of elec-
tronically administered Recent Physical Activity Questionnaire
(RPAQ) in Ten European Countries. PLoS One. 2014;9(3):e92829.
https://doi.org/10.1371/journal.pone.0092829
45. Mindell J. Appendix V: Measuring Physical Activity in Adults Using
the Recent Physical Activity Questionnaire (RPAQ). National Diet
and Nutrition Survey. Results from Years 1, 2, 3 and 4 (Combined) of
the Rolling Programme (2008/09 — 2011/12). Public Health En-
gland; 2014. Accessed November 1, 2024. https://www.gov.wales/
sites/default/files/statistics-and-research/2019-05/151209-national-
diet-nutrition-survey-rolling-programme-years-2-5-app-v-en.pdf
46. Nakade M, Takeuchi H, Kurotani M, Harada T. Effects of meal habits
and alcohol/cigarette consumption on morningness-eveningness
preference and sleep habits by Japanese female students aged
18—29. J Physiol Anthropol. 2009;28(2):83-90. https://doi.org/10.
2114/jpa2.28.83
47. White IR, Royston P, Wood AM. Multiple imputation using chained
equations: issues and guidance for practice. Stat Med. 2011;30(4):
377-399. https://doi.org/10.1002/sim.4067
48. Senaviratna NAMR, Cooray TMJA. Diagnosing multicollinearity of
logistic regression model. Asian J Prob Stat. 2019;5(2):1-9. https://
doi.org/10.9734/ajpas/2019/v5i230132
49. Aljuraiban GS, Chan Q, Oude Griep LM, et al. The impact of eating
frequency and time of intake on nutrient quality and body mass
index: the INTERMAP study, a population-based study. J Acad Nutr
Diet. 2015;115(4):528-536.e1. https://doi.org/10.1016/j.jand.2014.
11.017
50. Ha K, Song Y. Associations of meal timing and frequency with
obesity and metabolic syndrome among Korean adults. Nutrients.
2019;11(10):2437. https://doi.org/10.3390/nu11102437
51. Stata [computer program]. Release 16. StataCorp LLC; 2019.
52. Jakubowicz D, Barnea M, Wainstein J, Froy O. High caloric intake at
breakfast vs. dinner differentially influences weight loss of over-
weight and obese women. Obesity. 2013;21(12):2504-2512. https://
doi.org/10.1002/oby.20460
53. Bo S, Musso G, Beccuti G, et al. Consuming more of daily caloric
intake at dinner predisposes to obesity. A 6-year population-based
prospective cohort study. PLoS One. 2014;9(9):e108467. https://
doi.org/10.1371/journal.pone.0108467
54. Longo-Silva G, Bezerra de Oliveira PM, Pedrosa AKP, et al. Breakfast
skipping and timing of lunch and dinner: relationship with BMI and
obesity. Obes Res Clin Pract. 2022;16(6):507-513. https://doi.org/10.
1016/j.orcp.2022.10.012
55. Askari M, Heshmati J, Shahinfar H, Tripathi N, Daneshzad E. Ultra-
processed food and the risk of overweight and obesity: a systematic
review and meta-analysis of observational studies. Int J Obes.
2020;44(10):2080-2091.
https://doi.org/10.1038/s41366-020-
00650-z
56. Mambrini SP, Menichetti F, Ravella S, et al. Ultra-processed
food consumption and incidence of obesity and cardiometabolic
risk factors in adults: a systematic review of prospective
studies. Nutrients. 2023;15(11):2583. https://doi.org/10.3390/
nu15112583
57. Solomon TPJ, Chambers ES, Jeukendrup AE, Toogood AA, Blannin AK.
The effect of feeding frequency on insulin and ghrelin responses in
human subjects. Br J Nutr. 2008;100(4):810-819. https://doi.org/10.
1017/S000711450896757X
58. Basolo A, Bechi Genzano S, Piaggi P, Krakoff J, Santini F. Energy
balance and control of body weight: possible effects of meal timing
and circadian rhythm dysregulation. Nutrients. 2021;13(9):3276.
https://doi.org/10.3390/nu13093276
59. Beccuti G, Monagheddu C, Evangelista A, et al. Timing of food
intake: sounding the alarm about metabolic impairments? A sys-
tematic review. Pharmacol Res. 2017;125(Pt B):132-141. https://doi.
org/10.1016/j.phrs.2017.09.005
60. Lee A, Ader M, Bray GA, Bergman RN. Diurnal variation in glucose
tolerance: cyclic suppression of insulin action and insulin secretion
in normal-weight, but not obese, subjects. Diabetes. 1992;41(6):
750-759. https://doi.org/10.2337/diab.41.6.750
61. Morgan LM, Aspostolakou’ F, Wright J, Gama R. Diurnal variations in
peripheral insulin resistance and plasma non-esterified fatty acid
concentrations: a possible link? Ann Clin Biochem. 1999;36(4):447-
450. https://doi.org/10.1177/000456329903600407
62. Shaw E, Leung GKW, Jong J, et al. The impact of time of day on
energy expenditure: implications for long-term energy balance.
Nutrients. 2019;11(10):2383. https://doi.org/10.3390/nu11102383
63. Teixeira GP, Guimarães KC, Soares AGNS, et al. Role of chronotype in
dietary intake, meal timing, and obesity: a systematic review. Nutr
Rev. 2023;81(1):75-90. https://doi.org/10.1093/nutrit/nuac044
64. Merikanto I, Kronholm E, Peltonen M, Laatikainen T, Lahti T,
Partonen T. Relation of chronotype to sleep complaints in the gen-
eral Finnish population. Chronobiol Int. 2012;29(3):311-317. https://
doi.org/10.3109/07420528.2012.655870
65. Duffy J, Roepke. Differential impact of chronotype on weekday and
weekend sleep timing and duration. Nat Sci Sleep. 2010;2:213-220.
https://doi.org/10.2147/NSS.S12572
66. Taheri S, Lin L, Austin D, Young T, Mignot E. Short sleep duration is
associated with reduced leptin, elevated ghrelin, and increased
body mass index. PLoS Med. 2004;1(3):e62. https://doi.org/10.1371/
journal.pmed.0010062
67. Bonaccio M, Ruggiero E, Di Castelnuovo A, et al. Association be-
tween late-eating pattern and higher consumption of ultra-pro-
cessed food among Italian adults: findings from the INHES study.
Nutrients. 2023;15(6):1497. https://doi.org/10.3390/nu15061497
68. Fardet A. Minimally processed foods are more satiating and less
hyperglycemic than ultra-processed foods: a preliminary study with
98 ready-to-eat foods. Food Funct. 2016;7(5):2338-2346. https://doi.
org/10.1039/C6FO00107F
RESEARCH
■■ 2026 Volume 126 Number 3
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
15
69. Fardet A. Characterization of the degree of food processing in rela-
tion with its health potential and effects. Adv Food Nutr Res.
2018;85:79-129. https://doi.org/10.1016/bs.afnr.2018.02.002
70. Bandín C, Scheer FAJL, Luque AJ, et al. Meal timing affects glucose
tolerance, substrate oxidation and circadian-related variables: a
randomized, crossover trial. Int J Obes. 2015;39(5):828-833. https://
doi.org/10.1038/ijo.2014.182
71. Ravelli MN, Schoeller DA. Traditional self-reported dietary in-
struments are prone to inaccuracies and new approaches are needed.
Front Nutr. 2020;7:90. https://doi.org/10.3389/fnut.2020.00090
72. Hebert Jr, Clemow L, Pbert L, Ockene IS, Ockene JK. Social desir-
ability bias in dietary self-report may compromise the validity of
dietary intake measures. Int J Epidemiol. 1995;24(2):389-398.
https://doi.org/10.1093/ije/24.2.389
73. Kirkpatrick SI, Guenther PM, Subar AF, et al. Using short-term
dietary intake data to address research questions related to usual
dietary intake among populations and subpopulations: assump-
tions, statistical techniques, and considerations. J Acad Nutr Diet.
2022;122(7):1246-1262.
https://doi.org/10.1016/j.jand.2022.03.
010
74. Wang C, Almoosawi S, Palla L. Day-time patterns of carbohydrate
intake in adults by non-parametric multi-level latent class anal-
ysis—results from the UK national diet and nutrition survey (2008/
09—2015/16). Nutrients. 2019;11(10). https://doi.org/10.3390/
nu11102476
75. O’Connor LE, Higgins KA, Smiljanec K, et al. Perspective: A
research roadmap about ultra-processed foods and human
health for the United States food system: proceedings from an
Interdisciplinary, Multi-Stakeholder Workshop. In: Advances in
Nutrition. Vol 14. Elsevier; 2023:1255-1269. https://doi.org/10.
1016/j.advnut.2023.09.005
76. Wang YC, McPherson K, Marsh T, Gortmaker SL, Brown M. Health
and economic burden of the projected obesity trends in the USA and
the UK. Lancet. 2011;378(9793):815-825. https://doi.org/10.1016/
S0140-6736(11)60814-3
77. Popkin BM, Barquera S, Corvalan C, et al. Towards unified and
impactful policies to reduce ultra-processed food consumption
and promote healthier eating. Lancet Diabetes Endocrinol.
2021;9(7):462-470.
https://doi.org/10.1016/S2213-8587(21)00
078-4
AUTHOR INFORMATION
P. de Fragas Hinnig is a professor, Post-Graduation Program in Nutrition, Health Sciences Center, Federal University of Santa Catarina, Flo-
rianópolis, Brazil. F. Rauber is a researcher, Department of Preventive Medicine, School of Medicine, University of São Paulo, São Paulo, Brazil,
and a Collaborating Researcher, Center for Epidemiological Research in Nutrition and Health, University of São Paulo, São Paulo, Brazil. K. Chang
is a researcher, Public Health Policy Evaluation Unit, School of Public Health, Imperial College London, London, UK. E. P. Vamos is a senior
clinical lecturer, Public Health Policy Evaluation Unit, School of Public Health, Imperial College London, London, UK. R. Bertazzi Levy is a Senior
Researcher, Department of Preventive Medicine, School of Medicine, University of São Paulo, São Paulo, Brazil, a Collaborating Researcher,
Center for Epidemiological Research in Nutrition and Health, University of São Paulo, São Paulo, Brazil, and a Visiting Professor, Institute of
Biomedical Research of Salamanca, University of Salamanca, Salamanca, Spain.
Address correspondence to: Patrícia de Fragas Hinnig, PhD, Post-Graduation Program in Nutrition, Health Sciences Center, Federal University of
Santa Catarina, Florianópolis 88040-900, Brazil. E-mail: phinnig@yahoo.com.br
STATEMENT OF POTENTIAL CONFLICT OF INTEREST
No potential conflict of interest was reported by the authors.
FUNDING/SUPPORT
There is no funding to disclose.
AUTHOR CONTRIBUTIONS
P. de Fragas Hinnig, F. Rauber, and R. Bertazzi Levy designed the research and methodology. P. de Fragas Hinnig performed data analysis and
interpretation and wrote the original draft. F. Rauber and R. Bertazzi Levy assisted in data analysis and interpretation and revised and edited the
manuscript. K. Chang and E. P. Vamos revised and edited the manuscript. All authors read and approved the final manuscript.
DECLARATION OF GENERATIVE AI AND AI-ASSISTED TECHNOLOGIES IN THE WRITING PROCESS
During the preparation of this work, the authors used ChatGPT in order to help make the paragraphs more concise, reduce the number of
words, and refine the English grammar and style to improve overall readability. After using this tool, the authors reviewed and edited the
content as needed. The authors take full responsibility for the content of the publication.
RESEARCH
16
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
■■ 2026 Volume 126 Number 3
Table 2. Mean (SE) proportion of energy intake and UPF a (% of total energy) consumed in the morning, b midday, c and
evening d according to sociodemographic and behavioral characteristics of the UK population (aged 19 to 64 years), NDNS e
2008-2019 (n = 5749)
Characteristic
% of Total Energy Intake
% of Total Energy Intake From UPF
Morning
Midday
Evening
Morning
Midday
Evening
←̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅mean (SE)̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅→
Sex f
Male
17.0 (0.2)
34.3 (0.3)
48.7 (0.3)
10.6 (0.2)
21.0 (0.3)
24.2 (0.3)
Female
17.3 (0.2)
36.1 (0.2)***
46.6 (0.3)***
10.2 (0.1)
20.8 (0.2)
22.2 (0.3)***
Age group f
19-34 y
15.6 (0.3)
36.8 (0.4)
47.6 (0.4)
9.7 (0.2)
23.0 (0.3)
25.5 (0.4)
35-49 y
17.7 (0.2)
35.0 (0.3)
47.3 (0.4)
10.8 (0.2)
20.6 (0.2)
22.5 (0.3)
50-64 y
18.3 (0.3)***
33.6 (0.3)***
48.1 (0.3)
10.8 (0.2)***
18.8 (0.2)***
21.2 (0.3)***
Ethnicity f
White
17.1 (0.2)
35.0 (0.2)
47.9 (0.2)
10.5 (0.1)
21.4 (0.2)
23.8 (0.2)
Black
16.5 (1.0)
35.0 (1.4)
48.5 (1.6)
9.5 (0.7)
18.0 (1.0)***
18.7 (1.3)***
Asian
18.1 (0.7)
37.1 (0.9)*
44.7 (1.1)**
10.1 (0.5)
18.1 (0.8)***
19.8 (0.8)***
Mixed g
18.7 (1.1)
37.2 (1.7)
44.1 (1.9)
10.3 (0.8)
22.1 (1.7)***
20.9 (1.8)***
Any other group h
17.9 (1.2)
37.4 (1.5)
44.6 (1.8)
8.6 (0.8)*
15.2 (1.2)***
16.7 (1.4)***
Region f
England North
16.8 (0.3)
35.1 (0.4)
48.1 (0.4)
10.3 (0.3)
21.4 (0.4)
24.0 (0.4)
England Central/Midlands
16.9 (0.4)
35.3 (0.5)
47.8 (0.5)
10.4 (0.3)
21.9 (0.4)
24.5 (0.6)
England South (including London)
17.4 (0.2)
35.5 (0.3)
47.1 (0.4)
10.4 (0.2)
20.1 (0.3)**
21.8 (0.3)****
Scotland
17.2 (0.4)
33.5 (0.6)***
49.3 (0.7)
10.2 (0.3)
21.2 (0.5)
25.1 (0.7)
Wales
17.2 (0.5)
36.4 (0.5)***
46.4 (0.6)***
10.4 (0.3)
22.0 (0.5)
22.6 (0.7)
Northern Ireland
17.0 (0.4)
34.1 (0.4)
48.8 (0.5)
10.7 (0.3)
21.7 (0.4)
25.8 (0.4)**
Social class occupation ij
Higher managerial and professional
occupations
17.3 (0.3)
35.4 (0.4)
47.3 (0.4)
10.2 (0.2)**
20.0 (0.3)***
20.9 (0.4)***
Routine and manual occupations
17.3 (0.3)
35.6 (0.3)
47.1 (0.4)
10.7 (0.2)
21.6 (0.3)
25.1 (0.4)
Intermediate occupations
17.1 (0.3)
34.9 (0.4)
48.0 (0.5)
10.5 (0.3)
21.0 (0.4)
22.5 (0.4)
Lower managerial and professional
occupations
17.0 (0.3)
34.8 (0.4)
48.3 (0.5)
10.1 (0.2)
20.6 (0.3)
22.8 (0.5)
Physical activity ik
First quartile
16.6 (0.3)
36.2 (0.4)
47.2 (0.5)
10.1 (0.2)
21.4 (0.4)
23.7 (0.4)
Second quartile
17.3 (0.3)
35.5 (0.4)
47.2 (0.4)
10.3 (0.2)
21.0 (0.3)
22.6 (0.4)
Third quartile
16.3 (0.3)
35.1 (0.4)
48.5 (0.5)
9.8 (0.2)
20.5 (0.3)
23.0 (0.5)
Fourth quartile
18.4 (0.4)**
33.9 (0.4)***
47.6 (0.4)
11.5 (0.3)**
20.8 (0.4)
23.6 (0.4)
Smoking status f
Nonsmoker
17.7 (0.2)
35.3 (0.2)
47.0 (0.2)
10.8 (0.1)
20.5 (0.2)
22.1 (0.2)
Smoker
15.1 (0.3)***
35.0 (0.4)
49.9 (0.5)***
9.1 (0.2)***
22.3 (0.4)***
26.7 (0.5)***
Frequency of alcohol consumption i
≥ 3 times/wk
15.9 (0.3)
33.2 (0.4)
50.8 (0.4)
9.6 (0.2)
19.7 (0.3)
21.3 (0.3)
1 to 2 times/wk
17.1 (0.3)
35.6 (0.4)
47.3 (0.4)
10.6 (0.2)
21.2 (0.3)
23.3 (0.4)
< 1 time/wk
17.7 (0.3)
35.7 (0.4)
46.6 (0.4)
10.7 (0.2)
21.9 (0.3)
25.0 (0.4)
(continued on next page)
RESEARCH
■■ 2026 Volume 126 Number 3
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
17
Table 2. Mean (SE) proportion of energy intake and UPF a (% of total energy) consumed in the morning, b midday, c and evening d
according to sociodemographic and behavioral characteristics of the UK population (aged 19 to 64 years), NDNS e 2008-2019
(n = 5749) (continued)
Characteristic
% of Total Energy Intake
% of Total Energy Intake From UPF
Morning
Midday
Evening
Morning
Midday
Evening
Never
18.3 (0.5)***
36.8 (0.6)***
44.8 (0.8)***
10.6 (0.4)**
20.2 (0.6)**
21.7 (0.6)**
Sleep duration f
<7 h/d
17.8 (0.3)
34.4 (0.3)
47.8 (0.4)
10.9 (0.2)
20.6 (0.3)
23.4 (0.4)
7—8 h/d
17.3 (0.2)
35.0 (0.3)
47.6 (0.3)
10.4 (0.2)
20.7 (0.2)
22.6 (0.3)
> 8 h/d
15.4 (0.4)***
37.2 (0.5)***
47.4 (0.6)
9.3 (0.3)***
22.1 (0.5)**
24.1 (0.6)
Weight loss diet f
No
17.1 (0.2)
35.2 (0.2)
47.7 (0.2)
10.4 (0.1)
21.1 (0.2)
23.3 (0.2)
Yes
17.5 (0.3)
35.2 (0.5)
47.3 (0.5)
10.3 (0.2)
20.1 (0.4)***
22.5 (0.5)
a UPF = ultra-processed food.
b Morning: 5:00 AM to 10:59 AM.
c Midday: 11:00 AM to 4:59 PM.
d Evening: 5:00 PM to 4:59 AM.
e NDNS = National Diet and Nutrition Survey.
f Univariate linear regression (first category of sociodemographic and behavioral variables that appear in the row as reference).
g White and Black Caribbean, White and Black African, White and Asian.
h Ethnicities not included in the listed categories.
i Univariate linear regression for linear trend across categories.
j Routine and manual occupations: jobs that involve repetitive tasks and are often regulated by a basic labor contract; intermediate occupations: are considered to be more difficult
conceptually than routine manual and nonmanual jobs, but less difficult than managerial or professional jobs (clerical, sales, services); lower managerial and professional occupations:
positions that have an attenuated form of service relationship. Employees in these groups generally plan and supervise operations on behalf of the employer under the direction of senior
managers; higher managerial and professional occupations: positions that involve general planning and supervision of operations, and that require specialized educational training. 45
k First quartile represents the least physically active individuals, and the fourth quartile represents the most physically active.
*P < .05.
**P < .01.
***P < .001.
RESEARCH
18
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
■■ 2026 Volume 126 Number 3
Table 4. Sensitivity analyses describing ORs a (95% CI) for the association between tertiles of total energy intake percentage
and percentage of energy intake from ultra-processed foods, along with a continuous 10% increase, analyzed across 3 periods
of the day, b and overweight c in the UK population (aged 19-64 years), NDNS, d 2008-2019 (n = 5749)
Variable
Energy Consumption (% of Total Energy) in Each Period of the Day
Tertiles e
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First (reference) Second
Third
←̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅OR (95% CI)̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅→
Period of the day
Morning
Model 4 gh
1
0.86 (0.72 to 1.03)
0.82 (0.69 to 0.98)
.026
0.90 (0.83 to 0.97)
Model 5 gi
1
0.88 (0.73 to 1.05)
0.83 (0.70 to 0.98)
.030
0.90 (0.83 to 0.97)
Model 6 gj
1
0.89 (0.74 to 1.06)
0.84 (0.71 to 1.00)
.053
0.90 (0.83 to 0.98)
Midday
Model 4 gh
1
0.84 (0.71 to 1.00)
0.94 (0.79 to 1.12)
.483
0.99 (0.93 to 1.06)
Model 5 gi
1
0.86 (0.73 to 1.03)
0.95 (0.79 to 1.13)
.546
0.99 (0.94 to 1.06)
Model 6 gj
1
0.88 (0.74 to 1.05)
0.96 (0.81 to 1.15)
.678
1.00 (0.94 1.07)
Evening
Model 4 gh
1
1.20 (0.99 to 1.44)
1.21 (1.01 to 1.45)
.036
1.06 (0.99 to 1.13)
Model 5 gi
1
1.21 (1.01 to 1.46)
1.21 (1.01 to 1.44)
.041
1.06 (0.99 to 1.13)
Model 6 gj
1
1.20 (0.99 to 1.44)
1.18 (0.99 to 1.44)
.070
1.05 (0.99 to 1.12)
Variable
Consumption of Ultra-Processed Foods (% of Total Energy)
Tertiles k
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First (reference) Second
Third
←̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅OR (95% CI)̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅→
Period of the day
Morning
Model 4 lm
1
0.83 (0.70 to 0.99)
0.87 (0.73 to 1.04)
.130
0.92 (0.83 to 1.02)
Model 5 ln
1
0.86 (0.73 to 1.02)
0.94 (0.78 to 1.14)
.536
0.97 (0.86 to 1.09)
Model 6 lo
1
0.85 (0.71 to 1.01)
0.89 (0.74 to 1.08)
.231
0.93 (0.83 to 1.05)
Midday
Model 4 lm
1
0.91 (0.76 to 1.10)
1.19 (0.99 to 1.44)
.057
1.11 (1.02 to 1.20)
Model 5 lp
1
0.91 (0.76 to 1.10)
1.22 (1.01 to 1.47)
.038
1.12 (1.03 to 1.21)
Model 6 lq
1
0.88 (0.73 to 1.06)
1.14 (0.94 to 1.38)
.189
1.09 (1.01 to 1.19)
Evening
Model 4 lm
1
1.07 (0.90 to 1.28)
1.31 (1.08 to 1.59)
.007
1.09 (1.02 to 1.17)
(continued on next page)
RESEARCH
■■ 2026 Volume 126 Number 3
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
19
Table 4. Sensitivity analyses describing ORs a (95% CI) for the association between tertiles of total energy intake percentage and
percentage of energy intake from ultra-processed foods, along with a continuous 10% increase, analyzed across 3 periods of
the day, b and overweight c in the UK population (aged 19-64 years), NDNS, d 2008-2019 (n = 5749) (continued)
Variable
Consumption of Ultra-Processed Foods (% of Total Energy)
Tertiles k
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First (reference)
Second
Third
Model 5 lr
1
1.07 (0.90 to 1.28)
1.34 (1.09 to 1.63)
.005
1.11 (1.03 to 1.19)
Model 6 ls
1
1.07 (0.89 to 1.28)
1.32 (1.06 to 1.64)
.012
1.10 (1.02 to 1.19)
a OR = odds ratio.
b Morning: 5:00 AM to 10:59 AM, midday: 11:00 AM to 4:59 PM, evening: 5:00 PM to 4:59 AM.
c Overweight: body mass index ≥25.
d NDNS = National Diet and Nutrition Survey.
e The tertiles were categorized based on cutoff points for energy consumption: Morning: first tertile: ≤13.1% of total energy intake; second tertile: 13.2%-20.6% of total energy intake; third
tertile: ≥20.7% of total energy intake. Midday: first tertile: ≤30.7% of total energy intake; second tertile: 30.8%-39.6% of total energy intake; third tertile: ≥39.7% of total energy intake.
Evening: first tertile: ≤42.8% of total energy intake; second tertile: 42.9%-52.1% of total energy intake; third tertile: ≥52.2% of total energy intake.
f P value for linear trend across tertiles of dietary contribution.
g Adjustment model for energy consumption.
h Model 4: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
percentage of total energy from ultra-processed foods.
i Model 5: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
energy intake and total energy from ultra-processed foods.
j Model 6: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, total
energy intake, plus percentage of daily energy from sugar, percentage of daily energy from saturated fat, and daily fruit and vegetable intake in grams.
k For ultra-processed foods intake: Morning: first tertile: ≤7.2% of total energy intake; second tertile: 7.3%-12.6% of total energy intake; third tertile: ≥12.7% of total energy intake. Midday:
first tertile: ≤16.3% of total energy intake; second tertile: 16.4%-24.5% of total energy intake; third tertile: ≥24.6% of total energy intake. Evening: first tertile: ≤17.1% of total energy
intake; second tertile: 17.2%-27.1% of total energy intake; third tertile: ≥27.2% of total energy intake.
l Adjustment model for ultra-processed foods consumption.
m Model 4: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
total energy intake.
n Model 5: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
percentage of energy from ultra-processed foods in the midday and evening, and total energy intake.
o Model 6: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption,
percentage of energy from ultra-processed foods in the midday and evening plus percentage of daily energy from sugar, percentage of daily energy from saturated fat, and daily fruit and
vegetable intake in grams.
p Model 5: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
percentage of energy from ultra-processed foods in the morning and evening, and total energy intake.
q Model 6: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption,
percentage of energy from ultra-processed foods in the morning and evening, plus percentage of daily energy from sugar, percentage of daily energy from saturated fat, and daily fruit
and vegetable intake in grams.
r Model 5: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
percentage of energy from ultra-processed foods in the morning and midday, and total energy intake.
s Model 6: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption,
percentage of energy from ultra-processed foods in the morning and midday, plus percentage of daily energy from sugar, percentage of daily energy from saturated fat, and daily fruit and
vegetable intake in grams.
RESEARCH
20
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
■■ 2026 Volume 126 Number 3
Table 6. Sensitivity analyses describing OR a (95% CI) for the association between tertiles of total energy intake percentage and
percentage of energy intake from ultra-processed foods, along with a continuous 10% increase, analyzed across 3 periods of
the day, b and obesity c in the UK population (aged 19-64 years), NDNS d 2008-2019 (n = 5749)
Variable
Energy Consumption (% of Total Energy) in Each Period of the Day
Tertiles e
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First (reference) Second
Third
←̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅OR (95% CI)̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅→
Period of the day
Morning
Model 4 gh
1
0.89 (0.74 to 1.08)
1.03 (0.85 to 1.25)
.733
0.99 (0.91 to 1.08)
Model 5 gi
1
0.91 (0.75 to 1.09)
1.03 (0.85 to 1.25)
.722
0.99 (0.91 to 1.08)
Model 6 gj
—
0.92 (0.76 to 1.11)
1.05 (0.86 to 1.27)
.619
0.99 (0.91 to 1.08)
Midday
Model 4 gh
1
0.85 (0.71 to 1.03)
0.83 (0.70 to 1.02)
.070
0.96 (0.89 to 1.04)
Model 5 gi
1
0.87 (0.72 to 1.05)
0.84 (0.68 to 1.02)
.079
0.97 (0.89 to 1.04)
Model 6 gj
—
0.89 (0.74 to 1.08)
0.85 (0.70 to 1.04)
.111
0.97 (0.90 to 1.05)
Evening
Model 4 gh
1
1.08 (0.88 to 1.31)
1.17 (0.96 to 1.43)
.111
1.04 (0.97 to 1.11)
Model 5 gi
1
1.08 (0.89 to 1.32)
1.17 (0.96 to 1.42)
.119
1.03 (0.97 to 1.11)
Model 6 gj
—
1.08 (0.89 to 1.32)
1.15 (0.95 to 1.40)
.153
1.03 (0.96 to 1.10)
Variable
Consumption of Ultra-Processed Foods (% of Total Energy)
Tertiles k
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First (reference) Second
Third
←̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅OR (95% CI)̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅→
Period of the day
Morning
Model 4 lm
1
1.00 (0.84 to 1.20)
1.14 (0.95 to 1.37)
.162
1.09 (0.98 to 1.21)
Model 5 ln
1
1.06 (0.88 to 1.27)
1.27 (1.04 to 1.54)
.017
1.18 (1.05 to 1.32)
Model 6 lo
—
1.04 (0.87 to 1.25)
1.23 (1.01 to 1.49)
.040
1.15 (1.03 to 1.29)
Midday
Model 4 lm
1
0.92 (0.75 to 1.11)
1.20 (0.99 to 1.44)
.061
1.08 (0.99 to 1.17)
Model 5 lp
1
0.92 (0.76 to 1.13)
1.27 (1.06 to 1.53)
.013
1.12 (1.03 to 1.21)
Model 6 lq
—
0.91 (0.75 to 1.11)
1.24 (1.03 to 1.50)
.032
1.11 (1.02 to 1.20)
Evening
Model 4 lm
1
1.08 (0.90 to 1.30)
1.40 (1.16 to 1.70)
.001
1.13 (1.05 to 1.22)
(continued on next page)
RESEARCH
■■ 2026 Volume 126 Number 3
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
21
Table 6. Sensitivity analyses describing OR a (95% CI) for the association between tertiles of total energy intake percentage and
percentage of energy intake from ultra-processed foods, along with a continuous 10% increase, analyzed across 3 periods of
the day, b and obesity c in the UK population (aged 19-64 years), NDNS d 2008-2019 (n = 5749) (continued)
Variable
Consumption of Ultra-Processed Foods (% of Total Energy)
Tertiles k
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First (reference)
Second
Third
Model 5 lr
1
1.08 (0.90 to 1.31)
1.47 (1.21 to 1.78)
<.001
1.16 (1.08 to 1.26)
Model 6 ls
—
1.09 (0.90 to 1.32)
1.48 (1.21 to 1.81)
<.001
1.18 (1.08 to 1.28)
a OR = odds ratio.
b Morning: 5:00 AM to 10:59 AM, midday: 11:00 AM to 4:59 PM, evening: 5:00 PM to 4:59 AM.
c Obesity: body mass index ≥30.
d NDNS = National Diet and Nutrition Survey.
e The tertiles were categorized based on cutoff points for energy consumption: Morning: first tertile: ≤13.1% of total energy intake; second tertile: 13.2—20.6% of total energy intake; third
tertile: ≥20.7% of total energy intake. Midday: first tertile: ≤30.7% of total energy intake; second tertile: 30.8—39.6% of total energy intake; third tertile: ≥39.7% of total energy intake.
Evening: first tertile: ≤42.8% of total energy intake; second tertile: 42.9—52.1% of total energy intake; third tertile: ≥52.2% of total energy intake.
f P value for linear trend across tertiles of dietary contribution.
g Adjustment model for energy consumption.
h Model 4: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
percentage of total energy from ultra-processed foods.
i Model 5: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
energy intake and total energy from ultra-processed foods.
j Model 6: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, total
energy intake, plus percentage of daily energy from sugar, percentage of daily energy from saturated fat, and daily fruit and vegetable intake in grams.
k For ultra-processed foods intake: Morning: first tertile: ≤7.2% of total energy intake; second tertile: 7.3%-12.6% of total energy intake; third tertile: ≥12.7% of total energy intake. Midday:
first tertile: ≤16.3% of total energy intake; second tertile: 16.4%-24.5% of total energy intake; third tertile: ≥24.6% of total energy intake. Evening: first tertile: ≤17.1% of total energy
intake; second tertile: 17.2%-27.1% of total energy intake; third tertile: ≥27.2% of total energy intake.
l Adjustment model for ultra-processed foods consumption.
m Model 4: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
total energy intake.
n Model 5: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
percentage of energy from ultra-processed foods in the midday and evening and total energy intake.
o Model 6: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption,
percentage of energy from ultra-processed foods in the midday and evening plus percentage of daily energy from sugar, percentage of daily energy from saturated fat, and daily fruit and
vegetable intake in grams.
p Model 5: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
percentage of energy from ultra-processed foods in the morning and evening, and total energy intake.
q Model 6: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption,
percentage of energy from ultra-processed foods in the morning and evening, plus percentage of daily energy from sugar, percentage of daily energy from saturated fat, and daily fruit
and vegetable intake in grams.
r Model 5: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
percentage of energy from ultra-processed foods in the morning and midday, and total energy intake.
s Model 6: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption,
percentage of energy from ultra-processed foods in the morning and midday, plus percentage of daily energy from sugar, percentage of daily energy from saturated fat, and daily fruit and
vegetable intake in grams.
RESEARCH
22
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
■■ 2026 Volume 126 Number 3
Table 8. Sensitivity analyses describing OR a (95% CI) for the association between tertiles of total energy intake percentage and
percentage of energy intake from ultra-processed foods, along with a continuous 10% increase, analyzed across 3 periods of
the day, b and abdominal obesity c in the UK population (aged 19-64 years), NDNS d 2008-2019 (n = 4244)
Variable
Energy Consumption (% of Total Energy) in Each Period of the Day
Tertiles e
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First (reference) Second
Third
←̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅OR (95% CI)̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅→
Period of the day
Morning
Model 4 gh
1
0.88 (0.72 to 1.07)
0.97 (0.80 to 1.18)
.794
0.96 (0.88 to 1.05)
Model 5 gi
1
0.89 (0.72 to 1.08)
0.97 (0.80 to 1.18)
.808
0.96 (0.88 to 1.05)
Model 6 gj
1
0.90 (0.74 to 1.11)
0.99 (0.82 to 1.21)
.983
0.97 (0.88 to 1.06)
Midday
Model 4 gh
1
0.87 (0.72 to 1.05)
0.94 (0.77 to 1.14)
.498
0.98 (0.90 to 1.06)
Model 5 gi
1
0.88 (0.73 to 1.07)
0.94 (0.77 to 1.15)
.516
0.98 (0.90 to 1.06)
Model 6 gj
1
0.90 (0.74 to 1.09)
0.94 (0.77 to 1.15)
.520
0.98 (0.90 to 1.06)
Evening
Model 4 gh
1
1.03 (0.85 to 1.25)
1.16 (0.95 to 1.42)
.145
1.04 (0.97 to 1.12)
Model 5 gi
1
1.04 (0.86 to 1.25)
1.16 (0.95 to 1.42)
.151
1.04 (0.97 to 1.12)
Model 6 gj
1
1.03 (0.85 to 1.25)
1.14 (0.93 to 1.40)
.195
1.04 (0.96 to 1.12)
Variable
Consumption of Ultra-Processed Foods (% of Total Energy)
Tertiles k
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First (reference) Second
Third
←̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅OR (95% CI)̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅→
Period of the day
Morning
Model 4 lm
1
0.92 (0.77 to 1.11)
1.06 (0.87 to 1.28)
.574
1.01 (0.90 to 1.14)
Model 5 ln
1
0.95 (0.79 to 1.14)
1.13 (0.93 to 1.37)
.227
1.06 (0.94 to 1.20)
Model 6 lo
1
0.93 (0.77 to 1.13)
1.09 (0.90 to 1.33)
.389
1.04 (0.92 to 1.17)
Midday
Model 4 lm
1
0.78 (0.64 to 0.95)
1.10 (0.90 to 1.35)
.389
1.05 (0.96 to 1.15)
Model 5 lp
1
0.79 (0.65 to 0.96)
1.14 (0.94 to 1.40)
.229
1.07 (0.99 to 1.17)
Model 6 lq
1
0.78 (0.64 to 0.95)
1.10 (0.90 to 1.35)
.409
1.06 (0.97 to 1.15)
Evening
Model 4 lm
1
1.00 (0.83 to 1.20)
1.30 (1.06 to 1.59)
.013
1.11 (1.02 to 1.21)
(continued on next page)
RESEARCH
■■ 2026 Volume 126 Number 3
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
23
Table 8. Sensitivity analyses describing OR a (95% CI) for the association between tertiles of total energy intake percentage and
percentage of energy intake from ultra-processed foods, along with a continuous 10% increase, analyzed across 3 periods of
the day, b and abdominal obesity c in the UK population (aged 19-64 years), NDNS d 2008-2019 (n = 4244) (continued)
Variable
Consumption of Ultra-Processed Foods (% of Total Energy)
Tertiles k
P for trendf
Continuous (10% increase in
the consumption), OR (95% CI)
First (reference)
Second
Third
Model 5 lr
1
1.00 (0.83 to 1.21)
1.32 (1.08 to 1.62)
.008
1.12 (1.03 to 1.22)
Model 6 ls
1
1.00 (0.83 to 1.20)
1.30 (1.06 to 1.62)
.016
1.13 (1.03 to 1.23)
a OR = odds ratio.
b Morning: 5:00 AM to 10:59 AM, midday: 11:00 AM to 4:59 PM, evening: 5:00 PM to 4:59 AM.
c Abdominal obesity: waist circumference ≥ 102 cm for men and ≥ 88 cm for women.
d NDNS = National Diet and Nutrition Survey.
e The tertiles were categorized based on cutoff points for energy consumption: Morning: first tertile: ≤13.1% of total energy intake; second tertile: 13.2%-20.6% of total energy intake; third
tertile: ≥20.7% of total energy intake. Midday: first tertile: ≤30.7% of total energy intake; second tertile: 30.8%-39.6% of total energy intake; third tertile: ≥39.7% of total energy intake.
Evening: first tertile: ≤42.8% of total energy intake; second tertile: 42.9%-52.1% of total energy intake; third tertile: ≥52.2% of total energy intake.
f P value for linear trend across tertiles of dietary contribution.
g Adjustment model for energy consumption.
h Model 4: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
percentage of total energy from ultra-processed foods.
i Model 5: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
energy intake and total energy from ultra-processed foods.
j Model 6: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, total
energy intake, plus percentage of daily energy from sugar, percentage of daily energy from saturated fat, and daily fruit and vegetable intake in grams.
k For ultra-processed foods intake: Morning: first tertile: ≤7.2% of total energy intake; second tertile: 7.3%-12.6% of total energy intake; third tertile: ≥12.7% of total energy intake. Midday:
first tertile: ≤16.3% of total energy intake; second tertile: 16.4%-24.5% of total energy intake; third tertile: ≥24.6% of total energy intake. Evening: first tertile: ≤17.1% of total energy
intake; second tertile: 17.2%-27.1% of total energy intake; third tertile: ≥27.2% of total energy intake.
l Adjustment model for ultra-processed foods consumption.
m Model 4: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
total energy intake.
n Model 5: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
percentage of energy from ultra-processed foods in the midday and evening, and total energy intake.
o Model 6: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption,
percentage of energy from ultra-processed foods in the midday and evening plus percentage of daily energy from sugar, percentage of daily energy from saturated fat, and daily fruit and
vegetable intake in grams.
p Model 5: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
percentage of energy from ultra-processed foods in the morning and evening, and total energy intake.
q Model 6: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption,
percentage of energy from ultra-processed foods in the morning and evening, plus % of daily energy from sugar, percentage of daily energy from saturated fat, and daily fruit and
vegetable intake in grams.
r Model 5: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption, plus
percentage of energy from ultra-processed foods in the morning and midday, and total energy intake.
s Model 6: adjusted for sex, age, region, survey year, social class occupation, physical activity, smoking, sleep duration, following a weight-loss diet, frequency of alcohol consumption,
percentage of energy from ultra-processed foods in the morning and midday, plus % of daily energy from sugar, percentage of daily energy from saturated fat, and daily fruit and
vegetable intake in grams.
RESEARCH
24
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS
■■ 2026 Volume 126 Number 3