Since January 2020 Elsevier has created a COVID-19 resource centre with

free information in English and Mandarin on the novel coronavirus COVID-

19. The COVID-19 resource centre is hosted on Elsevier Connect, the
company's public news and information website.

Elsevier hereby grants permission to make all its COVID-19-related

research that is available on the COVID-19 resource centre - including this
research content - immediately available in PubMed Central and other
publicly funded repositories, such as the WHO COVID database with rights
for unrestricted research re-use and analyses in any form or by any means
with acknowledgement of the original source. These permissions are
granted for free by Elsevier for as long as the COVID-19 resource centre
remains active.
 Brain, Behavior, and Immunity 96 (2021) 18–27

Contents lists available at ScienceDirect

Brain Behavior and Immunity

journal homepage: www.elsevier.com/locate/ybrbi

Lifestyle risk factors and infectious disease mortality, including COVID-19,

among middle aged and older adults: Evidence from a community-based
cohort study in the United Kingdom
Matthew N. Ahmadi a, *, Bo-Huei Huang a, Elif Inan-Eroglu a, Mark Hamer b,
Emmanuel Stamatakis a
a
Charles Perkins Centre, School of Health Sciences, Faculty of Medicine and Health, The University of Sydney, NSW, Australia
b
Division of Surgery and Interventional Science, Faculty Medical Sciences, University College London, London, UK

A R T I C L E I N F O A B S T R A C T

Keywords: In this community-based cohort study, we investigated the relationship between combinations of modifiable
Physical activity lifestyle risk factors and infectious disease mortality. Participants were 468,569 men and women (56.5 ± 8.1,
Sedentary behaviour 54.6% women) residing in the United Kingdom. Lifestyle indexes included traditional and emerging lifestyle risk
Sleep
factors based on health guidelines and best practice recommendations for: physical activity, sedentary behaviour,
Diet
sleep quality, diet quality, alcohol consumption, and smoking status. The main outcome was mortality from
Alcohol
Smoking infectious diseases, including pneumonia, and coronavirus disease 2019 (COVID-19). Meeting public health
Population cohort guidelines or best practice recommendations among combinations of lifestyle risk factors was inversely associ­
ated with mortality. Hazard ratios ranged between 0.26 (0.23–0.30) to 0.69 (0.60–0.79) for infectious disease
and pneumonia. Among participants with pre-existing cardiovascular disease or cancer, hazard ratios ranged
between 0.30 (0.25–0.34) to 0.73 (0.60–0.89). COVID-19 mortality risk ranged between 0.42 (0.28–0.63) to 0.75
(0.49–1.13). We found a beneficial dose–response association with a higher lifestyle index against mortality that
was consistent across sex, age, BMI, and socioeconomic status. There was limited evidence of synergistic in­
teractions between most lifestyle behaviour pairs, suggesting that the dose–response relationship among different
lifestyle behaviours is not greater than the sum of the risk induced by each behaviour. Improvements in lifestyle
risk factors and meeting public health guidelines or best practice recommendations could be used as an ancillary
measure to ameliorate infectious disease mortality.

1. Introduction effects of infectious diseases are significant decreases in quality of life for
individuals, in addition to clinical and economic burden across com­
The increase in annual infectious disease cases and the proliferation munities. The direct costs of treating community acquired pneumonia is
of resistant strains of pathogens threatens the successful treatment of estimated to be between 3.7 and 12.1 billion USD annually, with an
community acquired infections (Cassini et al., 2019; Marston et al., additional $1.8 to $3.6 USD billion in indirect costs of economic pro­
2016; Tacconelli et al., 2018). An additional 60,900 deaths occur ductivity losses (Song et al., 2011; Welte et al., 2012; Weycker et al.,
annually due to antimicrobial resistance across the United States and 2010). Most recently, severe acute respiratory syndrome coronavirus 2,
Europe, whilst the incidence of sepsis now exceeds 48 million cases which causes coronavirus disease 2019 (COVID-19) has led to a global
worldwide (Gelband et al., 2015; Kadri, 2020; Rudd et al., 2020). Res­ health pandemic.
piratory infections, such as pneumonia, are the leading cause of death in Severe progression of infectious diseases is associated with multiple
developing countries, and the largest contributor to the overall burden lifestyle risk factors (Baik et al., 2000; Hamer et al., 2019). The role of
of disease in the world measured in disability adjusted life years (Ferkol lifestyle behaviours and risk of infectious disease mortality is becoming
and Schraufnagel, 2014; Nair et al., 2011). Among the detrimental increasingly important. This requires a better understanding of the

* Corresponding author at: Charles Perkins Centre, School of Health Sciences, Faculty of Medicine and Health, Level 6, 1 John Hopkins Drive, Camperdown, NSW
2006, Australia.
E-mail address: [email protected] (M.N. Ahmadi).

https://doi.org/10.1016/j.bbi.2021.04.022
Received 22 January 2021; Received in revised form 18 April 2021; Accepted 28 April 2021
Available online 1 May 2021
0889-1591/© 2021 Elsevier Inc. All rights reserved.
M.N. Ahmadi et al. Brain Behavior and Immunity 96 (2021) 18–27

relationship between combinations of different lifestyle risk factors that All participants provided written informed consent.
may increase the risk of mortality. To date, studies have only examined Participants consented to the use of their de-identified data and ac­
the individual associations of lifestyle risk factors and infectious diseases cess to their national health-related hospital and death records. Exclu­
(Hamer et al., 2019; Paulsen et al., 2017; Wang et al., 2017). For sions prior to the onset of analyses included participants who did not
example, smokers have shown an increased risk of both bacterial and have usable physical activity, sedentary behaviour, sleep, diet, alcohol
viral infection-related mortality (Carter et al., 2015; Huttunen et al., consumption, and smoking history information (n = 20,144). We then
2011), and poor diet quality has been associated with low resistance to excluded any remaining participants with an incomplete covariate
infections (Ambrus and Ambrus, 2004; Gordon, 1968; Katona and profile (n = 13,903). Missing values for a category were imputed using
Katona-Apte, 2008; Scrimshaw and SanGiovanni, 1997). Further, multivariate imputation by chained equations if at least 80% of all other
among individuals, who never drink alcohol or moderately drink, in­ data was present (Buuren and Groothuis-Oudshoorn, 2010).
fectious disease risk does not differ; risk, however, increases substan­
tially among heavy drinkers, leading to higher rates of morbidity and
mortality (Rehm et al., 2010; Samokhvalov et al., 2010). Higher vol­ 2.2. Measurements
umes of physical activity are associated with a lower incidence of in­
fectious diseases and related mortality (Baik et al., 2000; Hamer et al., During the baseline recruitment visit, participants were asked to
2019). Most recently, physical inactivity, a history of smoking, and complete a self-administered touchscreen questionnaire, which included
excessive alcohol consumption have been identified as lifestyle risk questions on socio-demographics and lifestyle exposures.
factors that contribute to increased risk of hospitalizations due to
COVID-19. More than a 4-fold increase in hospitalisation was observed 2.2.1. Physical activity
among participants engaging in all unfavourable behaviours (Hamer Physical activity was measured using the International Physical Ac­
et al., 2020). The additive influence of multiple lifestyle behaviours tivity Questionnaire (IPAQ) short form (Craig et al., 2003) and included
against infection related mortality, remains unknown. items on frequency and duration of walking, moderate intensity activity,
Prior literature suggests different lifestyle behaviours may have and vigorous intensity activity. Physical activity was expressed as MET-
synergistic effects (Stamatakis et al., 2015; Xiao et al., 2014). The risk of min/week and based on the IPAQ scoring procedure, participants who
immune-suppressive effects from an unhealthy lifestyle behaviour, such attained 600 MET-min/week met the physical activity guidelines of 150
as physical inactivity, may be amplified by unhealthy sleep habits and min of moderate-vigorous physical activity a week (Bull et al., 2020).
high sedentary time. Among the possible consequences is an increased Participants were classified as inactive if they attained 0 MET-min/
risk of hospitalisations and mortality events caused by respiratory in­ week, insufficiently active if they had<600 MET-min/week, and suffi­
fections (Fletcher et al., 2018; Ibarra-Coronado et al., 2015; Nieman ciently active if they had at least 600 MET-min/week.
et al., 2011; Opp and Krueger, 2015; Sallis et al., 2020). Studies that
have observed inconsistent relationships between inadequate sleep 2.2.2. Sedentary time
duration and respiratory infections did not consider the role of sleep Total sedentary time was based on three questions enquiring about
quality or the influence of combined lifestyle behaviours (Irwin, 2015; daily hours of TV, PC screen-based activities and driving. Sedentary time
Prather and Leung, 2016). Considering that individual lifestyle risk was classified as high (greater than7 h/d), medium (4 to 7 h/d) , or low
factors may have an additive influence on mortality risk, investigating (>=4 h/d).(Chau et al., 2015, 2013)
combinations of lifestyle behaviours together will elucidate more clini­
cally relevant information (Ding et al., 2015; Dunstan et al., 2012; 2.2.3. Sleep quality
Hamer et al., 2014; Hamilton et al., 2007; Stamatakis et al., 2015). Sleep quality was assessed using five healthy sleep characteristics
To our knowledge, no studies have examined the associations be­ which included (Fan et al., 2020): Morning chronotype, sleep duration
tween both established and emerging lifestyle risk factors, with infec­ (7–9 h), not usually insomnia, no snoring, and no frequent daytime
tious disease that include: physical activity, sedentary behaviour, sleep sleepiness. Following the sleep quality scoring by Fan et al, participants
quality, diet quality, alcohol consumption, and smoking status. The aim were given a score of “1′′ for every question they answered “yes” (Fan
of this study was to examine the association of combined lifestyle risk et al., 2020). Component scores were summed and participants were
factor indexes and risk of infectious disease mortality, including mor­ classified as having poor sleep quality (score = 0 to 1), moderate sleep
tality due to pneumonia and COVID-19. quality (score = 2 to 3), or good sleep quality (score = 4 to 5).

2. Materials and methods 2.2.4. Diet Quality

Diet quality was assessed using a modified Alternate Healthy Eating
2.1. Participants Index (AHEI), which is based on foods and nutrients that have been
shown to be predictive of disease (Chiuve et al., 2012). Participants are
The UK Biobank is a prospective cohort study which aims to inves­ given a score of 0 to 10 for each food category and the scoring criteria for
tigate the genetic, lifestyle, and environmental causes of a range of the AHEI is described in detail elsewhere (McCullough et al., 2002). For
diseases (Allen et al., 2012; Sudlow et al., 2015; UK Biobank, 2007). the current study, participants reported their daily diet in four cate­
Between 2006 and 2010, 502,656 adults aged between 40 and 69 years gories: fruits, vegetables, whole grains, and portions of red meat/ pro­
(229,182 men and 273, 474 women) were recruited. All participants cessed meat. All the component scores were summed and participants
were registered with the UK National Health Service (NHS) and lived were classified as having poor diet quality (score = 0 to 10), moderate
within ~ 40 km of 1 of the 22 study assessment centres. The UK Biobank diet quality (score = 11 to 30), and good diet quality (score = 31 to 40).
invited ~ 9.2 million people to participate through postal invitation
with a telephone follow-up, with a response rate of 5.7%. The UK Bio­ 2.2.5. Alcohol consumption
bank has approval from the North West Multi-Centre Research Ethics Participants reported their alcohol drinking status as: Never drinker,
Committee, the National Information Governance Board for Health and ex-drinker, or current drinker. Participants who were current drinkers,
Social Care in England and Wales, and the Community Health Index were asked about average weekly consumption of wine, spirits, and beer
Advisory Group in Scotland. In addition, an independent Ethics and intake. Based on current UK guidelines, participants were categorised as
Governance Council was formed in 2004 to oversee UK Biobank’s never drinkers, ex-drinkers, within guideline drinkers (<14 UK units of
continuous adherence to the Ethics and Governance Framework, which alcohol/wk; 1 unit = 8 g of alcohol), or above guideline drinkers (≥14
were developed for the study (http://www.uk-biobank.ac.uk/ethics/). UK units of alcohol/wk) (Health, 2016; Rosenberg et al., 2018).

19
M.N. Ahmadi et al. Brain Behavior and Immunity 96 (2021) 18–27

2.2.6. Smoking status interaction (RERI); attributable proportion due to interaction (AP); and
Participants were asked to report their current smoking status. They the synergistic effects (S). RERI and AP would be equal to zero and S
were classified as never smokers, previous smokers, and current would be equal to 1 if there is no interaction present between pairs of
smokers. behaviours (Andersson et al., 2005; Källberg et al., 2006). To reduce the
possibility of spurious associations due to reverse causation, we repeated
2.2.7. Healthy Lifestyle Index analyses after excluding all participants who died in the first five years of
Each lifestyle behaviour, except for alcohol consumption, was follow-up. Sensitivity analysis was conducted for infectious disease
assigned a score ranging from zero (least healthy behaviour) to two mortality by excluding all infectious disease mortality due to pneu­
(most healthy behaviour). Alcohol consumption was categorized into monia. In another set of sensitivity analyses, we excluded participants
four groups on the basis that ex-alcohol drinkers are generally at a with a history of smoking, cardiovascular disease, and cancer and
higher risk of all-cause mortality than lifelong never drinkers (Knott included self-reported health as a covariate. We also assessed the asso­
et al., 2015; Perreault et al., 2017). ciations of individual lifestyle risk factors with mortality among par­
Table 1 describes the categorisation for all six lifestyle risk factors ticipants who had the least healthy lifestyle index score. All analysis was
and the corresponding scores that were assigned to participants. All six performed using R software (version 4.0.2).
individual lifestyle behaviour scores were added together to obtain a
healthy lifestyle index score. Never drinkers and guideline drinkers were 3. Results
given the same index score because the behaviours have both been
shown to have similar protective health benefits (Friedman and Klatsky, 3.1. Sample
1993). A lifestyle behaviour score of 0–4 represented the least healthy
group and was an indication that participants had a score of 0 in mul­ Our analysis included 468,569 participants. Supplemental Fig. 1
tiple behaviour categories without a score of 2 in more than two cate­ provides a detailed flowchart of participants who were excluded due to
gories. A score of 10–12 represented the healthiest group, and was an missing or unusable data. The participants included in the study had a
indication that participants had a score of 2 in at least four out of the six corresponding 4,176 deaths due to infectious diseases and 3,170 deaths
categories. due to pneumonia. There were an additional 387 deaths due to COVID-
19. The number of participants with an event for each type of infectious
2.3. Outcomes disease is listed in Supplemental Table 1 . The absolute risk and person-
time rate for each healthy lifestyle index category is displayed in Sup­
Participant data was linked to the national datasets from the Na­ plemental Table 13. Table 2 presents the characteristics of the popula­
tional Health Service (NHS) Information Centre (England and Wales) tion at baseline. The median follow-up time was 11.3 years (IQR: 10.5 to
and the NHS Central Register Scotland (Scotland). Complete follow-up 11.9 years) with a total of 5,166,793 person-years of follow-up before
was available through June 28th, 2020. Mortality incidence data were death or censoring, and 54.6% of the participants were female. The
coded using the 10th Revision of the International Classification of average age of participants at baseline was 56.5 (±8.1) years. Among the
Diseases (ICD-10) and included if it was the underlying or contributory 29,281 participants classified as having the lowest healthy lifestyle
cause of death. Infectious disease mortality was identified using the behaviour index score (0 to 5 score), 62.7% were inactive, 41.9% re­
following ICD-10 codes: A00-B99 and J09-J18 (pneumonia). COVID-19 ported more than 7 h per day in discretionary sedentary time, and 14.4%
mortality was identified using ICD-10 codes U07.1-U07.2. had poor sleep quality. Among these participants, 53.8% had poor diet
quality, 45.8% were current smokers, and 87.3% were ex-drinkers or
2.4. Statistical analyses consuming more than 14 units of alcohol per week. Healthy lifestyle
behaviour index scores were more prevalent among females, those with
Hazard ratios (HRs) and 95% confidence intervals (CIs) were esti­
mated using Cox proportional hazards regression models for individual Table 1
lifestyle risk factors and healthy lifestyle index with infectious disease Lifestyle risk factor categories and index score.
outcome. The reference group for each individual lifestyle risk factor
Risk factor Category Definition Index
was the least favourable lifestyle behaviour. The timescale was in cal­ score
endar time (months). Multivariable proportional regression models
Inactive 0 min 0
were adjusted for the following covariates: age at baseline, sex, socio­
Physical Insufficient 1–149 min 1
economic status based on the Townsend deprivation index (Townsend Activity Sufficient ≥ 150 min 2
et al., 1988), ethnicity (White, South Asian, Black, Chinese, and other), High SB greater than 7 hrs 0
body mass index (weight divided by squared height), corticosteroid use, Sedentary Mod SB 4–7 hrs 1
and comorbidities (cardiovascular diseases, cancers, diabetes, chronic behaviour
Behaviour Low SB < 4 hrs 2
respiratory disease [ICD-10 codes J.40 to J.47], liver disease, end-stage Poor index ≤ 1 sleep score 0
renal disease, immune disorders/HIV, and hypertension defined as ≥ Sleep Moderate index sleep score 1
140/90 mmHg) slepattern
To examine the associations between individual lifestyle risk factors Good index greater than 3 sleep 2
score
and healthy lifestyle index with COVID-19 mortality, we used binomial
Poor quality 0 diet score 0
regression to account for all mortality events occurring only between Diet Moderate quality 1 diet score 1
March to June 2020. The adjusted risk ratio models included all the Good quality 2 diet score 2
covariates previously listed. Ex-drinker * 0
To evaluate the consistency of our findings in different population *Alcohol Above guideline * 1
Never drinker * 2
subgroups, we conducted a set of stratified analysis by: sex (male; fe­ Within guideline * 2
male); age (<50 years; < 60 years; and ≥ 60 years); body mass index Current 0
(BMI) category (normal weight; overweight; obese); and socioeconomic Smoking Previous 1
status (Townsend index quintiles). In addition, we examined the asso­ Never 2
ciations among participants who had a history of cardiovascular disease *In the United Kindom, 1 unit = 8g of alcohol; Heavy drinker ≥14 units; To
and cancer. Three measures were used to investigate interaction be­ derive a combined lifestyle behaviour index score, never drinker and within
tween pairs of lifestyle behaviours: The relative excess risk due to guideline drinker were combined into the same category.

20
M.N. Ahmadi et al. Brain Behavior and Immunity 96 (2021) 18–27

Table 2
Study population characteristics at baseline. Values are means (SD) unless stated otherwise.
Lifestyle Behaviour Index Score
Characteristic 0–4 6 7 8 9 10 10–12
Sample size (n) 29,281 33,641 54,524 75,083 84,975 80,357 110,582
Follow-up duration (years) 10.7 (2.0) 10.9 (1.8) 11.0 (1.6) 11.0 (1.5) 11.0 (1.4) 11.1 (1.4) 11.1 (1.3)
Age (years) 55.7 (8.0) 56.3 (7.9) 56.5 (8.0) 56.6 (8.0) 56.5 (8.1) 56.5 (8.1) 56.6 (8.3)
Women (%) 33.9 38.1 42.3 47.8 54.2 61.8 71.0
Physical activity, n (%)
Inactive 18,355 (62.7) 14,267 (42.4) 17,485 (32.1) 18,280 (24.3) 15,400 (18.1) 8,859 (11.0) 2,575 (2.3)
Insufficient 7935 (27.1) 11,830 (35.2) 19,565 (35.9) 25,525 (34.0) 27,763 (32.7) 25,445 (31.7) 22,546 (20.4)
Sufficient 2991 (10.2) 7,544 (22.4) 17,474 (32.0) 31,278 (41.7) 41,812 (49.2) 46,053 (57.3) 85,461 (77.3)
Sedentary
High 12,280 (41.9) 9,328 (27.7) 10,976 (20.1 10,513 (14.0) 7,917 (9.3) 4,778 (5.9) 1,693 (1.5)
Mod 14,857 (50.7) 19,861 (59.0) 33,738 (61.9) 46,350 (61.7) 50,083 (58.9) 43,161 (53.7) 41,879 (37.9)
Low 2,144 (7.3) 4,452 (13.2) 9,810 (18.0) 18,220 (24.3) 26,975 (31.7) 32,418 (40.3) 67,010 (60.6)
Sleep
Poor quality (0–1) 4,223 (14.4) 2,618 (7.8) 2,702 (5.0) 2,319 (3.1) 1,614 (1.9) 831 (1.0) 213 (0.2)
Moderate quality (2–3) 19,984 (68.2) 21,595 (64.2) 32,430 (59.5) 39,146 (52.1) 37,002 (43.5) 28,627 (35.6) 22,351 (20.2)
Good quality (4–5) 5,074 (17.3) 9,428 (28.0) 19,392 (35.6) 33,618 (44.8) 46,359 (54.6) 50,899 (63.3) 88,018 (79.6)
Diet
Poor quality 15,755 (53.8) 11,986 (35.6) 14,254 (26.1) 13,914 (18.5) 10,843 (12.8) 6,407 (8.0) 2,105 (1.9)
Moderate quality 11,110 (37.9) 15,717 (46.7) 26,137 (47.9) 34,933 (46.5) 36,892 (43.4) 32,400 (40.3) 29,704 (26.9)
Good quality 2,416 (8.3) 5,938 (17.7) 14,133 (25.9) 26,236 (34.9) 37,240 (43.8) 41,550 (51.7) 78,773 (71.2)
Alcohol*
Ex-drinker 5,378 (18.4) 3,074 (9.1) 3,318 (6.1) 2,761 (3.7) 1,432 (1.7) 395 (0.5) 0 (0)
Above guideline 20,170 (68.9) 22,511 (66.9) 33,212 (60.9) 38,912 (51.8) 32,824 (38.6) 17,903 (22.3) 4,450 (4.0)
Non-drinker 662 (2.3) 1,116 (3.3) 2,251 (4.1) 3,565 (4.7) 4,390 (5.2) 3,981 (5.0) 3,625 (3.3)
Within guideline 3,071 (10.5) 6,940 (20.6) 15,743 (28.9) 29,845 (39.7) 46,329 (54.5) 58,078 (72.3) 102,507 (92.7)
Smoking
Current 13,416 (45.8) 9,168 (27.3) 9,691 (17.8) 8,091 (10.8) 5,057 (6.0) 2,466 (3.1) 692 (0.6)
Previous 12,131 (41.4) 16,390 (48.7) 26,429 (48.5) 33,555 (44.7) 32,164 (37.9) 23,784 (29.6) 18,537 (16.8)
Never 3,734 (12.8) 8,083 (24.0) 18,404 (33.8) 33,437 (44.5) 47,754 (56.2) 54,107 (67.3) 91,353 (82.6)
Townsend deprivation index [median − 0.9 (-3.0, 2.4) − 1.7 (-3.4, 1.3) − 2.0 (-3.5, 0.9) − 2.1 (-3.6, 0.5) − 2.3 (-3.7, 0.2) − 2.4 (-3.7, 0.0) − 2.5 (-3.8, − 0.3)
(IQR)]
Body Mass Index 28.8 (5.3) 28.6 (5.1) 28.2 (4.9) 27.9 (4.8) 27.5 (4.7) 27.0 (4.6) 26.1 (4.3)
Ethnicity (%)
White 95.8 95.9 95.5 95.2 94.8 94.5 94.9
South Asian 1.3 1.4 1.6 1.8 2.0 2.0 1.7
Black 1.3 1.2 1.4 1.4 1.4 1.6 1.6
Chinese 0.1 0.2 0.2 0.3 0.3 0.4 0.4
Other 1.5 1.4 1.3 1.3 1.4 1.5 1.4
Comorbidities (%)
Cancer 8.2 7.9 8.3 8.1 8.3 8.4 8.5
Cardiovascular disease 38.3 35.8 33.9 31.4 29.0 27.0 23.7
Diabetes 8.0 7.0 6.1 5.4 4.8 4.1 3.1
Chronic respiratory illness 16.4 14.6 13.6 13.2 12.6 12.2 11.3
Liver disease 0.6 0.4 0.3 0.3 0.2 0.2 0.2
End-stage renal disease 0.2 0.1 0.1 0.1 <0.1 <0.1 <0.1
Immune disorders/HIV 0.5 0.4 0.4 0.4 0.4 0.3 0.3

lower body mass index, and higher socioeconomic status. Individuals who were ex-smokers or had never smoked had a
significantly lower risk for infectious disease mortality (ex-smokers:
0.50 [ 0.46 to 0.54]; never smokers: 0.37 [0.34 to 0.41]) and pneumonia
3.2. Individual lifestyle risk factors mortality (ex-smokers: 0.46 [0.42 to 0.51]; never smokers: 0.33 [0.30 to
0.36]) compared to individuals who were current smokers. In contrast,
3.2.1. Infectious disease and pneumonia mortality there was weak evidence for an association of diet quality. Compared to
The hazard ratios of each individual lifestyle behaviour for infectious those with the poorest diet quality (referent group), only participants
disease and pneumonia mortality are provided in Tables 3 and 4, with good diet quality had an attenuated risk for infectious disease
respectively. In the fully adjusted models, we found a direct association mortality (0.85 [0.77 to 0.93]) and pneumonia mortality (0.82 [0.75 to
between all three movement behaviours (physical activity, sedentary 0.91]). When ex-drinkers (referent group) were compared to current
behaviour, sleep) and infectious disease mortality and pneumonia drinkers we observed a 44% to 47% reduction in infectious disease
mortality. When individuals with good sleep quality were compared to mortality (within guideline drinkers: 0.56 [0.50 to 0.63]; above guide­
individuals with poor sleep quality, we observed a 20% decrease in in­ line drinkers: 0.53 [0.47 to 0.60]).
fectious disease mortality (HR [95%CIs]: 0.80 [0.70 to 0.92]) and
pneumonia mortality (0.80 [0.68 to 0.95]). The associations for seden­ 3.2.2. COVID-19 mortality
tary time followed the same pattern, and when individuals with low Table 5 shows the risk ratio of each lifestyle behaviour category for
sedentary time were compared to individuals with high sedentary time, COVID-19 mortality. In the fully adjusted models, individuals who were
we observed ≈21% decrease in infectious disease mortality (0.78 [0.72 sufficiently active (RR [95%CIs]: 0.70 [0.54 to 0.89]), had never smoked
to 0.87]) and pneumonia mortality (0.79 [0.67 to 0.94]). Compara­ (0.54 [0.39 to 0.74]), and were current drinkers (within guideline
tively, when individuals who were sufficiently active were compared to drinkers: 0.60 (0.40 to 0.89]; above guideline drinkers: 0.62 [0.41 to
those who were inactive, we observed a 37% decrease in infectious 0.93]) had lower COVID-19 mortality risk compared to the referent
disease mortality (0.64 [0.59 to 0.69]) and pneumonia mortality (0.63 groups of each lifestyle risk factor.
[0.58 to 0.69]) (Tables 3 and 4).

21
M.N. Ahmadi et al. Brain Behavior and Immunity 96 (2021) 18–27

Table 3
Lifestyle risk factors and infectious disease mortality hazard ratio.
Risk factor N Events Model 1 Model 2
HR (95% CI) HR (95% CI)

Physical
Activity Inactive 95,221 1288 1.00 (ref) 1.00 (ref)
Insufficient 140,609 1173 0.65 (0.60, 0.70) 0.77 (0.71, 0.83)
Sufficient 232,613 1715 0.52 (0.48, 0.56) 0.64 (0.59, 0.69)

Sedentary
Behaviour High 57,485 748 1.00 (ref) 1.00 (ref)
Moderate 249,929 2354 0.70 (0.65, 0.76) 0.86 (0.79, 0.93)
Low 161,029 1074 0.60 (0.55, 0.66) 0.79 (0.72, 0.87)

Sleep
Poor 14,520 212 1.00 (ref) 1.00 (ref)
Moderate 201,135 2004 0.66 (0.57, 0.76) 0.83 (0.72, 0.97)
Good 252,788 1960 0.54 (0.47, 0.62) 0.80 (0.70, 0.92)

Diet
Poor 75,264 750 1.00 (ref) 1.00 (ref)
Moderate 186,893 1668 0.82 (0.75, 0.89) 0.94 (0.87, 1.03)
Good 206,286 1758 0.67 (0.62, 0.73) 0.85 (0.77, 0.93)

Alcohol
Ex-drinker 16,257 340 1.00 (ref) 1.00 (ref)
Above guideline 169,542 1584 0.39 (0.35, 0.44) 0.53 (0.47, 0.60)
Never drinker 19,522 211 0.55 (0.46, 0.65) 0.76 (0.64, 0.91)
Within guideline 261,842 2041 0.40 (0.35, 0.45) 0.56 (0.50, 0.63)

Smoking
Current 48,581 905 1.00 (ref) 1.00 (ref)
Previous 162,990 1814 0.42 (0.39, 0.45) 0.50 (0.46, 0.54)
Never 256,872 1457 0.28 (0.26, 0.30) 0.37 (0.34, 0.41)

Table 4
Lifestyle risk factors and pneumonia mortality hazard ratio.
Risk factor N Events Model 1 Model 2
HR (95% CI) HR (95% CI)

Physical
Activity Inactive 95,221 984 1.00 (ref) 1.00 (ref)
Insufficient 140,609 893 0.64 (0.59, 0.71) 0.77 (0.70, 0.84)
Sufficient 232,613 1293 0.51 (0.47, 0.55) 0.63 (0.58, 0.69)

Sedentary
Behaviour High 57,485 583 1.00 (ref) 1.00 (ref)
Moderate 249,929 1773 0.68 (0.62, 0.75) 0.83 (0.76, 0.92)
Low 161,029 814 0.60 (0.53, 0.66) 0.78 (0.70, 0.87)

Sleep
Poor 14,520 160 1.00 (ref) 1.00 (ref)
Moderate 201,135 1521 0.66 (0.56, 0.78) 0.83 (0.70, 0.98)
Good 252,788 1489 0.54 (0.46, 0.63) 0.80 (0.68, 0.95)

Diet
Poor 75,264 584 1.00 (ref) 1.00 (ref)
Moderate 186,893 1278 0.80 (0.73, 0.88) 0.94 (0.85, 1.03)
Good 206,286 1308 0.64 (0.58, 0.70) 0.82 (0.75, 0.91)

Alcohol
Ex-drinker 16,257 261 1.00 (ref) 1.00 (ref)
Above guideline 169,542 1240 0.39 (0.34, 0.45) 0.54 (0.47, 0.61)
Never drinker 19,522 156 0.53 (0.43, 0.65) 0.75 (0.61, 0.92)
Within guideline 261,842 1513 0.38 (0.34, 0.44) 0.55 (0.48, 0.63)

Smoking
Current 48,581 727 1.00 (ref) 1.00 (ref)
Previous 162,990 1393 0.39 (0.36, 0.43) 0.46 (0.42, 0.51)
Never 256,872 1050 0.25 (0.23, 0.28) 0.33 (0.30, 0.36)

3.3. Healthy lifestyle index disease and pneumonia mortality. For both infectious disease and
pneumonia, there was a dose–response association with higher lifestyle
3.3.1. Infectious disease and pneumonia mortality index scores. For example, there was a 34% (HR [95%CIs]: 0.66 [0.59 to
Fig. 1 shows the healthy lifestyle index hazard ratios for infectious 0.75]) to 71% (0.29 [0.26 to 0.33]) reduction in infectious disease

22
M.N. Ahmadi et al. Brain Behavior and Immunity 96 (2021) 18–27

Table 5
Lifestyle risk factors and COVID-19 mortality risk ratio.
Risk factor N Events Model 1 Model 2
RR (95% CI) RR (95% CI)

Physical
Activity Inactive 95,221 112 1.00 (ref) 1.00 (ref)
Insufficient 140,609 115 0.75 (0.58, 0.97) 0.87 (0.67, 1.14)
Sufficient 232,613 160 0.57 (0.44, 0.72) 0.70 (0.54, 0.89)

Sedentary
Behaviour High 57,485 68 1.00 (ref) 1.00 (ref)
Moderate 249,929 217 0.72 (0.55, 0.95) 0.90 (0.68, 1.90)
Low 161,029 102 0.65 (0.48, 0.89) 0.87 (0.64, 1.20)

Sleep
Poor 14,520 17 1.00 (ref) 1.00 (ref)
Moderate 201,135 181 0.75 (0.46, 1.24) 0.96 (0.58, 1.58)
Good 252,788 189 0.66 (0.40, 1.08) 0.97 (0.59, 1.61)

Diet
Poor 75,264 62 1.00 (ref) 1.00 (ref)
Moderate 186,893 140 0.83 (0.61, 1.12) 0.92 (0.68, 1.25)
Good 206,286 185 0.85 (0.64, 1.14) 1.03 (0.77, 1.39)

Alcohol
Ex-drinker 16,257 29 1.00 (ref) 1.00 (ref)
Above guideline 169,542 150 0.46 (0.31, 0.69) 0.62 (0.41, 0.93)
Never drinker 19,522 25 0.79 (0.46, 1.35) 0.87 (0.50, 1.50)
Within guideline 261,842 183 0.44 (0.30, 0.69) 0.60 (0.40, 0.89)

Smoking
Current 48,581 59 1.00 (ref) 1.00 (ref)
Previous 162,990 183 0.66 (0.49, 0.89) 0.75 (0.55, 1.02)
Never 256,872 145 0.45 (0.33, 0.61) 0.54 (0.39, 0.74)

Fig. 1. Healthy lifestyle index hazard ratio for infectious diseases and pneumonia mortality. Models are adjusted for age, sex, socioeconomic status, ethnicity, BMI,
cardiovascular disease, cancer, diabetes, hypertension, use of anti-hypertensive medication, use of corticosteroids, chronic lung/respiratory disease, liver diseases,
diabetes, end-stage renal disease, and immune disorders/HIV. The original combined lifestyle behaviour scores ranged from 0 to 12. This score has been re-classified
as follows: scores 0 to 4 = least Healthy group; score of 5 = 6th Healthiest group; score of 6 = 5th Healthiest group; score of 7 = 4th Healthiest group; score of 8 = 3rd
Healthiest group; score of 9 = 2nd Healthiest group; scores 10 to 12 = Healthiest group.

mortality for participants who were not classified in the least healthy Additional analysis for infectious disease and pneumonia among only
behaviour group. Similarly, the pneumonia mortality risk was gradually participants with cancer or cardiovascular disease showed a dos­
attenuated with a higher lifestyle index; e.g. a 31% (0.69 [0.60 to 0.79]) e–response association with higher lifestyle index scores (Supplemental
to 74% (0.26 [0.23 to 0.30]) lower pneumonia mortality risk for par­ Figs. 2 and 3). For infectious disease, participants with cancer had a 28%
ticipants when compared to those in the least healthy behaviour group. (0.72 [0.60 to 0.86]) to 65% (0.35 [0.29 to 0.42]) reduction in mortality

23
M.N. Ahmadi et al. Brain Behavior and Immunity 96 (2021) 18–27

risk, whilst participants with cancer had a 30% (0.72 [0.61 to 0.79]) to 3.3.4. Sensitivity analysis
68% (0.32 [0.28 to 0.37]) reduction compared to participants classified Removing participants with an event occurring in the first five years
in the least healthy behaviour group. Likewise, the pneumonia mortality of follow-up, a history of smoking, cardiovascular disease, or cancer had
risk among was gradually attenuated with a higher lifestyle index; no material impact on the dose–response associations with infectious
participants with cancer had a 27% (0.73 [0.60 to 0.89]) to 69% (0.31 disease mortality (Supplemental Tables 14 and 15, and Supplemental
[0.25 to 0.38]) reduction in mortality risk, and participants with car­ Figures 4 and 5). The associations of individual lifestyle risk factors with
diovascular disease had a 29% (0.71 [0.61 to 0.82]) to 70% (0.30 [0.25 infectious disease mortality were not appreciably different when par­
to 0.34]) reduction. ticipants who had the least healthy lifestyle behaviour index score were
analysed separately (Supplemental Tables 16 and 17). Three of the in­
3.3.2. COVID-19 mortality dividual lifestyle risk factors showed beneficial associations against in­
Fig. 2 displays the healthy lifestyle index risk ratios for COVID-19 fectious disease mortality when pneumonia events were excluded:
mortality. Across the lifestyle groupings, we observed a similar reduc­ engaging in at least some physical activity; not being a current smoker;
tion for COVID-19 mortality risk as in infectious disease and pneumonia and consuming at least some alcohol (Supplemental Table 18).
mortality above. Among the 4th healthiest to healthiest lifestyle index,
COVID-19 mortality risk was attenuated by 44% (RR [95% CIs]: 0.56 4. Discussion
[0.38 to 0.82]) to 58% (0.42 [0.28 to 0.63]) for individuals who were
not classified in the least healthy behaviour group. In this prospective cohort study, we examined the additive rela­
tionship between multiple lifestyle risk factors - physical activity,
3.3.3. Population impact sedentary behaviour, sleeping quality, diet quality, alcohol consump­
Supplemental Tables 2 to 11 and Supplemental Fig. 2 to 3 display tion, and smoking. We found a clear beneficial dose response association
results stratified by sex, age, body mass index, socioeconomic status, and with a healthier lifestyle index score against mortality from infectious
participants diagnosed with cardiovascular disease or cancer. There disease, pneumonia, and COVID-19. These associations were indepen­
were generally consistent dose–response patterns with higher lifestyle dent of multiple markers of overall health status. We found limited ev­
indexes across all strata, including participants in the highest mortality idence of synergistic interactions between pairs of behaviours,
risk groups. For example, participants in the lowest socioeconomic sta­ suggesting that any beneficial associations conferred by different life­
tus quintile had an infectious disease mortality risk between 0.74 [0.59 style behaviours is not greater than the sum of the risk induced by each
to 0.92] to 0.31 [0.24 to 0.40]. Mortality risk among participants who behaviour. This interpretation is supported by the results of the indi­
were obese or over 60 years, and not classified in the lowest lifestyle vidual risk factors and outcomes. Results for COVID-19 mortality were
index category was markedly low; among these participants, hazard consistent, although the low number of events made the statistical
ratios were between 0.70 [0.57 to 0.86] to 0.31 [0.20 to 0.47] for in­ comparisons less clear. The patterns of attenuation, however, were
fectious disease mortality. Likewise, participants diagnosed with car­ comparable to infectious disease and pneumonia mortality. Our results
diovascular disease or cancer had an incremental decrease for mortality are encouraging, not least for middle-aged and older adults who are at
risk as the healthy lifestyle index improved with hazard ratios between the highest risk of mortality from respiratory infections, who can
0.72 [ 0.60 to 0.85] to 0.32 [0.28 to 0.37]. The only pair of lifestyle potentially gain protection against the consequences of infectious dis­
behaviours that showed a statistically significant synergistic interaction ease through modifiable lifestyle behaviours.
(Supplemental Table 12) was not meeting physical activity guidelines We observed a dose–response for infectious disease mortality with
and being a current smoker (RERI [95% CI] = 0.4 [0.06–0.8]; S = 1.3 higher lifestyle index scores. Infectious disease mortality in a smaller
[1.1–1.5], attributable portion due to interaction = 14.0% (2.8%- analysis of the Health Survey for England and Scottish Health Survey
25.2%)]. The lack of significant synergistic interactions among most examining traditional lifestyle behaviours- that included physical ac­
lifestyle behaviour pairs suggests that the dose–response relationship tivity, smoking, and alcohol consumption- reported protective associa­
among the different lifestyle behaviours is not greater than the sum of tions against mortality among 97,844 participants if they engaged in at
the risk induced by each behaviour. least some moderate to vigorous physical activity, and had never
smoked (Hamer et al., 2019). The study did not examine the additive
effects of lifestyle risk factors that led to a decrease in infectious disease

Fig. 2. Healthy lifestyle index risk ratio for COVID-

19 mortality. Models are adjusted for age, sex, so­
cioeconomic status, ethnicity, BMI, cardiovascular
disease, cancer, diabetes, hypertension, use of anti-
hypertensive medication, use of corticosteroids,
chronic lung/respiratory disease, liver diseases,
diabetes, end-stage renal disease, and immune dis­
orders/HIV. The original combined lifestyle behav­
iour scores ranged from 0 to 12. This score has been
re-classified as follows: scores 0 to 4 = least Healthy
group; score of 5 = 6th Healthiest group; score of 6
= 5th Healthiest group; score of 7 = 4th Healthiest
group; score of 8 = 3rd Healthiest group; score of 9
= 2nd Healthiest group; scores 10 to 12 = Healthiest
group.

24
M.N. Ahmadi et al. Brain Behavior and Immunity 96 (2021) 18–27

mortality risk. Analysis of 64,027 HUNT Study participants showed that risk factors were measured with self-report questionnaires. Due to social
bloodstream-specific infectious disease mortality was associated with desirability bias, misclassification is potentially non-random, and the
individual health behaviours, specifically moderate to vigorous physical results are most likely biased toward the null, with participants more
activity levels and smoking status (Paulsen et al., 2017). Other epide­ likely to report desirable behaviours. Therefore, the preventable infec­
miological studies have assessed other traditional individual behaviours tious disease mortality related to the healthy lifestyle indices is likely to
with infectious disease using various lifestyle behaviour measures be underestimated, as indicated by PF. Second, the sleep quality scoring
(Almirall et al., 2008; Inoue et al., 2007; Wang et al., 2014). The current included sleep chronotype, which might be influenced more by genetic
study is the first to examine the protective benefits for a combined traits than behavioural factors (Adan et al., 2012; Hur et al., 1998;
healthy lifestyle and among individuals with comorbidities, who are Koskenvuo et al., 2007). Third, although the UK Biobank cohort is not
most at risk of infectious disease mortality. The health benefits were representative of the general population (UK Biobank participants are
found to be additive and can be attained through a combination of healthier than the general population), prior epidemiological evidence
lifestyle behaviours. The dose–response nature of the associations be­ has shown that there is little evidence for bias attributable to nonpar­
tween healthy lifestyle indexes was consistent across infectious disease, ticipation and exposure-disease relationships are widely generalizable
pneumonia, and COVID-19 mortality. (Fry et al., 2017). This reinforces the epidemiological principle that
We found consistent beneficial associations for all six individual associations are less dependent on the representativeness of the cohort,
lifestyle behaviour categories with infectious disease and pneumonia relative to prevalence (Galea et al., 2007).
mortality. With only one exception, however, there was no evidence of
synergistic interactions between pairs of behaviours. Specifically, 4.1. Conclusions
meeting physical activity guidelines and not being a current smoker
were the only lifestyle behaviours to have a synergistic interaction This large prospective cohort study examined the additive impact of
against the risk of infectious disease mortality. Habitual moderate to healthy lifestyle behaviour combinations, which included the analysis of
vigorous physical activity enhances a number of immune parameters traditional and emerging lifestyle factors. We found that in middle aged
such as increasing natural killer cell activity, neutrophils, number of and older adults, including those with cardiovascular disease and can­
circulating lymphocytes, and cytokine production (Mackinnon, 1999; cer, healthier lifestyle behaviours may protect against the most severe
Matthews et al., 2002; Nieman, 1994; Nieman et al., 1990). Conversely, consequences of infectious disease. The findings based on public health
smoking affects many of the same immune-parameters but in the guidelines and best practice recommendations provides information
opposite direction (Hersey et al., 1983; Sopori, 2002). that clinicians and researchers can readily translate into practice and
Meeting health guidelines or best practice recommendations in future research.
combinations of different lifestyle behaviours can significantly reduce
the risk of infectious disease mortality among both the low and high-risk Declaration of Competing Interest
segments of the population, regardless of sex, age, weight, or socio­
economic status. In addition to preventive immunology measures, The authors declare that they have no known competing financial
public health efforts focused on improvements in meeting minimum interests or personal relationships that could have appeared to influence
lifestyle recommendations could be used as an ancillary measure to the work reported in this paper.
ameliorate the most severe health consequences of infectious disease,
especially among middle aged and older adults. Participants with
Acknowledgement
existing chronic conditions such as cardiovascular disease and cancer—
for whom our study has also shown to gain health benefits—might
We are grateful to the UK Biobank participants. This research has
choose to engage in a number of differing healthy lifestyle behaviours
been conducted using the UK Biobank Resource.
and can still attain protective benefits against infectious disease, pneu­
monia, and COVID-19 mortality. These findings offer additional re­
References
sources for primary care to prescribe improvements in lifestyle risk
factors that can be used as a powerful ancillary measure against mor­ Adan, A., Archer, S.N., Hidalgo, M.P., Di Milia, L., Natale, V., Randler, C., 2012.
tality from infectious disease. Circadian typology: A comprehensive review. Chronobiol. Int. 29, 1153–1175.
To our knowledge, this is the first study to examine a comprehensive https://doi.org/10.3109/07420528.2012.719971.
Allen, N., Sudlow, C., Downey, P., Peakman, T., Danesh, J., Elliott, P., Gallacher, J.,
lifestyle risk factor index score incorporating multiple modifiable be­ Green, J., Matthews, P., Pell, J., Sprosen, T., Collins, R., 2012. UK Biobank: Current
haviours (physical activity, sedentary behaviour, sleep quality, diet status and what it means for epidemiology. Heal. Policy Technol. 10.1016/j.
quality, alcohol consumption, and smoking status) in relation to infec­ hlpt.2012.07.003.
Almirall, J., Bolíbar, I., Serra-Prat, M., Roig, J., Hospital, I., Carandell, E., Agustí, M.,
tious disease mortality risk. We were able to provide a comprehensive Ayuso, P., Estela, A., Torres, A., De Salas, J.M., Costa, J., Tristany, M., Grau, M.J.,
assessment for sleep quality that accounted for five sleep characteristics. Sancho, S., Miguel, E., Fradera, M., Ochoa, I., Castany, M.J., Quilez, A., Marina, V.,
We were, also, able to separate never drinkers from ex-drinkers who may Subias, P., Jimeno, B., Bradnovich, A., Rodriguez, M., Ramon, E., Gardella, A.,
Ginés, C., Mestres, X., Armada, A., Mallafré, J., Roger, M., Gros, M.T., Les, N.,
have quit drinking due to prior alcohol-related problems. The dietary Joanola, J., Doménech, J., Bundó, M., Trilla, M., Massons, J., Montero, J., Zurilla, E.,
measure was comprehensive and included fruits, vegetables, grains, and Torán, P., Aizpurua, M.M., Lozano, G., Casals, J., Sorribes, J., Torrellas, D.,
red/processed meat. We also did not conflate the lifestyle behaviours Anglada, J.L.L., Salabarnada, J., Sanz, E., Gorgas, F., Ribas, A., Fau, E., Pellicer, I.,
Morales, S., Casas, A., Bernad, J., De Montoliu, A., Gaya, J., Vallés, R., Vazquez, A.,
with their outcomes, as some lifestyle behaviour indices have previously
Peiró, R., Aresté, G., Mengual, G.N., Viñes, M.C., Almerich, E., Lopez, M.A., Bel, J.,
done by including weight status or other metabolic health indicators in Gosalves, A., Macip, S., Carrillo, E., Paulo, P., Pol, M., Sala, J., Mir, P., Carrera, P.,
the index (Bonaccio et al., 2019; Lee et al., 2011). We examined modi­ Legazpi, I., Planellas, F., Beltran, A., Planas, T., Rovira, J., De Ciurana, M.,
Xargay, C., Cortés, R.B., Paredes, J., Olive, J.G., Montoya, P., Burcet, F.X.,
fiable lifestyle behaviours in a large cohort with more than 10 years
Alcantarilla, D., De Ribot, J., Gorgot, M.R., Bas, R.A., Rosa, M.B., Castells, J.,
follow-up for mortality, and the longest person-years follow-up in the Castillo, J.A., Torrent, J.C., Rodríguez, D., Rius, X., Navarro, M., Benet, J.M.,
field, and quantified the population health impact from different life­ Caballero, J.C., Llach, J., Sanglas, J., Morales, V., Solanella, J., Cardona, J.I.,
style behaviour combinations and synergistic interactions. The use of Ferrer, J., Solís, M.J., Güerri, F., Murillo, M.J., Llovet, M.D., Guinea, I.,
Juáreztarrag, M.M., Alvarez, M., Bladé, J., Sánchez Oró, I., Martin, F., Cabré, J.J.,
lifestyle behaviour indices such as ours based on current guidelines and Frigola, J., Boj, J., Chacon, A., Satué, E., Gallego, F., Montanyés, D., Basora, T.,
best practice category thresholds for risk allows for policy-relevant Martin, E., Gil, P., Roselló, M., Marimon, J., Costa, B., Gutiérrez, C., Moltó, E.,
lifestyle behaviours to be easily translated and assessed across settings Isach, A., Izzeddin, N., Campani, M., Ferré, J., Fernández, A., Jové, J., Vilalta, J.,
Ribera, E., Fernández, M., Ochoa, O., Vila Palacios, A., Donado-Mazarrón, A.,
and populations. Ramon, C., Caballol, R., Navarro, L., Masqué, J., Llor, L.L., Garcia, G., Arasa, M.J.,
Opposing these strengths were several limitations. First, all lifestyle Yoldi, A., Viloria, I., Vallespi, F., Clua, J.L.L., Basart, J.M., Ciurana, E., Monclus, J.F.,

25
M.N. Ahmadi et al. Brain Behavior and Immunity 96 (2021) 18–27

Casanovas, A., Viñas, L., Checa, E., Marin, J., Santigosa, J., Arayo, F., Campillos, C., Dunstan, D.W., Howard, B., Healy, G.N., Owen, N., 2012. Too much sitting - A health
Juarez, A., Perez, J., Paredes, E., Rodriguez, J., Solà, Y., Pujol, J., Navarro, M.A., hazard. Diabetes Res. Clin. Pract. 97, 368–376. https://doi.org/10.1016/j.
Casado, I., Madrid, M., Berdié, J., Andrés, M.J., Barragan, B., Torras, A., Gonzalez, S., diabres.2012.05.020.
Fuentes, A., El Hosn-Makarem, J.A., Pelaez, F., Ubieto, A., Fallada, A., Josa, A., Fan, M., Sun, D., Zhou, T., Heianza, Y., Lv, J., Li, L., Qi, L., 2020. Sleep patterns, genetic
Favà, E., Ros, M., Alonso, M., Pinazo, M.J., Borrás, S., Nolla, C., Bitrià, J., Araujo, O., susceptibility, and incident cardiovascular disease: A prospective study of 385 292
Collado, A., Glaria, T., Montero, J.C., Flores, P., Serra, P., Torrellas, E., Buxadé, I., UK biobank participants. Eur. Heart J. 41, 1182–1189. https://doi.org/10.1093/
Mussoll, J., Gomez, M., Serrano, C., Pubill, M., Gost, J., Burdoy, E., Busquets, L.L., De eurheartj/ehz849.
Castro, S., Bartolomé, M., Corona, E., Valverde, R., Verde, Y., Alegre, M., Papiol, M., Ferkol, T., Schraufnagel, D., 2014. The global burden of respiratory disease. Ann. Am.
Martí, O., Catalá, M., Martinez, M.A., Diaz, E., Borrás, A., Aznar, F., Riera, F., Thorac. Soc. 11, 404–406. https://doi.org/10.1513/AnnalsATS.201311-405PS.
Colom, I., Calvet, E., Nicolás, J., Ruiz, J., Mateu, C., Arranz, J., Fé, A., Puig, B., Fletcher, G.F., Landolfo, C., Niebauer, J., Ozemek, C., Arena, R., Lavie, C.J., 2018.
Gutiérrez, M., Morant, L.L., Llull, M., Estelrich, J., Quintana, L., Llabrés, C., Promoting Physical Activity and Exercise: JACC Health Promotion Series. J. Am.
Sureda, M.M., Cuevas, L., Muñoz, Y., Llobera, J., Iñiguez, M., Guasch, J., Sánchez, F., Coll. Cardiol. 72, 1622–1639. https://doi.org/10.1016/j.jacc.2018.08.2141.
Pascual, J., Castelló, A., Gómez, L.L., Marin, P., Garcia, C., Moreno, P., Coll, M., Friedman, G.D., Klatsky, A.L., 1993. Is alcohol good for your health? N. Engl. J. Med.
Palma, C., Ribera, C., Aldosa, J., Alonso, C., Bergés, A., Burgués, J., Burgués, L.L., 1882–1883.
Busquets, J., Casal, D., Cerdán, N., Crespo, X., Duat, R., Escoda, J., Fiter, M., Font, A., Fry, A., Littlejohns, T.J., Sudlow, C., Doherty, N., Adamska, L., Sprosen, T., Collins, R.,
Fraysse, M., Fuentes, S., García-Núñez, R., López-Pinacho, T., Masardo, M., Allen, N.E., 2017. Comparison of Sociodemographic and Health-Related
Martínez-Illescas, J., Morisset, P., Pallarés, M., Pérez-Serra, R., Plà, A., Pons, E., Characteristics of UK Biobank Participants With Those of the General Population.
Ruiz, A., Rodríguez-Picart, J.C., Sánchez-Claver, J., Sylvestre, E., Tolosa, R., Am. J. Epidemiol. 186, 1026–1034. https://doi.org/10.1093/aje/kwx246.
Touceda, J.A., Varela, J.R., Vidal, J.C., 2008. New evidence of risk factors for Galea, S., Ph, D.R., Tracy, M., 2007. Participation Rates in Epidemiologic Studies. Ann.
community-acquired pneumonia: A population-based study. Eur. Respir. J. 31, Epidemiol. 17, 643–653. https://doi.org/10.1016/j.annepidem.2007.03.013.
1274–1284. https://doi.org/10.1183/09031936.00095807. Gelband, H., Miller-Petrie, M., Pant, S., Gandra, S., Levinson, J., Barter, D., White, A.,
Ambrus, J.L., Ambrus, J.L., 2004. Nutrition and infectious diseases in developing Laxminarayan, R., 2015. The state of the world’s antibiotics 2015. Wound Heal.
countries and problems of acquired immunodeficiency syndrome. Exp. Biol. Med. South. Africa 8, 30–34.
229, 464–472. https://doi.org/10.1177/153537020422900603. Gordon, J.E., 1968. Weanling diarrhoea—a synergism of infection and nutrition, in:
Andersson, T., Alfredsson, L., Källberg, H., Zdravkovic, S., Ahlbom, A., 2005. Calculating Interactions of Malnutrition and Infection. p. 216.
measures of biological interaction. Eur. J. Epidemiol. 20, 575–579. https://doi.org/ Hamer, M., Kivimäki, M., Gale, C.R., Batty, G.D., 2020. Lifestyle risk factors,
10.1007/s10654-005-7835-x. inflammatory mechanisms, and COVID-19 hospitalization: A community-based
Baik, I., Curhan, G.C., Rimm, E.B., Bendich, A., Willett, W.C., Fawzi, W.W., 2000. cohort study of 387,109 adults in UK. Brain. Behav. Immun. 87, 184–187. https://
A prospective study of age and lifestyle factors in relation to community-acquired doi.org/10.1016/j.bbi.2020.05.059.
pneumonia in US men and women. Arch. Intern. Med. 160, 3082–3088. https://doi. Hamer, M., O’Donovan, G., Stamatakis, E., 2019. Lifestyle risk factors, obesity and
org/10.1001/archinte.160.20.3082. infectious disease mortality in the general population: Linkage study of 97,844
Bonaccio, M., Di Castelnuovo, A., Costanzo, S., De Curtis, A., Persichillo, M., Cerletti, C., adults from England and Scotland. Prev. Med. (Baltim) 123, 65–70. https://doi.org/
Donati, M.B., de Gaetano, G., Iacoviello, L., Investigators, M.S., 2019. Impact of 10.1016/j.ypmed.2019.03.002.
combined healthy lifestyle factors on survival in an adult general population and in Hamer, M., Stamatakis, E., Steptoe, A., 2014. Effects of substituting sedentary time with
high-risk groups: prospective results from the Moli-sani study. J. Intern. Med. 286, physical activity on metabolic risk. Med. Sci. Sports Exerc. 46, 1946–1950. https://
207–220. https://doi.org/10.1111/joim.12637. doi.org/10.1249/MSS.0000000000000317.
Bull, F., Saad Al-Ansari, S., Biddle, S., Borodulin, K., Buman, M., Cardon, G., Carty, C., Hamilton, M., Hamilton, D., Zderic, T., 2007. Role of low energy expenditure and sitting
Chaput, J.-P., Chastin, S., Chou, R., Dempsey, P., DiPietro, L., Ekelund, U., Firth, J., in obesity, metabolic syndrome, type 2 diabetes, and cardiovascular disease.
Friedenreich, C., Garcia, L., Gichu, M., Jago, R., Katzmarzyk, P., Lambert, E., Diabetes 56, 2655–2667. https://doi.org/10.2337/db07-0882.CVD.
Leitzmann, M., Milton, K., Ortega, F., Ranasinghe, C., Stamatakis, E., Willumsen, J., Alcohol guidelines review–Report from the Guidelines development group to the UK
2020. World Health Organization 2020 Guidelines on Physical Activity and Chief Medical Officers.
Sedentary Behaviour. Br. J. Sports Med. 1451–1462 https://doi.org/10.1136/ Hersey, P., Prendergast, D., Edwards, A., 1983. Effects of cigarette smoking on the
bjsports-2020-102955. immune system Follow-up studies in normal subjects after cessation of smoking.
Buuren, S.van, Groothuis-Oudshoorn, K., 2010. mice: Multivariate imputation by Med. J. Aust. 2, 425–429. https://doi.org/10.5694/j.1326-5377.1983.tb122565.x.
chained equations in R. J. Stat. Softw. 1–68 https://doi.org/10.1080/ Hur, Y.-M., Bourchar Jr, T., Lykken, D., 1998. Genetic and environmental influence on
10580530.2013.739883. morningness-eveningness. Pers. Individ. Dif. 25, 917–925.
Carter, B.D., Abnet, C.C., Feskanich, D., Freedman, N.D., Hartge, P., Lewis, C.E., Huttunen, R., Heikkinen, T., Syrjänen, J., 2011. Smoking and the outcome of infection.
Ockene, J.K., Prentice, R.L., Speizer, F.E., Thun, M.J., Jacobs, E.J., 2015. Smoking J. Intern. Med. 269, 258–269. https://doi.org/10.1111/j.1365-2796.2010.02332.x.
and Mortality — Beyond Established Causes. N. Engl. J. Med. 372, 631–640. https:// Ibarra-Coronado, E.G., Pantaleón-Martínez, A.M., Velazquéz-Moctezuma, J., Prospéro-
doi.org/10.1056/nejmsa1407211. García, O., Méndez-Díaz, M., Pérez-Tapia, M., Pavón, L., Morales-Montor, J., 2015.
Cassini, A., Högberg, L.D., Plachouras, D., Quattrocchi, A., Hoxha, A., Simonsen, G.S., The Bidirectional Relationship between Sleep and Immunity against Infections.
Colomb-Cotinat, M., Kretzschmar, M.E., Devleesschauwer, B., Cecchini, M., J. Immunol. Res. 2015 https://doi.org/10.1155/2015/678164.
Ouakrim, D.A., Oliveira, T.C., Struelens, M.J., Suetens, C., Monnet, D.L., Strauss, R., Inoue, Y., Koizumi, A., Wada, Y., Iso, H., Watanabe, Y., Date, C., Yamamoto, A.,
Mertens, K., Struyf, T., Catry, B., Latour, K., Ivanov, I.N., Dobreva, E.G., Tambic Kikuchi, S., Inaba, Y., Toyoshima, H., Tamakoshi, A., 2007. Risk and protective
Andraševic, A., Soprek, S., Budimir, A., Paphitou, N., Žemlicková, H., Schytte factors related to mortality from pneumonia among middle-aged and elderly
Olsen, S., Wolff Sönksen, U., Märtin, P., Ivanova, M., Lyytikäinen, O., Jalava, J., community residents: The JACC study. J. Epidemiol. 17, 194–202. https://doi.org/
Coignard, B., Eckmanns, T., Abu Sin, M., Haller, S., Daikos, G.L., Gikas, A., 10.2188/jea.17.194.
Tsiodras, S., Kontopidou, F., Tóth, Á., Hajdu, Á., Guólaugsson, Ó., Kristinsson, K.G., Irwin, M.R., 2015. Why sleep is important for health: A psychoneuroimmunology
Murchan, S., Burns, K., Pezzotti, P., Gagliotti, C., Dumpis, U., Liuimiene, A., perspective. Annu. Rev. Psychol. 66, 143–172. https://doi.org/10.1146/annurev-
Perrin, M., Borg, M.A., de Greeff, S.C., Monen, J.C., Koek, M.B., Elstrøm, P., psych-010213-115205.
Zabicka, D., Deptula, A., Hryniewicz, W., Caniça, M., Nogueira, P.J., Fernandes, P.A., Kadri, S.S., 2020. Key Takeaways from the U.S. CDC’s 2019 Antibiotic Resistance Threats
Manageiro, V., Popescu, G.A., Serban, R.I., Schréterová, E., Litvová, S., Report for Frontline Providers. Crit. Care Med. 939–945 https://doi.org/10.1097/
Štefkovicová, M., Kolman, J., Klavs, I., Korošec, A., Aracil, B., Asensio, A., Pérez- CCM.0000000000004371.
Vázquez, M., Billström, H., Larsson, S., Reilly, J.S., Johnson, A., Hopkins, S., 2019. Källberg, H., Ahlbom, A., Alfredsson, L., 2006. Calculating measures of biological
Attributable deaths and disability-adjusted life-years caused by infections with interaction using R. Eur. J. Epidemiol. 21, 571–573. https://doi.org/10.1007/
antibiotic-resistant bacteria in the EU and the European Economic Area in 2015: a s10654-006-9037-6.
population-level modelling analysis. Lancet Infect. Dis. 19, 56–66. https://doi.org/ Katona, P., Katona-Apte, J., 2008. The interaction between nutrition and infection. Clin.
10.1016/S1473-3099(18)30605-4. Infect. Dis. 46, 1582–1588. https://doi.org/10.1086/587658.
Chau, J.Y., Grunseit, A., Midthjell, K., Holmen, J., Holmen, T.L., Bauman, A.E., Van Der Knott, C.S., Coombs, N., Stamatakis, E., Biddulph, J.P., 2015. All cause mortality and the
Ploeg, H.P., 2015. Sedentary behaviour and risk of mortality from all-causes and case for age specific alcohol consumption guidelines: Pooled analyses of up to 10
cardiometabolic diseases in adults: Evidence from the HUNT3 population cohort. Br. population based cohorts. BMJ 350. https://doi.org/10.1136/bmj.h384.
J. Sports Med. 49, 737–742. https://doi.org/10.1136/bjsports-2012-091974. Koskenvuo, M., Hublin, C., Partinen, M., Heikkilä, K., Kaprio, J., 2007. Heritability of
Chau, J.Y., Grunseit, A.C., Chey, T., Stamatakis, E., Brown, W.J., Matthews, C.E., diurnal type: A nationwide study of 8753 adult twin pairs. J. Sleep Res. 16, 156–162.
Bauman, A.E., Van Der Ploeg, H.P., 2013. Daily sitting time and all-cause mortality: https://doi.org/10.1111/j.1365-2869.2007.00580.x.
A meta-analysis. PLoS One 8, 1–14. https://doi.org/10.1371/journal.pone.0080000. Lee, C., Sui, X., Hooker, S.P., Ebert, J.R.H., Blair, S.N., 2011. Combined Impact of
Chiuve, S.E., Fung, T.T., Rimm, E.B., Hu, F.B., McCullough, M.L., Wang, M., Stampfer, M. Lifestyle Factors on Cancer Mortality in Men. https://doi.org/10.1016/j.
J., Willett, W.C., 2012. Alternative dietary indices both strongly predict risk of annepidem.2011.04.010.
chronic disease. J. Nutr. 142, 1009–1018. https://doi.org/10.3945/jn.111.157222. Mackinnon, L.T., 1999. Advances in exercise immunology. Human Kinetics.
Craig, C.L., Marshall, A.L., Sjöström, M., Bauman, A.E., Booth, M.L., Ainsworth, B.E., Marston, H.D., Dixon, D.M., Knisely, J.M., Palmore, T.N., Fauci, A.S., 2016.
Pratt, M., Ekelund, U., Yngve, A., Sallis, J.F., Oja, P., 2003. International physical Antimicrobial resistance. JAMA - J. Am. Med. Assoc. 316, 1193–1204. https://doi.
activity questionnaire: 12-Country reliability and validity. Med. Sci. Sports Exerc. org/10.1001/jama.2016.11764.
35, 1381–1395. https://doi.org/10.1249/01.MSS.0000078924.61453.FB. Matthews, C.E., Ockene, I.R.A.S., Freedson, P.S., Rosal, M.C., Merriam, P.A., Hebert, J.R.,
Ding, D., Rogers, K., van der Ploeg, H., Stamatakis, E., Bauman, A.E., 2015. Traditional 2002. Moderate to vigorous physical activity and risk of upper-respiratory tract
and Emerging Lifestyle Risk Behaviors and All-Cause Mortality in Middle-Aged and infection. Med. Sci. Sport. Exerc. 34, 1242–1248.
Older Adults: Evidence from a Large Population-Based Australian Cohort. PLoS Med. McCullough, M.L., Feskanich, D., Stampfer, M.J., Giovannucci, E.L., Rimm, E.B., Hu, F.
12, 1–21. https://doi.org/10.1371/journal.pmed.1001917. B., Spiegelman, D., Hunter, D.J., Colditz, G.A., Willett, W.C., 2002. Diet quality and

26
M.N. Ahmadi et al. Brain Behavior and Immunity 96 (2021) 18–27

major chronic disease risk in men and women: Moving toward improved dietary Sallis, J.F., Adlakha, D., Oyeyemi, A., Salvo, D., 2020. An international physical activity
guidance. Am. J. Clin. Nutr. 76, 1261–1271. https://doi.org/10.1093/ajcn/ and public health research agenda to inform coronavirus disease-2019 policies and
76.6.1261. practices. J. Sport Heal. Sci. 9, 328–334. https://doi.org/10.1016/j.
Nair, H., Brooks, W.A., Katz, M., Roca, A., Berkley, J.A., Madhi, S.A., Simmerman, J.M., jshs.2020.05.005.
Gordon, A., Sato, M., Howie, S., Krishnan, A., Ope, M., Lindblade, K.A., Carosone- Samokhvalov, A.V., Irving, H.M., Rehm, J., 2010. Alcohol consumption as a risk factor
Link, P., Lucero, M., Ochieng, W., Kamimoto, L., Dueger, E., Bhat, N., Vong, S., for pneumonia: A systematic review and meta-analysis. Epidemiol. Infect. 138,
Theodoratou, E., Chittaganpitch, M., Chimah, O., Balmaseda, A., Buchy, P., 1789–1795. https://doi.org/10.1017/S0950268810000774.
Harris, E., Evans, V., Katayose, M., Gaur, B., O’Callaghan-Gordo, C., Goswami, D., Scrimshaw, N.S., SanGiovanni, J.P., 1997. Synergism of nutrition, infection, and
Arvelo, W., Venter, M., Briese, T., Tokarz, R., Widdowson, M.A., Mounts, A.W., immunity: An overview. Am. J. Clin. Nutr. 66 https://doi.org/10.1093/ajcn/
Breiman, R.F., Feikin, D.R., Klugman, K.P., Olsen, S.J., Gessner, B.D., Wright, P.F., 66.2.464S.
Rudan, I., Broor, S., Simões, E.A., Campbell, H., 2011. Global burden of respiratory Song, J.H., Thamlikitkul, V., Hsueh, P.R., 2011. Clinical and economic burden of
infections due to seasonal influenza in young children: A systematic review and community-acquired pneumonia amongst adults in the Asia-Pacific region. Int. J.
meta-analysis. Lancet 378, 1917–1930. https://doi.org/10.1016/S0140-6736(11) Antimicrob. Agents 38, 108–117. https://doi.org/10.1016/j.
61051-9. ijantimicag.2011.02.017.
Nieman, D.C., 1994. Exercise, Infection, and Immunity. Int. J. Sports Med. 15, Sopori, M., 2002. Effects of cigarette smoke on the immune system. Nat. Rev. Immunol.
S131–S141. https://doi.org/10.1055/s-2007-1021128. 2, 372–377.
Nieman, D.C., Henson, D.A., Austin, M.D., Sha, W., 2011. Upper respiratory tract Stamatakis, E., Rogers, K., Ding, D., Berrigan, D., Chau, J., Hamer, M., Bauman, A., 2015.
infection is reduced in physically fit and active adults. Br. J. Sports Med. 45, All-cause mortality effects of replacing sedentary time with physical activity and
987–992. https://doi.org/10.1136/bjsm.2010.077875. sleeping using an isotemporal substitution model: A prospective study of 201,129
Nieman, D.C., Nehlsen-Cannarella, S.L., Markoff, P.A., Balk-Lamberton, A.J., Yang, H., mid-aged and older adults. Int. J. Behav. Nutr. Phys. Act. 12 https://doi.org/
Chritton, D.B.W., Lee, J.W., Arabatzis, K., 1990. The effects of moderate exercise 10.1186/s12966-015-0280-7.
training on natural killer cells and acute upper respiratory tract infections. Int. J. Sudlow, C., Gallacher, J., Allen, N., Beral, V., Burton, P., Danesh, J., Downey, P.,
Sports Med. 11, 467–473. https://doi.org/10.1055/s-2007-1024839. Elliott, P., Green, J., Landray, M., Liu, B., Matthews, P., Ong, G., Pell, J., Silman, A.,
Opp, M.R., Krueger, J.M., 2015. Sleep and immunity: A growing field with clinical Young, A., Sprosen, T., Peakman, T., Collins, R., 2015. UK Biobank: An Open Access
impact. Brain. Behav. Immun. 47, 1–3. https://doi.org/10.1016/j.bbi.2015.03.011. Resource for Identifying the Causes of a Wide Range of Complex Diseases of Middle
Paulsen, J., Mehl, A., Dewan, A., Solliga, E., Dama, J.K., Asvold, B., 2017. Associations of and Old Age. PLoS Med. 12, 1–10. https://doi.org/10.1371/journal.pmed.1001779.
obesity and lifestyle with the risk and mortality of bloodstream infection in a general Tacconelli, E., Sifakis, F., Harbarth, S., Schrijver, R., van Mourik, M., Voss, A.,
population : a 15-year follow-up of 64 027 individuals in the HUNT Study. Int. J. Sharland, M., Rajendran, N.B., Rodríguez-Baño, J., Bielicki, J., de Kraker, M.,
Epidemiol. 46, 1573–1581. https://doi.org/10.1093/ije/dyx091. Gandra, S., Gastmeier, P., Gilchrist, K., Gikas, A., Gladstone, B.P., Goossens, H.,
Perreault, K., Bauman, A., Johnson, N., Britton, A., Rangul, V., Stamatakis, E., 2017. Does Jafri, H., Kahlmeter, G., Leus, F., Luxemburger, C., Malhotra-Kumar, S., Marasca, G.,
physical activity moderate the association between alcohol drinking and all-cause, McCarthy, M., Navarro, M.D., Nuñez-Nuñez, M., Oualim, A., Price, J., Robert, J.,
cancer and cardiovascular diseases mortality? A pooled analysis of eight British Sommer, H., von Cube, M., Vuong, C., Wiegand, I., Witschi, A.T., Wolkewitz, M.,
population cohorts. Br. J. Sports Med. 51, 651–657. https://doi.org/10.1136/ 2018. Surveillance for control of antimicrobial resistance. Lancet Infect. Dis. 18,
bjsports-2016-096194. e99–e106. https://doi.org/10.1016/S1473-3099(17)30485-1.
Prather, A.A., Leung, C.W., 2016. Association of insufficient sleep with respiratory Townsend, P., Phillimore, P., Beattie, A., 1988. Health and deprivation: inequality and
infection among adults in the United States. JAMA Intern. Med. 176, 850–852. the North. Routledge.
https://doi.org/10.1001/jamainternmed.2016.0787. UK Biobank, 2007. UK Biobank: Protocol for a large-scale prospective epidemiological
Rehm, J., Baliunas, D., Borges, G.L.G., Graham, K., Irving, H., Kehoe, T., Parry, C.D., resource. UKBB-PROT-09-06 (Main Phase).
Patra, J., Popova, S., Poznyak, V., Roerecke, M., Room, R., Samokhvalov, A.V., Wang, H.E., Baddley, J., Griffin, R.L., Judd, S., Howard, G., Donnelly, J.P., Safford, M.M.,
Taylor, B., 2010. The relation between different dimensions of alcohol consumption 2014. Physical inactivity and long-term rates of community-acquired sepsis. Prev.
and burden of disease: An overview. Addiction 105, 817–843. https://doi.org/ Med. (Baltim) 65, 58–64. https://doi.org/10.1016/j.ypmed.2014.04.017.
10.1111/j.1360-0443.2010.02899.x. Wang, S., Liu, X., Chen, Q., Liu, C., Huang, C., Fang, X., 2017. The role of increased body
Rosenberg, G., Bauld, L., Hooper, L., Buykx, P., Holmes, J., Vohra, J., 2018. New national mass index in outcomes of sepsis: A systematic review and meta-analysis. BMC
alcohol guidelines in the UK: Public awareness, understanding and behavioural Anesthesiol. 17, 1–11. https://doi.org/10.1186/s12871-017-0405-4.
intentions. J. Public Heal. (United Kingdom) 40, 549–556. https://doi.org/10.1093/ Welte, T., Torres, A., Nathwani, D., 2012. Clinical and economic burden of community-
pubmed/fdx126. acquired pneumonia among adults in Europe. Thorax 67, 71–79. https://doi.org/
Rudd, K.E., Johnson, S.C., Agesa, K.M., Shackelford, K.A., Tsoi, D., Kievlan, D.R., 10.1136/thx.2009.129502.
Colombara, D.V., Ikuta, K.S., Kissoon, N., Finfer, S., Fleischmann-Struzek, C., Weycker, D., Strutton, D., Edelsberg, J., Sato, R., Jackson, L.A., 2010. Clinical and
Machado, F.R., Reinhart, K.K., Rowan, K., Seymour, C.W., Watson, R.S., West, T.E., economic burden of pneumococcal disease in older US adults. Vaccine 28,
Marinho, F., Hay, S.I., Lozano, R., Lopez, A.D., Angus, D.C., Murray, C.J.L., 4955–4960. https://doi.org/10.1016/j.vaccine.2010.05.030.
Naghavi, M., 2020. Global, regional, and national sepsis incidence and mortality, Xiao, Q., Keadle, S.K., Hollenbeck, A.R., Matthews, C.E., 2014. Sleep duration and total
1990–2017: analysis for the Global Burden of Disease Study. Lancet 395, 200–211. and cause-specific mortality in a large US cohort: Interrelationships with physical
https://doi.org/10.1016/S0140-6736(19)32989-7. activity, sedentary behavior, and body mass index. Am. J. Epidemiol. 180,
997–1006. https://doi.org/10.1093/aje/kwu222.

27