Sage Journals: Discover world-class research

Abstract

Introduction

Lifestyle-related health behaviors are major contributors to preventable chronic diseases. We aimed to characterize temporal trends in health behaviors of US children, adolescents, and adults and examine the association of sociodemographic characteristics with clustering of unhealthy behaviors.

Methods

The study population included 46,793 participants from the 2005 to 2016 National Health and Nutrition Examination Survey from three age groups: children (aged 2–11), adolescents (aged 12–19), and adults (aged ≥20). We calculated weighted prevalences of unhealthy behaviors—poor diet quality, low physical activity, screentime, fast food consumption, smoking (adolescents and adults), alcohol use, and short sleep duration (adults)—for each survey wave to examine temporal trends across age groups. Multivariable logistic regression estimated associations of sociodemographic characteristics with unhealthy behavior clustering (≥2 behaviors), stratified by age.

Results

Results of the study demonstrated that between 2005 and 2016, the proportion of children and adolescents not meeting the physical activity guidelines increased and screentime increased among all ages. Dietary quality improved and smoking prevalence declined among adolescents and adults, while fast food consumption declined among adolescents. Unhealthy behavior clustering among children increased by 13%. The odds of unhealthy behavior clustering were higher among children and adolescents that were older, non-Hispanic Black, or lived in unmarried households, and among adults who were younger, non-Hispanic black, had lower educational attainment, were uninsured or had public insurance, and had lower poverty-to-income ratios.

Conclusion

The findings suggest that screentime and physical inactivity are growing areas of concern in the U.S. population, and that disparities in adherence to multiple healthy behaviors may contribute to disparities in chronic diseases.

Keywords

Children adolescents adults physical activity diet

Introduction

Cardiovascular disease (CVD) is the leading cause of death in the United States, currently responsible for approximately one in four adult deaths.¹ Lifestyle-related health behaviors including smoking, alcohol consumption, poor diet, physical inactivity, and poor sleep hygiene are modifiable risk factors for chronic disease development. Clustering of these multiple unhealthy behaviors within individuals is associated with increased risk of all-cause, CVD, and cancer mortality.^2,3 Previous research has highlighted disparities in behavioral risk factors across sociodemographic factors. The literature demonstrates greater adherence to dietary intake recommendations among higher socioeconomic status (SES) groups and higher prevalence of childhood obesity among U.S. Hispanic and non-Hispanic Black children.^4,5 Prior studies report that many young people do not meet recommended levels of physical activity, with youth from low SES backgrounds engaging in less physical activity and more sedentary behaviors.^6–8 Finally, clustering of unhealthy behaviors may be more common among youth from racial/ethnic minority populations or lower SES households.⁹

There have been limited published studies examining differences in trends and predictors of multiple lifestyle factors between US children, adolescents, and adults over an extended time period. Such findings could be used to improve existing prevention approaches, contribute to the development of new approaches, and provide an updated epidemiological view of health behavior patterns across age groups. Our objective was to characterize temporal trends in health behaviors of US children, adolescents, and adults from 2005 to 2016. Our secondary objective was to examine the association of sociodemographic factors with clustering of unhealthy behaviors, using the National Health and Nutrition Examination Survey (NHANES).

Methods

Study design and population

This study was a secondary data analyses of repeat cross-sectional data from the 2005 to 2016 NHANES. NHANES uses a multistage, stratified sampling design to assess the health and nutritional status of a nationally representative sample of the civilian, non-institutionalized, US population.¹⁰ The NHANES protocol was approved by the NCHS Research Ethics Review board. Data analysis was performed between January and April of 2019.

The study population included children (ages 2–11 years), adolescents (12–19 years), and adults (20 years and older) who completed a 24-h dietary recall (n = 50,038). We examined these age groups separately to determine how health behavior patterns varied by life stage. We excluded women who were pregnant at the time of the survey (n = 656, 1.3%) and participants who were missing data on health behaviors examined in the analysis (n = 2589, 5.2%), for a total analytic sample of 46,793 participants. Pregnant women were excluded from the study as their health behaviors may not be reflective of their baselines when not pregnant. Participants with incomplete or missing data were excluded to maintain reliability and reproducibility of the results of the study while also minimizing random error.

Health behavior assessment

Table 1 summarizes details of the health behaviors assessed in each survey wave by age group. Certain years were not included in analysis of specific health behaviors due to lack of data availability or changes in data collection methods.

Table 1.

Characteristics of study population (n = 46,793 total).

	Weighted percent or mean (SE)
	Children	Adolescents	Adults
Total N	11,019	7898	27,876
Population represented	39,582,566	32,346,374	205,729,248
Age at screening—years	6.5 (0.1)	15.6 (0.1)	47.7 (0.3)
Gender
Male	51.1%	50.4%	49.2%
Female	48.9%	49.6%	50.8%
Race
White	54.9%	57.1%	69.0%
Hispanic	22.2%	20.3%	13.3%
Black	14.3%	14.8%	11.0%
Other race	8.6%	7.8%	6.7%
Nativity
Foreign-born	3.5%	8.7%	15.7%
US-born	96.5%	91.3%	84.3%
Education category^a
<High school diploma	19.5%	20.2%	16.2%
High school diploma/GED	22.1%	23.0%	22.8%
Some college/associates	30.5%	33.0%	32.0%
Bachelors or higher	27.9%	23.8%	28.9%
Marital status^a
Not married	22.4%	29.3%	37.7%
Married/living with partner	77.6%	70.7%	62.3%
Family income to poverty ratio^b	2.4 (0.1)	2.5 (0.1)	3.0 (0.1)
Insurance
None	7.6%	13.0%	18.2%
Private	52.1%	56.3%	62.4%
Public	40.3%	30.6%	19.4%
Usual source of healthcare
No	3.3%	12.5%	14.9%
Yes	96.7%	87.5%	85.1%

For children and adolescents, this reflects the participant’s household reference person’s education or marital status. For adults, this reflects the participant’s own education or marital status.

Calculated by dividing reported family income by the federal poverty limit; ranges from 0 to 5. For example, a score of 1 indicates family income equal to the federal poverty limit, which a score of 2 reflects family income 200% of the federal poverty limit.

Diet was assessed by 24-h dietary recall administered by trained interviewers using a standardized, computer-assisted interview system. Dietary intakes were converted to overall dietary quality using the Healthy Eating Index (HEI) 2015, a summary score which includes 13 different dietary components (e.g. total vegetables, total fruit, whole grains, added sugars).¹¹ As in other studies,^7,9,12,13 a HEI-2015 score of ≤50 (out of 100 possible) was used to define poor dietary quality.

Physical activity and screentime were assessed using the GPAQ.¹⁴ For adults and adolescents, we calculated the total minutes spent engaged in moderate/vigorous activity per week across for work, transportation, and recreation; for children, we used the number of days that their parents reported they were physically active for at least an hour. We defined “not meeting physical activity guidelines” as less than 150 min/week of moderate to vigorous activity for adults, and as not being physically active for at least an hour a day 7 days/week for children and adolescents.¹⁵ Screen time was assessed using self- or proxy-reported average hours per day spent using a computer or watching television/videos. Excessive screen time was defined as >2 h per day outside of school/work.^16,17

Fast food consumption was assessed using two items from the Flexible Conumer Behavior Survey module,¹⁸ which assessed the total number of meals in the past 7 days that were prepared away from home and the number from a fast-food or pizza place. Excessive fast food consumption was defined as ≥3 times per week, a level of consumption associated with weight gain and insulin resistance.¹⁹

Smoking history data were collected for adults and adolescents only. We classified adults as current smokers if they reported ever smoking 100 cigarettes in their lifetime and currently smoking. We classified adolescents as current smokers if they reported smoking on at least 1 of the past 30 days.²⁰

Alcohol consumption was assessed among adults only as the number of alcoholic drinks consumed per week. We defined heavy alcohol use as >7 drinks/week for women and >14 drinks/week for men.²¹

Sleep duration on weeknights/worknights was collected from adults only. Insufficient sleep was defined as was <7 h of sleep/night.²² Due to changes in the method for ascertaining sleep duration, we excluded sleep data from 2015 to 2016.

Unhealthy behavior clustering

Unhealthy behavior clustering was defined as reporting ≥2 unhealthy behaviors in a given survey cycle. In order to facilitate comparisons across years, we only calculated this variable for survey cycles in which all applicable health behaviors were collected for a particular age group (children: 2009–2016, adolescents: 2011–2016, adults: 2011–2014).

Sociodemographic characteristics

Sociodemographic characteristics included age, sex, race/ethnicity, birthplace (U.S. vs foreign-born), education, marital status, household poverty to income ratio (PIR), health insurance status, and having a usual source of health care. For children and adolescents, the educational attainment and marital status of the household reference person completing the survey (typically a parent or guardian) were assessed.

Statistical analysis

All analyses accounted for the complex NHANES survey design. We used the 1-day dietary weights, and used Stata’s “svy” package with the “subpop” option to appropriately calculate variance estimates that are nationally representative for each age group. We calculated weighted proportions of each age group with unhealthy levels of each health behavior, as well as the proportion engaging in ≥2 unhealthy behaviors, in each survey cycle, to examine descriptively how prevalence of unhealthy behaviors have changed over time. We tested for the presence of linear trends in the prevalence of each unhealthy behavior over time using logistic regression, with survey wave entered as a continuous variable.

Multiple imputation with chained equations was used to impute missing covariate data (10.3% of participants were missing at least one covariate).²³ We used multivariable logistic regression to estimate associations of the sociodemographic characteristics listed above with the odds of having ≥2 unhealthy behaviors, separately for each age group.

Results

Our analysis included 11,019 children, 7898 adolescents, and 27,876 adults across six survey cycles (overall and age group-specific sample sizes by survey cycle are presented in Table 2). Table 1 presents the distribution of demographic and socioeconomic characteristics across the three age groups.

Table 2.

Weighted trends in unhealthy behavior prevalence over 10 years (2007–2016) among U.S. children, adolescents, and adults.^a

	Trends in unhealthy behaviors across NHANES survey cycles—% (95% CI)^b
	2005–2006	2007–2008	2009–2010	2011–2012	2013–2014	2015–2016	p-trend^c	Difference, 2007–2016 (95% CI)
Unhealthy diet (HEI ≤ 50)
Children	57.2 (52.8, 61.5)	57.9 (53.2, 62.7)	50.4 (47.1, 53.8)	49.0 (43.7, 54.4)	54.0 (49.8, 58.2)	54.9 (50.4, 59.3)	0.2	−2.3 (−8.5, 3.9)
Adolescents	72.2 (68.9, 75.4)	72.1 (67.3, 76.8)	67.9 (62.0, 73.8)	63.1 (58.5, 67.6)	64.3 (59.6, 69.0)	63.4 (59.3, 67.4)	<0.001	−8.8 (−14.0, −3.6)*
Adults	53.0 (50.0, 56.1)	50.9 (45.5, 56.3)	46.6 (44.2, 49.0)	45.9 (43.3, 48.6)	48.4 (45.2, 51.7)	48.3 (44.8, 51.8)	0.04	−4.7 (−9.3, −0.1)*
Not meeting PA guidelines^d
Children	–	–	22.8 (18.1, 27.4)	23.6 (19.4, 27.8)	30.2 (25.4, 35.0)	40.7 (36.4, 45.0)	<0.001	17.9 (11.6, 24.3)*
Adolescents	–	41.8 (38.5, 45.1)	48.4 (43.8, 52.9)	44.0 (40.0, 48.1)	51.9 (47.9, 55.9)	47.9 (43.4, 52.5)	0.02	6.1 (0.5, 11.7)*
Adults	–	36.0 (32.9, 39.1)	36.2 (33.0, 39.4)	34.5 (31.6, 37.4)	37.6 (35.9, 39.3)	33.4 (30.1, 36.8)	0.4	−2.6 (−7.2, 2.0)
Screen time >2 h/day^e
Children	36.3 (32.9, 39.7)	42.8 (38.2, 47.4)	41.6 (37.3, 45.8)	42.9 (38.7, 47.1)	44.6 (39.0, 50.2)	50.5 (44.0, 56.9)	<0.001	14.2 (6.9, 21.4)*
Adolescents	56.3 (53.1, 59.4)	–	–	62.7 (58.4, 67.0)	63.6 (60.7, 66.5)	63.7 (59.5, 68.0)	0.001	7.4 (2.1, 12.7)*
Adults	52.2 (50.0, 54.4)	–	–	64.3 (61.2, 67.4)	63.0 (61.3, 64.8)	64.1 (61.0, 67.2)	<0.001	11.9 (8.1, 15.7)*
Fast food ≥3 times/week^f
Children	–	15.6 (11.4, 19.7)	16.0 (12.4, 19.5)	17.8 (13.3, 22.3)	13.5 (10.4, 16.6)	16.0 (13.2, 18.9)	0.8	0.4 (−4.6, 5.5)
Adolescents	–	28.9 (25.6, 32.3)	25.9 (22.1, 29.6)	23.9 (19.9, 27.8)	24.2 (20.8, 27.7)	21.3 (16.4, 26.2)	0.01	−7.6 (−13.6, −1.7)*
Adults	–	23.2 (20.5, 26.0)	20.8 (19.2, 22.5)	22.3 (20.5, 24.1)	21.7 (19.9, 23.5)	22.1 (19.3, 24.9)	0.7	−1.1 (−5.1, 2.8)
Current smoking
Adolescents	16.9 (14.0, 19.7)	16.3 (12.0, 20.5)	16.3 (13.5, 19.2)	14.2 (10.8, 17.6)	7.1 (4.3, 9.9)	5.4 (3.3, 7.6)	<0.001	−11.5 (−15.0, −7.9)*
Adults	24.9 (23.2, 26.7)	23.6 (20.9, 26.4)	20.1 (18.2, 22.0)	20.7 (18.3, 23.0)	19.1 (17.0, 21.1)	18.2 (16.3, 20.1)	<0.001	−6.7 (−9.3, −4.1)*
Heavy alcohol use
Adults	9.3 (7.8, 10.9)	8.1 (6.5, 9.8)	9.3 (7.7, 10.9)	10.5 (8.8, 12.1)	8.5 (7.3, 9.7)	8.2 (6.7, 9.8)	0.6	−1.1 (−3.3, 1.1)
Sleep <7 h/night^f
Adults	35.3 (32.7, 37.9)	37.1 (34.0, 40.1)	36.2 (34.1, 38.3)	36.9 (34.8, 39.0)	35.2 (33.0, 37.4)	–	0.9	−0.1 (−3.4, 3.4)
% With ≥2 unhealthy behaviors^g
Children	–	–	40.3 (37.0, 43.5)	42.3 (38.3, 46.2)	44.0 (39.3, 48.6)	53.6 (47.6, 59.5)	<0.001	13.3 (6.5, 20.1)*
Adolescents	–	–	–	68.7 (64.1, 73.3)	68.9 (65.5, 72.3)	70.0 (65.0, 75.1)	0.7	1.3 (−5.5, 8.1)
Adults	–	–	–	71.7 (68.3, 75.0)	71.8 (69.2, 74.4)	–	0.9	0.1 (−4.1, 4.4)
Mean # unhealthy behaviors
Children	–	–	1.3 (1.2, 1.4)	1.3 (1.2, 1.4)	1.4 (1.3, 1.5)	1.6 (1.5, 1.8)	<0.001	0.3 (0.2, 0.5)*
Adolescents	–	–	–	2.1 (2.0, 2.2)	2.1 (2.0, 2.2)	2.0 (1.9, 2.1)	0.4	−0.1 (−0.2, 0.1)
Adults	–	–	–	2.3 (2.3, 2.4)	2.3 (2.3, 2.4)	–	0.8	0.0 (−0.1, 0.1)

HEI: Healthy Eating Index 2015; PA: physical activity.

Age groups defined as: children: aged 2–11 years, adolescents: aged 12–19 years, adults: aged 20 years and older.

Confidence intervals calculated using logistic regression with survey cycle entered as a categorical variable, followed by marginal standardization to calculate predicted proportions and 95% confidence interval.

p for trend calculated using logistic regression with survey cycle entered as a continuous variable.

Physical activity measures changed before the 2007–2008 cycle for adults and adolescents and before the 2009–2010 cycle for children, so physical activity prior to those years are not included.

Screen time not collected for adolescents or adults in cycles 2007–2008 and 2009–2010.

Fast food consumption not assessed in 2005–2006 cycle for any age group.

Sleep assessment method changed in 2015–2016, so this wave is not included in the analysis.

Unhealthy behavior clustering was only assessed in survey cycles where all age-group-specific health behaviors were collected.

Temporal trends in health behaviors

Table 2 presents weighted proportions of children, adolescents, and adults with each unhealthy behavior pattern across the 6 NHANES survey waves (2005/2006–2015/2016). Results are described for each health behavior below.

Dietary quality

Among adolescents, the percent with poor dietary quality (HEI-2015 ≤50) decreased by 8.8 percentage points, from 72.2% to 63.4%, (95% CI: −14.0, −3.6) from 2005/2006 to 2015/2016. A smaller decrease was noted among adults (-4.7 percentage points, 95% CI: −9.3, −0.1) and children (−2.3 percentage points, 95% CI: −8.5, 3.9) (Table 2). In 2015/2016, poor dietary quality was most prevalent among adolescents (63.4%), followed by children (54.9%) then adults (48.3%, Figure 1). It should also be noted that the largest improvements in dietary quality among all groups occurred between the years 2005/2006 and 2011/2012. From that point forward, rates of poor diet quality remained stagnant or increased.

Figure 1.

Prevalence of unhealthy behaviors in 2015–2016 by age group.

Physical activity

The proportion of children and adolescents who did not meet physical activity guidelines increased over time, by 17.9 (11.6, 24.3) percentage points from 2009/2010 to 2015/2016 among children and by 6.1 (0.5, 11.7) percentage points from 2007/2008 to 2015/2016 among adolescents, respectively. Adults not meeting physical activity guidelines decreased by 2.6 percentage points (95% CI: −7.2, 2.0) from 2007/2008 to 2015/2016. In 2015/2016, lack of adherence to physical activity guidelines was most prevalent among adolescents (47.9% compared to 40.7% among children and 33.4% among adults, Figure 1).

Screentime

The proportion of U.S. children, adolescents, and adults spending >2 h in screen time increased between 2005/2006 and 2015/2016. The largest increase was among children (14.2 percentage points, 95% CI: 6.9, 21.4), followed by adults (11.9 percentage points, 95% CI: 8.1, 15.7), then adolescents (7.4 percentage points, 95% CI: 2.1, 12.7). In 2015/2016, adults and adolescents had similar prevalence of excessive screentime, with lower prevalence among children (Figure 1).

Fast food consumption

Weekly fast food consumption frequency ≥3 times per week did not change significantly for children (0.4, 95% CI: −4.6, 5.5) or adults (−1.1, 95% CI: −5.1, 2.8) over the course of the study. However, adolescents demonstrated a significant decrease in weekly fast food consumption by 7.6 percentage points (95% CI: −13.6, −1.7) between 2007/2008 and 2015/2016. In 2015/2016, 16.0% of children ate fast food ≥3 times per week compared to 21.3% of adolescents and 22.1% of adults.

Smoking

The prevalence of cigarette smoking declined from 2005/2006 to 2015/2016 among both adolescents and adults. Among adolescents, current smoking prevalence declined by 11.5 percentage points (−15.0, −7.9), from 16.9% to 5.4%. Among adults, smoking prevalence declined by 6.7 percentage points (−9.3, −4.1) from 24.9% to 18.2%.

Alcohol use

The prevalence of heavy alcohol use among adults was fairly stable across time, declining by only 1.1 percentage points (95% CI: −3.3, 1.1) from 9.3% in 2005/2006 to 8.2% in 2015/2016.

Sleep

Just over one-third of adults reported insufficient sleep on weeknights/worknights, with minimal change over time from 35.3% in 2005/2006 to 35.2% in 2013/2014. Sleep estimates are not presented for 2015/2016 due to a change in the sleep assessment method.

Unhealthy behavior clustering

The proportion of children engaging in ≥2 unhealthy behaviors increased by 13.3 (6.5, 20.1) percentage points, from 40.3% in 2009/2010 to 53.6% in 2015/2016. Fewer years were available for evaluation among adolescents and adults since all health behaviors were not assessed in all years. The proportion of adolescents and adults engaging in ≥2 unhealthy behaviors was approximately 70% and changed little across available survey waves (Table 2). The mean number of unhealthy behaviors increased from 1.3 to 1.6 among children (p for trend <0.001) from 2009/2010 to 2015/2016, while remaining stable for adolescents and adults.

Associations of sociodemographic characteristics with unhealthy behavior clustering

Children’s odds of engaging in ≥2 unhealthy behaviors increased by 21% for each additional year of age (odds ratio (OR): 1.21, 95% CI: 1.18, 1.25, Table 3). Non-Hispanic Black children had 51% higher odds of reporting ≥2 unhealthy behaviors (OR 1.51, 95% CI: 1.27, 1.80) compared to non-Hispanic White children. Foreign-born children and those residing in married households were less likely to engage in ≥2 unhealthy behaviors (OR: 0.56 (0.40, 0.78) and OR: 0.75 (0.62, 0.90), respectively).

Table 3.

Association of sociodemographic characteristics with unhealthy behavior clustering by age group.

	≥2 Unhealthy behaviors—odds ratio (95% CI)
	Children	Adolescents	Adults
Age, years^a	1.21 (1.18, 1.25)*	1.08 (1.03, 1.14)*	0.97 (0.95, 0.99)*
Gender
Female	Ref	Ref	Ref
Male	1.10 (0.95, 1.26)	0.84 (0.65, 1.09)	1.18 (0.99, 1.40)
Race
White	Ref	Ref	Ref
Hispanic	0.94 (0.77, 1.15)	1.19 (0.91, 1.55)	0.86 (0.69, 1.06)
Black/African American	1.51 (1.27, 1.80)*	1.42 (1.02, 1.98)*	1.79 (1.51, 2.11)*
Other race	1.03 (0.75, 1.41)	0.93 (0.62, 1.40)	1.22 (0.93, 1.58)
Birthplace
US born	Ref	Ref	Ref
Foreign-born	0.56 (0.40, 0.78)*	1.02 (0.71, 1.47)	0.47 (0.37, 0.58)*
Education^b
<HS	1.07 (0.79, 1.46)	1.23 (0.81, 1.86)	2.45 (1.94, 3.10)*
HS degree/GED	1.27 (0.94, 1.76)	1.56 (1.03, 2.36)*	2.50 (2.07, 3.01)*
Some college/associate’s	1.19 (0.97, 1.46)	1.20 (0.82, 1.77)	1.90 (1.52, 2.38)*
Bachelor’s or higher	Ref	Ref	Ref
Marital status^b
Not married	Ref	Ref	Ref
Married/living with partner	0.75 (0.62, 0.90)*	0.62 (0.45, 0.84)*	0.96 (0.79, 1.17)
Insurance status
Private	Ref	Ref	Ref
Public	0.97 (0.79, 1.19)	1.03 (0.74, 1.44)	1.34 (1.08, 1.66)*
None	1.29 (0.97, 1.70)	1.11 (0.67, 1.82)	1.31 (1.04, 1.63)*
Usual health care source
Yes	Ref	Ref	Ref
No	1.01 (0.65, 1.56)	1.09 (0.72, 1.66)	0.97 (0.81, 1.17)
Family income-poverty ratio
Tertile 1 (lowest)	Ref	Ref	Ref
Tertile 2	1.16 (0.96, 1.39)	0.91 (0.66, 1.27)	0.90 (0.73, 1.11)
Tertile 3 (highest)	1.03 (0.79, 1.35)	1.03 (0.67, 1.57)	0.69 (0.56, 0.85)*
Survey cycle^c
2009/2010	Ref	–	–
2011–2012	1.10 (0.86, 1.40)	Ref	Ref
2013–2014	1.14 (0.91, 1.43)	1.03 (0.80, 1.31)	1.01 (0.86, 1.19)
2015–2016	1.78 (1.34, 2.36)*	1.08 (0.82, 1.44)	–

Reflects a 1-year difference for children and adolescents; reflects a 5-year difference for adults.

Model only includes survey cycles where all age-group-applicable health behaviors were assessed. Survey cycles 2009–2010 through 2015–2016 were included for children, cycles 2011–2012 through 2015–2016 were included for adolescents, and cycles 2011–2012 and 2013–2014 were included for adults.

p < 0.05.

Among adolescents, each 1-year increase in age was associated with an 8% increase in the odds of unhealthy behavior clustering (OR: 1.08, 95% CI: 1.03, 1.14). Non-Hispanic Black adolescents had 42% higher odds of having ≥2 unhealthy behaviors compared to non-Hispanic White adolescents (OR 1.42, 95% CI: 1.02, 1.98). Adolescents whose household reference person had a high school education had 56% higher odds (OR 1.56, 95% CI: 1.03, 2.36) of having ≥2 unhealthy behaviors compared to those with a Bachelor’s degree or higher, while adolescents residing in a married households had 38% lower odds (OR: 0.62, 95% CI: 0.45, 0.84).

Adults had 3% lower odds of unhealthy behavior clustering for each 5-year increase in age (OR 0.97, 95% CI: 0.95, 0.99). The odds of having multiple unhealthy behaviors were higher among males compared to females (OR 1.18, 95% CI: 0.99, 1.40) and among non-Hispanic Black adults compared to non-Hispanic White adults (OR 1.79, 95% CI: 1.51, 2.11). An inverse relationship was observed between clustering of unhealthy behaviors and higher levels of education in adults. Adults with public or no health insurance had higher odds of engaging in multiple unhealthy behaviors than adults with private insurance (OR 1.34 (1.08, 1.66) and 1.31 (1.04, 1.63), respectively), while foreign-born adults had lower odds (OR 0.47, 95% CI: 0.37, 0.58) compared to U.S. born adults. Finally, adults in the highest versus lowest PIR tertile had 31% lower odds of unhealthy behavior clustering (OR: 0.69, 95% CI: 0.56, 0.85).

Discussion

This study provides an updated view of trends in the prevalence of multiple health behaviors among U.S children, adolescents, and adults, and elucidates similarities and differences in the sociodemographic correlates of unhealthy behavior clustering across age groups. We identified several positive health behavior changes over the last decade, including improvements in diet and reductions in smoking prevalence among adolescents and adults, as well as a reduction in fast food consumption among adolescents. However, other behavioral trends worsened over time, including reductions in adherence to physical activity guidelines for children and adolescents, an increase in screen time for all age groups, and an increase in the prevalence of unhealthy behavior clustering among children. Disparities in unhealthy behavior clustering were apparent by race and age for all age groups, and by household marital status for children and adolescents. Differences by educational attainment, nativity, health insurance status, and household income were also observed, but varied across age groups.

The increase in prevalence of children and adolescents who did not meet national guidelines for physical activity is larger than that seen in earlier studies. The Health Behaviour in School-aged Children study reported that that the percent of U.S. students aged 11–15 meeting physical activity guidelines declined by approximately 3 percentage points from 2002 to 2010,²⁴ while an analysis of Youth Risk Behavior Surveillance System data from 1999 to 2005 indicated stable patterns in physical activity among students in grades 9–12.²⁵ The decline in physical activity in recent years could be due to various social, economic, and cultural factors. One potential factor may be the decline or complete removal of recess time in schools over recent years.²⁶ Schools are a principle setting for children to engage in their daily physical activity and are critical for interventions to improve physical activity levels.²⁷ Another potential contributor is the rise in use of social media among children and adolescents, which may have also contributed to the progressive increase in screentime over the same period.²⁸ The continued decline in physical activity over the years is an issue of public health significance, as childhood obesity has been associated with increased risk of obesity into adulthood and the development of future chronic diseases.^29,30

The results of this study demonstrate a decrease in the proportion of adolescents and adults with poor dietary quality, as well as a decline in adolescent fast food consumption. However, most of the decline occurred early in the study period and stagnated. The prevalence of an unhealthy diet remained high across all age groups, and was the most prevalent unhealthy behavior among children. This is concerning as childhood and adult obesity rates continue to rise in the US, including a significant increase in obesity among 2–5 years old children, posing a major healthcare threat to the population.³¹

Clustering of unhealthy behaviors was lower among foreign-born participants, non-Hispanic White individuals, youth living in married households, and adults with private health insurance, higher educational attainment and higher income. Foreign-born participants may be less likely to follow a westernized diet high in saturated fats and sucrose and low in fiber, patterns which have been shown to contribute to increased occurrences of cardiometabolic diseases.^32,33 Individuals with higher levels of education may correlate with higher income earning potential, private health insurance, and larger social networks, all of which may reduce the use of stress-associated behaviors to cope with prolonged social and economic hardships.³⁴ Individuals with higher SES and private health insurance may also have increased access to healthcare providers, leading to higher likelihood of early intervention and suggested follow-up. However, the increased odds of health behavior clustering among non-Hispanic Black individuals across age groups is of particular concern, given persistent racial disparities in deaths from preventable chronic diseases. Racial disparities in health outcomes are complex and multifactorial, rooted in systemic racism and influenced by inequities in environmental exposures (e.g. access to healthy food, safe recreational spaces) and access to health care.³⁵

Our study adds to the existing literature by enabling comparisons across multiple behaviors and age groups, to identify key areas of concern and understand whether trends are consistent across different age groups. These findings demonstrate the importance of addressing specific lifestyle risk factors in order to reduce disparities and overall chronic disease prevalence. This study utilized a large, nationally representative data set, which allowed for accurate estimates of the prevalence of health behaviors among U.S. children, adolescents, and adults over time. A strength of this examination of health behavior patterns and is clustering from a life course perspective, enabling comparison across different life stages. Nevertheless, this study had several important limitations, including the lack of data from certain survey years due to changes in assessment methods or measures. This limited our ability to examine longitudinal trends in health behavior clustering among adolescents and adults. Also, because NHANES uses a repeat cross-sectional design, we were able to examine population-level changes but not changes in health behaviors among individuals over time. Additionally, self- or parent-reported data was used for all the health behaviors, which may be subject to recall bias or social desirability bias. Objective measurements of health behaviors were not available in all years and for all age groups. Lastly, with the recent emergence of vaping and electronic cigarettes, we were unable to examine their use as data was not available in earlier survey results.

Conclusions

In a nationally representative sample of U.S. children, adolescents, and adults examining trends from 2005 to 2016, dietary patterns largely improved, screentime increased, and physical activity declined among youth. Results suggest screen time and physical inactivity as particularly important areas to target with public health prevention campaigns.

Footnotes

Author contributions

Naser Mubarak N: Manuscript Preparation, Data Analysis. Austin Wynn: Manuscript Revision. Brandon Tapasak: Manuscript editing. Andrew Collins: Manuscript Revision. Norah Mubarak: Manuscript Preparation. Carla Gonzalez: Manuscript Preparation. Gustavo Marino: Manuscript Editing. Stephanie Mayne: Manuscript Design, Data Analysis, and Manuscript Editing. Acknowledgement that all authors have contributed significantly and that all authors agree with the content of the manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethics approval and consent to participate

Data obtained and analyzed was from a publicly available data set from the CDC. IRB exempt status.

Informed consent

The manuscript does not contain any individual person’s data in any form.

Significance for public health

Health behaviors such as smoking, physical activity, alcohol consumption, and sleep hygiene have been recognized as significant contributors toward the development and rate of progression of chronic diseases. Chronic diseases, including cardiovascular disease, cancer, and COPD have become leading causes of death. The results of this study illustrating the increased prevalence of certain unhealthy behaviors across different age groups over time, and differences in unhealthy behavior clustering across sociodemographic groups, may be used to elucidate existing disparities found within the prevalence of certain chronic diseases. This study was carried out between May 2019 and June 2021.

Availability of data and materials

All data generated or analyzed during this study are included in this published article.

References

Virani

Alonso

Benjamin

, et al. Heart Disease and Stroke Statistics-2020 update: a report from the American Heart Association. Circulation 2020; 141(9): e139–e596.

Kvaavik

Batty

Ursin

, et al. Influence of individual and combined health behaviors on total and cause-specific mortality in men and women: the United Kingdom health and lifestyle survey. Arch Intern Med 2010; 170(8): 711–718.

van Dam

Spiegelman

, et al. Combined impact of lifestyle factors on mortality: prospective cohort study in US women. BMJ 2008; 337: a1440.

Darmon

Drewnowski

Does social class predict diet quality?

Am J Clin Nutr 2008; 87(5): 1107–1117.

Hales

Carroll

Fryar

, et al. Prevalence of obesity among adults and youth: United States, 2015-2106. NCHS data brief, no 288. Hyattsville, MD: National Center for Health Statistics, 2017.

Janssen

Boyce

Simpson

, et al. Influence of individual- and area-level measures of socioeconomic status on obesity, unhealthy eating, and physical inactivity in Canadian adolescents. Am J Clin Nutr 2006; 83(1): 139–145.

Jackson

Yang

Zhang

Income disparities and cardiovascular risk factors among adolescents. Pediatrics 2018; 142(5).

Drenowatz

Eisenmann

Pfeiffer

, et al. Influence of socio-economic status on habitual physical activity and sedentary behavior in 8- to 11-year old children. BMC Public Health 2010; 10: 214.

Mayne

Virudachalam

Fiks

AG.

Clustering of unhealthy behaviors in a nationally representative sample of U.S. children and adolescents. Prev Med 2020; 130: 105892.

10.

Curtin

Mohadjer

Dohrmann

, et al. The National Health and Nutrition Examination Survey: sample design, 1999-2006. Vital Health Stat 2 2012; (155): 1–39.

11.

Krebs-Smith

Pannucci

Subar

, et al. Update of the Healthy Eating Index: HEI-2015. J Acad Nutr Diet 2018; 118(9): 1591–1602.

12.

Hurley

Oberlander

Merry

, et al. The healthy eating index and youth healthy eating index are unique, nonredundant measures of diet quality among low-income, African American adolescents. J Nutr 2009; 139(2): 359–364.

13.

Ford

Greenlund

Hong

Ideal cardiovascular health and mortality from all causes and diseases of the circulatory system among adults in the United States. Circulation 2012; 125(8): 987–995.

14.

Bull

Maslin

Armstrong

Global Physical Activity Questionnaire (GPAQ): nine country reliability and validity study. J Phys Act Health 2009; 6(6): 790–804.

15.

U.S. Department of Health and Human Services. Physical activity guidelines for Americans. 2nd ed. Washington, DC: U.S. Department of Health and Human Services, 2018.

16.

Council on Communications and Media. Children, adolescents, and the media. Pediatrics 2013; 132(5): 958–961.

17.

National Heart, Lung, and Blood Institute. Expert panel on integrated guidelines for cardiovascular health and risk reduction in children and adolescents: summary report. Pediatrics 2011; 128(Suppl 5): S213–S256.

18.

Economic Research Service. Flexible consumer behavior survey, https://www.ers.usda.gov/topics/food-choices-health/food-consumption-demand/flexible-consumer-behavior-survey/ (accessed 29 August 2019).

19.

Pereira

Kartashov

Ebbeling

, et al. Fast-food habits, weight gain, and insulin resistance (the CARDIA study): 15-year prospective analysis. Lancet 2005; 365(9453): 36–42.

20.

Cornelius

Wang

Jamal

, et al. Tobacco product use among adults - United States, 2019. MMWR Morb Mortal Wkly Rep 2020; 69(46): 1736–1742.

21.

U.S. Department of Health and Human Services and U.S. Department of Agriculture. 2015–2020 dietary guidelines for Americans. 8th ed. http://health.gov/dietaryguidelines/2015/guidelines/ (2015).

22.

Watson

Badr

Belenky

, et al. Recommended amount of sleep for a healthy adult: a joint consensus statement of the American Academy of Sleep Medicine and Sleep Research Society. Sleep 2015; 38(6): 843–844.

23.

White

Royston

Wood

AM.

Multiple imputation using chained equations: Issues and guidance for practice. Stat Med 2011; 30(4): 377–399.

24.

Kalman

Inchley

Sigmundova

, et al. Secular trends in moderate-to-vigorous physical activity in 32 countries from 2002 to 2010: a cross-national perspective. Eur J Public Health 2015; 25(Suppl 2): 37–40.

25.

Treuth

Wang

How active are American adolescents and have they become less active?

Obes Rev 2010; 11(12): 847–862.

26.

Murray

Ramstetter

The crucial role of recess in school. Pediatrics 2013; 131(1): 183–188.

27.

Parrish

Chong

Moriarty

, et al. Interventions to change school recess activity levels in children and Adolescents: A systematic review and meta-analysis. Sports Med 2020; 50: 2145–2173.

28.

Council on Communications and Media. Media use in school-aged children and adolescents. Pediatrics 2016; 138(5): e20162592.

29.

Freedman

Khan

Serdula

, et al. The relation of childhood BMI to adult adiposity: the Bogalusa Heart Study. Pediatrics 2005; 115(1): 22–27.

30.

Llewellyn

Simmonds

Owen

, et al. Childhood obesity as a predictor of morbidity in adulthood: a systematic review and meta-analysis. Obes Rev 2016; 17(1): 56–67.

31.

Skinner

Ravanbakht

Skelton

, et al. Prevalence of obesity and severe obesity in US children, 1999-2016. Pediatrics 2018; 141(3).

32.

Fung

McCullough

van Dam

, et al. A prospective study of overall diet quality and risk of type 2 diabetes in women. Diabetes Care 2007; 30(7): 1753–1757.

33.

Yang

Zhang

Gregg

, et al. Added sugar intake and cardiovascular diseases mortality among US adults. JAMA Intern Med 2014; 174(4): 516–524.

34.

Ross

CL.

The links between education and health. Am Sociol Rev 1995; 60(5): 719–745.

35.

Williams

Mohammed

Leavell

, et al. Race, socioeconomic status, and health: complexities, ongoing challenges, and research opportunities. Ann N Y Acad Sci 2010; 1186: 69–101.

An examination of temporal trends in health behaviors among United States children,adolescents,and adults from 2005 to 2016

Abstract

Introduction

Methods

Results

Conclusion

Keywords

Introduction

Methods

Study design and population

Health behavior assessment

Unhealthy behavior clustering

Sociodemographic characteristics

Statistical analysis

Results

Temporal trends in health behaviors

Dietary quality

Physical activity

Screentime

Fast food consumption

Smoking

Alcohol use

Sleep

Unhealthy behavior clustering

Associations of sociodemographic characteristics with unhealthy behavior clustering

Discussion

Conclusions

Footnotes

Author contributions

Declaration of conflicting interests

Funding

Ethics approval and consent to participate

Informed consent

Significance for public health

Availability of data and materials

References