The current study examined the prevalence and clustering of 5 health-risk behaviors among adolescents in Hawaii, including physical inactivity, low fruit and vegetable consumption, junk food consumption, excessive television time, and inadequate sleep. High school students were recruited from 5 classrooms in Oahu Hawaii. Data were collected in the spring semester of 2011. Proportions were used to describe the prevalence of single and multiple health risk behaviors. Significant health behavior clusters were revealed using an observed-to-expected (O/E) ratio method. Participating adolescents (n = 114) were 11th and 12th grade students with a mean age of 16.28 (SD = 0.62). Participants were predominantly female (75%) and Filipino-American (68%). Seventy-seven percent of adolescents were physically inactive, 90% watched excessive TV, 66% consumed inadequate fruits and vegetables, 94% reported inadequate levels of sleep, and 80% consumed excessive junk food. Overall, 94% reported at least 3 risk factors, 73% reported at least 4 risk factors, and 37% reported all 5 risk factors. No significant clusters were found. Conclusion: Health-risk behaviors cluster and occur more often than expected among adolescents living in Hawaii. Non-significant clustering may be due to insufficient variability within the sample data; future examinations of this highly understudied population are necessary.
Over the last 20 to 30 years, the prevalence of youth obesity has spread to epidemic proportions [
Evidence suggests that common health-risk behaviors cluster or co-occur among youth [
The current research targeted adolescents (predominantly Filipino) residing in Hawaii, where the obesity epidemic has become a severe concern [
The current study objective was to assess the prevalence and distribution of multiple health risk behaviors including physical inactivity, excessive television time, low fruit and vegetable consumption, excessive junk food consumption, and inadequate sleep. Investigating the co-occurrence and clustering of these behaviors will aid in the understanding of behavior-related risk among Filipino-American adolescents, as well as inform early intervention efforts.
Participating adolescents were 11th (85%) and 12th (15%) grade high school students residing in Oahu, Hawaii. Through a university-teacher partnership, students were recruited from five separate high school classrooms. The school was located in a highly populated city with a predominantly Filipino American community and a median household income of $49,444. During the 2010/11 school year, the high school student population was primarily Filipino American (65.5%) followed by Samoan (9.3%), Part-Hawaiian (6.5%), and other or mixed ethnicities (18.7%). Forty-nine percent of students were included in the free/reduced school lunch program during that same year, which was slightly higher than the average Hawaii public high school at 47% [
Informed consent preceded all study procedures, which were approved by the University Institutional Review Board and the State of Hawaii Department of Education. Participating adolescents completed self-reported surveys in-class, followed by corresponding focus groups. A trained social worker led the follow-up focus group discussion (≈30 - 35 minutes), encouraging participating adolescents to expand and/or adjust their survey responses. This two-step assessment ensured understanding of survey questions, and allowed adolescents to elaborate on the meaning and reasoning of their answers. Data analyzed in the current study included participants’ self-reported demographics and health behaviors.
Participants responded to questions pertaining to their demographics, physical activity, television viewing time, fruit and vegetable consumption, junk food consumption and sleep. Demographic data included age, grade, gender and ethnicity. The current analysis is part of a larger, subsequent study that examines potential correlates within the physical environment.
Physical activity was measured as the frequency (days per week) and duration (minutes per day) of moderate and vigorous activity. Moderate physical activity was defined as an “activity that requires some effort, and makes your breathing harder than normal. Some examples would be fast walking, bicycling, swimming, weight lifting, baseball, softball, tennis, volleyball, hula, yoga or dancing.” Vigorous physical activity was defined as an “activity that makes your heart beat quickly, and makes you breath very hard (cannot maintain a conversation). Some examples would be running, jogging, fast bicycling, aerobic dance, rollerblading, paddling, fast swimming, soccer, basketball, football, or martial arts.” The specific types of physical activity used to describe the two intensity levels were based on results of previous focus groups with Filipino adolescents [
Fruit and vegetable consumption was measured as sizes established from the food guide pyramid; therefore, 1 serving of fruit and 1 serving of fruit juice were defined to participants as ‘‘one medium piece of fresh fruit, 1/2 cup of fruit salad, 1/4 cup of raisins, apricots, or other dried fruit, 6 oz. of 100% orange, apple, or grape juice (do not count fruit punch, lemonade, Gatorade, Sunny Delight or fruit drink)”. One serving of vegetables was defined as “1 medium carrot or other fresh vegetable, 1 small bowl of green salad, 1/2 cup of fresh or cooked vegetables, or 3/4 cup of vegetable soup (do not count French fries, onion rings, potato chips, fried tempura or fried okra).”
Participants responded to two separate questions regarding fruit and vegetable consumption. The first asked, “How many servings of fruits do you usually eat each day”, and the second asked, “How many servings of vegetables do you usually eat each day”. For each question, response options included “none, 1, 2, 3, 4, 5, and 6 or more”. Participants who reported less than 5 servings of fruit and vegetables combined were considered at risk for inadequate fruit and vegetable consumption. Youth are recommended to consume 2 cups of fruit (4 servings) and 2 1/2 cups of vegetables each day [
Junk food consumption was assessed by giving participants a list of common snacks and foods high in saturated fat and sugar (i.e., junk foods). During a previous visit to the classroom, participants were asked to list the high fat and sugar foods they typically consume, which were then operationalized into junk food servings prior to survey data collection. One junk food serving was specified for each listed snack, and participants were asked to report the number of servings they consume in a typical day. Junk food servings were determined via kcal; one junk food serving was represented as high fat/sugar foods with portion sizes containing ≥300 kcal. The calories and portion sizes were based on the U.S. Department of Agriculture’s Calorie Counter National Nutrient Database for Standard Reference [
Junk food servings were specified to participants in portion sizes. One serving was defined for the following foods: Ice cream (1/2 cup), soda pop (1 can/bottle/medium size), French fries (1 cup/1 medium size), spam musubi (1 with/without egg), chips (1 snack bag/1 cup/10-12 chips), pretzels (1 cup/15 - 20 pretzels/1 grab bag), gummy candy (1/4 cup jelly beans/2 fun size packs/9 gummy worms), chocolate bar (1 bar/2 fun bars/4 small pieces), cookies (4 medium cookies/2 big cookies, skittles (2 fun size/1 package/1/4 cup, one brownie, popcorn/caramel corn (1 cup/1 package). Participating students also reported commonly eating junk foods defined by multiple servings including the following: Burger with or without cheese (1 medium patty/1/4 pound = 2 servings), pizza (1 slice = 2 servings), cake (1 piece = 2 servings), and nachos with beef, cheese and sour cream (6 servings). Participants who reported consuming ≥3 servings of junk food daily were considered “At Risk” for excessive junk food consumption (J).
Television time was measured by asking participants to report how much time they spend watching television on a typical day with the question, “On average, how many hours per day do you watch television or movies”. Those reporting ˃ 2 hours per day were considered “At Risk” for excessive television time (TV).
To assess sleep, participants responded to the following question regarding how much they typically sleep each day, “On average, how many hours per day do you sleep”. Those reporting < 9 hours of sleep per day were considered “At Risk” for inadequate sleep (S).
The observed prevalence of single health behavior risk factors was described as a binary variable (At Risk = 1; Not at Risk = 0), and single risk behavior prevalence rates were calculated for each of the five health risk behaviors. Prevalence rates for each of the five behaviors were used to determine clustering of behavioral risk factors [
Clustering of behavioral risk factors was assessed using the observed to expected ratio (O/E) method, and the strongest associations among risk factors were identified using 95% confidence intervals. Clustering occurs when the observed proportion is greater than the expected proportion, indicating that certain combinations of behaviors occur more often than expected based on single behavior prevalence rates [
After calculating the observed and expected proportions for all 32 non-overlapping combinations, clustering of risk behaviors was assessed by dividing the observed proportions by the expected values for each combination (O/E ratio). Proportions > 1 indicate significant clustering. Ninety-five percent confidence intervals were calculated for each of the ratios [
Characteristics of the study population are presented in
The most common single behavioral risk factor was inadequate sleep, with a prevalence of approximately 94%. Excessive television time was the second most common, with approximately 90% at risk, followed by excessive junk food consumption (79%).
Adolescents | ||||
---|---|---|---|---|
Male | Female | |||
N (%) 114 | 24 | 76 | ||
Mean Age (SD) | 16.3 (0.603) | 16.2 (0.524) | ||
Grade Level | 12th N (%) | 11th N (%) | 12th N (%) | 11th N (%) |
Grade Level | 4 (3.50) | 25 (21.90) | 13 (11.40) | 73 (64.04) |
Ethnicity | N | % | N | % |
Filipino American | 20 | 17.86 | 58 | 51.79 |
Filipino American Mixed Ethnicity | 7 | 6.25 | 16 | 14.29 |
Pacific Islander Mix | 1 | 0.89 | 6 | 5.36 |
Other | - | - | 4 | 3.57 |
Data collected from high school students in Oahu, Hawaii.
At Risk (%) | Not at Risk (%) | |
---|---|---|
Inadequate Physical Activity (PA)a | 71.9 | 23.4 |
Excessive Television Time (TV)b | 89.5 | 10.5 |
Inadequate Fruit & Vegetable Intake (FV)c | 65.8 | 34.2 |
Excessive Junk Food Intake (J)d | 78.9 | 21.1 |
Inadequate Sleep (S)e | 93.9 | 6.1 |
Data collected from high school students in Oahu, Hawaii. aInadequate physical activity: engaging in less than 60 minutes of activity per day; bExcessive television time: watching more than 2 hours of television per day; cInadequate fruit and vegetable intake: less than 5 servings of fruit and vegetables combined per day; dExcessive junk food intake: consuming 3 or more servings of junk food per day; eInadequate sleep: less than 9 hours of sleep per night.
All participants reported at least one or more behavioral risk factors, however, the majority of the study population (approximately 73%) reported having four or more risk behaviors. Three (2.6%) of the 114 participants reported having ONLY one risk factor, and five (4.4%) participants reported having ONLY two risk factors. Twenty percent reported three risk factors, and the largest proportion of adolescents reported four (36%) or five (37%) behavioral risk factors. These results are also shown in
The combination of having all five risk factors showed clustering with an O/E ratio of 1.17 (CI: 0.89 - 1.46), indicating that the proportion of occurrence for all five risk factors was 17% greater than would be expected if risk factors occurred independently. Among three-behavior patterns, PA/FV/S and TV/FV/S showed the greatest degree of clustering, occurring 92% and 73% more than expected (see
This study is an initial report of the prevalence and clustering of multiple health risk behaviors in a representative sample of Filipino-American adolescents residing in Hawaii, a highly understudied population. Approximately 93% of the study population had three or more risk behaviors, and 73% had four or five risk behaviors. In total, 100% of the sample reported at least one risk behavior. To our knowledge, few studies have examined the prevalence of multiple behavioral risk factors in youth, and none have examined such among similar ethnically diverse youth populations. Although not directly comparable, our results do mirror previous multiple health behavior studies among youth [
A similar study of 11 - 15 year olds in San Diego reported about 50% of participants did not meet physical activity guidelines (60 minutes per day), 33% watched over two hours of television per day, and over 88% did not meet dietary guidelines for fruit and vegetable servings per day (5 servings per day) [
In a 2005 study of Australian 14-year olds, approximately 10% of participants had three or four co-occurring health risk factors including, excessive television time, tobacco use, overweight and high blood pressure [
In a representative sample of Canadian children and adolescents, ages 10 to 17, Alamian and Paradis [
Number of Health Risk Behaviors | At Risk (% of Sample) |
---|---|
1 | 2.63 |
2 | 4.39 |
3 | 20.17 |
4 | 35.96 |
5 | 36.84 |
Data collected from high school students in Oahu, Hawaii.
No. | Combinations | Observed % | Expected % | O/E Ratio | 95% CI (P-Value) | |
---|---|---|---|---|---|---|
0 | - | 0 | - | - | - | |
1 | PA | 0 | 0.03 | 0 | - | |
1 | TV | 0 | 0.11 | 0 | - | |
1 | FV | 0 | 0.02 | 0 | - | |
1 | J | 0.88 | 0.05 | 18.06 | −18.18 - 55.24 (0.12) | |
1 | S | 1.75 | 0.19 | 8.78 | −5.35 - 32.46 (0.02) | |
Total | 2.6 | 0.42 | 6.23 | |||
2 | PA, TV | 0 | 0.28 | 0 | - | |
2 | PA, FV | 0 | 0.06 | 0 | - | |
2 | PA, J | 0 | 0.12 | 0 | - | |
2 | PA, S | 0.88 | 0.51 | 1.72 | −1.69 - 5.13 (0.10) | |
2 | TV, FV | 0 | 0.21 | 0 | - | |
2 | TV, J | 0 | 0.41 | 0 | - | |
2 | TV, S | 1.75 | 1.70 | 1.03 | −0.41 - 2.47 (0.03) | |
2 | FV, J | 0 | 0.09 | 0 | - | |
2 | FV, S | 1.75 | 0.38 | 4.56 | −3.39 - 10.29 | |
2 | J, S | 0 | 0.75 | 0 | - | |
Total | 4.38 | 4.50 | 0.96 | |||
3 | PA, TV, FV | 1.75 | 0.54 | 3.22 | −1.27 - 7.68 (0.02) | |
3 | PA, TV, J | 1.75 | 1.06 | 1.66 | −0.65 - 3.96 (0.01) | |
3 | PA, TV, S | 1.75 | 4.36 | 0.40 | −0.16 - 0.96 | |
3 | PA, FV, J | 0 | 0.24 | 0 | - | |
3 | PA, FV, S | 1.75 | 0.98 | 1.78 | −0.71 - 4.29 (0.03) | |
3 | PA, J, S | 1.75 | 1.91 | 0.92 | −0.36 - 2.20 (0.01) | |
3 | TV, FV, J | 0 | 0.79 | 0 | - | |
3 | TV, FV, S | 4.39 | 3.28 | 1.34 | −0.16 - 2.58 (0.19) | |
3 | TV, J, S | 6.14 | 6.37 | 0.96 | 0.31 - 1.20 (0.03) | |
3 | FV, J, S | 0.88 | 1.44 | 0.61 | −0.60 - 1.82 (0.14) | |
Total | 20.18 | 20.99 | 0.96 | |||
4 | PA, TV, FV, J | 1.75 | 2.04 | 0.86 | −0.38 - 2.06 (0.024) | |
4 | PA, TV, FV, S | 5.26 | 8.39 | 0.63 | 0.13 - 1.12 (0.001) | |
4 | TV, FV, J, S | 10.53 | 12.26 | .86 | 0.39 - 1.33 (0.012) | |
4 | PA, FV, J, S | 0.88 | 3.68 | 0.24 | −0.23 - 0.71 (0.12) | |
4 | PA, TV, J, S | 17.54 | 16.30 | 1.08 | 0.64 - 1.51 | |
Total | 35.96 | 42.67 | 0.84 | |||
5 | PA, TV, FV, J, S | Total | 36.84 | 31.37 | 1.17 | 0.89 - 1.46 |
Data collected from high school students in Oahu, Hawaii. PA: Physical inactivity; TV: Excessive television time; FV: Inadequate fruit and vegetable consumption; J: Excessive junk food consumption; S: Inadequate sleep.
Without common behavioral assessments, comparison of current results to previous research is difficult. Unlike the comparison studies, the current study did not measure risk factors such as tobacco or alcohol use. However, based on our sample and common behaviors assessed, the disparity between Filipino-American adolescents living in Hawaii and their Canadian, Australian and American counterparts is apparent. The current study is the first investigation among this highly understudied and unique population. Additional investigations among similar youth populations are warranted.
LimitationsCurrent study limitations are noteworthy. First, given the small sample size, results may not generalize to the greater population of Filipino-American adolescents; however, current results may be indicative of Filipino- American adolescents living in Hawaii. Second, the lack of variability within the sample limited the likelihood of uncovering statistically significant behavioral clusters. More specifically, the majority of the current sample (approx. 93%) was at risk for three or more risk factors, limiting the ability to detect patterns of specific behavior combinations that are most problematic.
These results suggest a need for future research among similar minority populations. The high prevalence rates for single and multiple health risk behaviors we report indicate a need for future population-level examinations. To inform development of effective behavioral interventions, future research should examine the social and physical environments contributing to the co-occurrence of health risk behaviors.
This research was funded by the National Cancer Institute R25 CA 90956.