Abstract
Objective:
Our objective was to examine racial/ethnic disparities in injury-related risk behaviors and deaths among teens in Missouri, with a focus on Hispanic people—the fastest-growing racial/ethnic group in the state.
Methods:
We used data from the 2013 Missouri Youth Risk Behavior Survey, which included 1616 students in grades 9 through 12 from 32 public and charter high schools. The overall response rate was 69%. We compared the prevalence of 10 injury-related risk behaviors among racial/ethnic groups and used multivariate logistic regression models to control for respondent age and sex. Using data from the 2000-2014 Missouri death records, we also compared injury-related death rates among racial/ethnic groups of teens aged 15 to 19 years.
Results:
Hispanic students had a significantly higher prevalence than non-Hispanic white students for 9 of 10 risk behaviors and a significantly higher prevalence than non-Hispanic black students for 6 of the 10 risk behaviors included in the study. However, Hispanic teens aged 15 to 19 years had a significantly lower death rate from suicide, homicide, and unintentional injury combined (39.8 per 100000 population, 95% confidence interval [CI], 32.2-48.5) when compared with their non-Hispanic white (54.3 per 100000 population, 95% CI, 52.3-54.6) or non-Hispanic black (94.1 per 100000 population, 95% CI, 87.9-100.3) counterparts.
Conclusion:
Injury-related risk behaviors were more prevalent among Hispanic students than non-Hispanic white or non-Hispanic black students. Further efforts are needed to understand risk behaviors among Hispanic teens to guide intervention efforts.
According to US population estimates, the total Hispanic population in Missouri was 232954, or 3.5% of the state population in 2013, which was substantially lower than the proportion of Hispanic population in the United States (16.3%) during the same year. However, Hispanic people are the fastest-growing racial/ethnic group in Missouri. From 2000 to 2013, the number of Hispanic people in Missouri increased from 118592 to 232954, a 96% increase. By contrast, the non-Hispanic population in Missouri increased by only 6%, and the Hispanic population in the United States increased by 53% during the same period. 1
In 2013, 41% of the Hispanic population in Missouri was ≤19 years of age, compared with 25% of the non-Hispanic population. 1 Therefore, to find opportunities to improve the Hispanic population’s health and reduce health disparities, it is especially important to monitor the health status and risk behaviors among young Hispanic people.
In 2012, injuries accounted for 190 of the 243 (78.2%) Missouri teens aged 15 to 19 years who died during the year. Of these 190 injury deaths, 73 (38.4%) were caused by motor vehicle crashes, 48 (25.3%) by homicides, 41 (21.6%) by suicides, and 27 (14.2%) by all other injuries and adverse effects. 2 Despite having lower socioeconomic status on average than non-Hispanic white people, Hispanic people often have mortality rates that are comparable to, or in some cases lower than, those of non-Hispanic white people. Some ascribe this “Hispanic paradox” to behaviors and traditions in the Hispanic culture, whereas others attribute this phenomenon to underreporting of Hispanic ethnicity on death records. 3
The objective of this study was to describe the burden of, and identify racial/ethnic disparities in, injury-related risk behaviors and deaths among high school students in Missouri, with a focus on Hispanic teens.
Methods
Data Source
The Missouri Youth Risk Behavior Survey (YRBS) monitors behaviors of public high school students (primarily aged 13-18 years) in 6 categories identified by the Centers for Disease Control and Prevention as the leading causes of morbidity and mortality among young people 4 : (1) behaviors that contribute to unintentional injuries and violence, (2) sexual behaviors that contribute to unintended pregnancy and sexually transmitted diseases, (3) alcohol and other drug use, (4) tobacco use, (5) unhealthy dietary behaviors, and (6) inadequate physical activity. The YRBS is a paper-and-pencil survey that is conducted in odd-numbered years nationally and in selected states, including Missouri. The survey is distributed in randomly selected high schools with randomly selected classes of students in grades 9 through 12. Details of the sampling design and response rates for the 2013 Missouri YRBS are described elsewhere. 5 The 2013 Missouri YRBS included 1616 students in grades 9 through 12 from 32 public and charter high schools. Response rates were 80% (32 of 40) for sampled high schools and 87% (1616 of 1864) for sampled high school students, with an overall response rate of 69%. 5 Data were weighted to represent students in public or charter high schools in Missouri. The Missouri Department of Health and Senior Services Institutional Review Board approved the 2013 Missouri YRBS.
The 2013 Missouri YRBS had 17 questions on injury-related risk behaviors. We included in our analysis the 10 questions that had sample sizes of at least 100 for each of 3 racial/ethnic groups (non-Hispanic white, non-Hispanic black, and Hispanic). The only exception was for “having been physically hurt by someone they were dating,” which had a sample size of 97 for non-Hispanic black students.
We determined race/ethnicity using 2 questions: “Are you Hispanic or Latino?” and “What is your race?” We classified students as Hispanic/Latino (hereinafter Hispanic) if they answered “yes” to the first question, regardless of race. Students who answered “no” to the first question and only “black or African American” to the second question were referred to as non-Hispanic black. Students who selected only “white” to the second question were referred to as non-Hispanic white. Students who selected any other race were referred to as non-Hispanic other.
We used data from death records to compare injury-related deaths among non-Hispanic white, non-Hispanic black, and Hispanic teens aged 15 to 19 years in Missouri and the United States. We obtained death rates and 95% confidence intervals (CIs) for suicide; homicide; motor vehicle traffic; and suicide, homicide, and unintentional injury combined in Missouri and the United States during 2000 to 2014 from the Centers for Disease Control and Prevention’s WONDER (Wide-Ranging Online Data for Epidemiologic Research). 6
Statistical Analysis
We calculated the prevalence of the 10 injury risk behaviors for each racial/ethnic group and compared them using the Wald χ 2 test for significance. We used multivariate logistic regression to control for respondent age and sex. Hispanic race/ethnicity was the referent category for all analyses because the primary focus of this study was to compare Hispanic students with other racial/ethnic groups. We analyzed data using SAS 7 9.4 and accounted for the survey design effects in the analysis. We compared death rates with 95% CIs. If there was no overlap between 95% CIs, we determined the rates to be significant.
Results
The survey population comprised 76.6% non-Hispanic white, 15.8% non-Hispanic black, 4.1% Hispanic, and 3.4% non-Hispanic other students. The weighted distribution for all racial/ethnic groups in the survey population was similar to the enrollment percentages for female and male students in each grade provided by the Missouri Department of Elementary and Secondary Education. However, in the survey population, Hispanic students (27.7% vs 22.1%), and non-Hispanic black students (30.9% vs 23.0%) were overrepresented and non-Hispanic white students (22.3% vs 24.0%) were slightly underrepresented by seniors, compared with the enrollment population. In addition, the percentage of Hispanic male students in the survey population was higher than in the enrollment population (58.4% vs 50.1%; Table 1).
Characteristics of high school students in the Missouri Youth Risk Behavior Survey and student enrollment overall, by race/ethnicity, sex, and grade: 2013
Abbreviation: CI, confidence interval.
a2013 Youth Risk Behavior Survey population in Missouri.
b2013 Missouri public high school enrollment.
cThe survey data were weighted by race, sex, and grade to represent the student population. Survey design effects including the weight were taken into account when calculating the weighted number and percentage. The weighted numbers and percentages by race, sex, and grade are not exactly the same as those in the enrollment because of missing values in these variables. For the same reason, the numbers in subcategories do not sum to the total.
dIncludes American Indian / Alaska Native, Asian American, and Native Hawaiian / other Pacific Islander.
Compared with non-Hispanic white students, Hispanic students had a significantly higher prevalence for all but 1 of the 10 risk behaviors included in this study (ie, weapon-carrying behavior). Compared with non-Hispanic black students, Hispanic students had a significantly higher prevalence for 6 of the 10 risk behaviors: depression, the 3 suicide-related behaviors, weapon-carrying behavior, and riding with someone who had been drinking alcohol. In addition, Hispanic students had a significantly higher prevalence than non-Hispanic other students for 2 of the 10 risk behaviors. After adjusting for age and sex, all differences remained significant except for weapon-carrying behavior and riding in a car with a driver who had been drinking alcohol (Table 2).
Injury-related risk behaviors among high school students, by race/ethnicity: Missouri Youth Risk Behavior Survey, 2013
Abbreviations: CI, confidence interval; aOR, adjusted odds ratio.
aRespondents who answered each question. The number of respondents in each racial/ethnic group may not sum to the total number because of missing responses to the race and/or ethnicity questions.
bPercentages take into account survey design effects, including weights.
cBased on Wald χ 2 tests using univariate logistic regressions with Hispanic as the reference group. P < .05 is considered significant.
dAdjusted for age and sex.
eIncludes American Indian / Alaska Native, Asian American, and Native Hawaiian / other Pacific Islander.
The 2000-2014 Missouri death records data showed that compared with non-Hispanic white teens aged 15 to 19 years, Hispanic teens aged 15 to 19 years had a significantly lower death rate from motor vehicle traffic (16.8 per 100000 population, 95% CI, 12.0-22.8, vs 33.3 per 100000 population, 95% CI, 31.7-34.9), and compared with non-Hispanic black teens, Hispanic teens had a significantly lower death rate from homicide (10.2 per 100000 population, 95% CI, 6.6-15.1, vs 63.6 per 100000 population, 95% CI, 58.5-68.7). The difference in the suicide death rate among the 3 racial/ethnic groups in Missouri was not significant. Overall, Hispanic teens had a significantly lower death rate from suicide, homicide, and unintentional injury combined as compared with non-Hispanic white and black teens. The pattern from the US data was similar, except that Hispanic teens had a significantly lower suicide death rate than that of non-Hispanic white teens in the United States (5.4 per 100 000 population, 95% CI, 5.2-5.6, vs 8.4 per 100 000 population, 95% CI, 8.3-8.5; Table 3).
Injury-related death rates among teens aged 15-19 years: Missouri and the United States, 2000-2014a
Abbreviation: CI, confidence interval.
aData source: US Centers for Disease Control and Prevention. 6
bPer 100 000 population.
Discussion
This study is the first to include data on most intentional and unintentional injury-related behaviors among young people in Missouri. In this study, Hispanic teens had a significantly higher prevalence of injury-related risk behaviors than their counterparts from other racial/ethnic origins. These risk behaviors could lead to motor vehicle crashes, violence, and suicide. However, Hispanic teens had a significantly lower death rate from suicide, homicide, and unintentional injuries combined as compared with their non-Hispanic white or black peers.
Our findings are consistent with literature on risk behaviors among Hispanic young people. Using data from the 2011 national YRBS, Eaton et al found that the prevalence of riding with a driver who had been drinking alcohol, being injured in a physical fight, experiencing dating violence, feeling sad or hopeless, and attempting suicide was significantly higher among Hispanic students than among non-Hispanic white students. 8 A report from the Colorado Department of Public Health and Environment also showed that a significantly higher percentage of Hispanic (12.5%) than non-Hispanic white (5.1%) high school students rarely or never wore a seat belt while riding in a car driven by someone else and a significantly higher percentage of Hispanic (30.9%) than non-Hispanic white (22.3%) students felt sad or helpless for 2 or more weeks in 2009. 9
We also found similar racial/ethnic disparities in injury-related death among young people in the United States. Hispanic young people have a significantly lower death rate from suicide, homicide, and unintentional injury combined as compared with non-Hispanic white or black young people. One possible explanation for the lower death rates among Hispanic teens is that Hispanic ethnicity is underreported in death records 10 ; such underreporting could lead to inaccurate mortality rates for Hispanic people, including teens. Therefore, if only death data were used in public health assessments and planning, public health professionals would see only a partial picture and would miss the information on risk behaviors and related injuries.
These risk behaviors could exact a high morbidity cost, leading to physical and emotional distress from injuries, emergency department visits, hospitalizations, and missed school days. However, little is known about racial/ethnic disparities in nonfatal injuries among teens in Missouri and the United States. 11 The National Electronic Injury Surveillance System-All Injury Program, a nonfatal injury surveillance system, provides national estimates of injuries treated in US hospital emergency departments. Although the system has a large overall sample size, the proportion of the sample with missing data on race/ethnicity is high: 22.8% for self-harm, 19.6% for assault, 23.1% for transportation-related injury, and 21.3% for all injury among teens aged 15 to 19 years. Therefore, race/ethnicity-specific rates are not calculated in WISQARS (Web-based Injury Statistics Query and Reporting System). 11
Limitations
This study had several limitations. First, the YRBS is based on self-report, and the extent of under- or overreporting of behaviors cannot be determined. Second, the sample sizes for Hispanic and non-Hispanic black students in the Missouri YRBS were relatively small (<150) for the risk behaviors included in this study and were too small (<100) to generate stable prevalence estimates for the injury-related risk behaviors not included, such as texting, e-mailing, or talking on a cellular telephone while driving. Therefore, this study captured only a partial picture of disparities in injury-related risk behaviors among Missouri teens. Finally, Hispanic students in the survey were overrepresented by seniors and males. In general, older age and male sex are associated with increased risk behaviors among teens. However, after adjusting for age and sex, almost all differences remained significant. Therefore, this overrepresentation of seniors and males was unlikely to affect the validity of this study. Data from the national YRBS and Colorado YRBS also showed a high percentage of Hispanic teens engaging in risk behaviors that could lead to intentional and unintentional injuries, 7,8 suggesting that we should continue to monitor these disparities and further explore their underlying causes and effects on health.
Conclusion
A higher percentage of Hispanic students than students from other racial/ethnic groups in Missouri engaged in behaviors that may lead to motor vehicle crashes, violence, or suicide. We also found a paradox between racial/ethnic disparities in injury-related deaths and risk behaviors. High-quality data should be collected and analyzed to better understand these disparities. Further efforts are needed to understand risk behaviors among Hispanic young people to guide intervention efforts.
Footnotes
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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was partly supported by the Preventive Health and Health Services Block Grant, cooperative agreement 2B01OT009030-15, between the Centers for Disease Control and Prevention and the Missouri Department of Health and Senior Services. The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention.