**4. Discussion**

A large proportion (29%) of participants had elevated blood pressure in this study of Hispanic adolescents, which was more common among males and those with chronic disease indicators (AN, overweight or obesity, and abdominal obesity). This is consistent with prior studies characterizing hypertension in children and adolescents [43,44]. A major strength of this study is that it is among the first to examine the potential impact of working on blood pressure in adolescents, specifically Hispanics. Working was a significant predictor of elevated blood pressure after adjusting for potential confounders at baseline. In addition, despite not being statistically significant, the mean probabilities of elevated blood pressure were higher for adolescents who were working. This association is also consistent with studies of adult working populations [24–28]. A detailed comparison with literature focusing on adults is difficult given differences in data collection and study variable definitions. The mechanism of how work exposures influence blood pressure could not be examined in this study. However, some of the participants in the study reported long work hours in addition to other demands related to school and home responsibilities. Insufficient time for engaging in healthy and stress relieving behaviors may be a factor. As an example, increased stress and limited time can negatively impact sleep quality and quantity. A recent systematic review indicated evidence of a link between sleep parameters and obesity, hypertension, and insulin sensitivity, but there were an insufficient number of high-quality studies available for assessing causality [45]. With respect to psychosocial job exposures, a study of Brazilian adolescent workers evaluated a connection between job demands, job control, and social support at work on various health indicators. The researchers found associations between psychological job demands and reduced sleep quantity during the week as well as work injury and body pain [46]. In the present study, self-reported sleep quantity was not associated with elevated blood pressure. The lack of association could be due to under or misreporting of sleep parameters since the data were self-reported. Furthermore, the outcome was elevated blood pressure rather than diagnosed hypertension. In addition, work-related as well as general stress may increase unhealthy coping behaviors in adolescents such as excessive food consumption, consuming foods higher in sugar and saturated fat, or consuming alcohol [30–34]. Many of these unhealthy coping mechanisms are associated with hypertension and coronary heart disease [47,48].

Additional strengths of this study included using an existing migrant education program infrastructure to access a hard-to-reach population [36]. Partnering closely with the school administrators and staff allowed for maintaining contact over time with a young, Hispanic population who are clearly at risk for chronic disease. Additional strengths included response proportions that were over 80% at baseline and during each assessment. Over 90% of eligible participants continued to participate in each follow-up assessment.

The key limitations of this study include a low prevalence of students currently engaged in work. While the use of pre-existing data made it possible to examine an under-researched topic with few resources, the low number of current workers also could have restricted the study's statistical power, while also prohibiting the examination of specific types of work and work-related exposures, both physical and psychosocial. Examples of areas of interest for future studies include increased

time demands and long work hours combined with other non-work demands on time, psychosocial job stress, and physical job exposures. Future research is needed that addresses other populations beyond Hispanics and that which is designed and powered specifically to examine specific work exposures within the context of nutritional and other risk factors. Another important limitation is a high rate of attrition due to school drop-out, relocating to another school district, and early graduation. Approximately 25% of those in the baseline assessment were no longer eligible to participant in the second follow-up assessment. A subsequent pilot study examined the impact of this loss to follow-up. Blood pressure levels were higher in those who dropped out; however levels were not significantly different when compared to those who remained in the study [36]. This loss to follow-up may have contributed to only a minimal average difference in elevated blood pressure levels comparing those currently working with those who were not in the follow-up assessments. Beyond the potential impact of loss to follow-up, the reason for the lack of an association across all years is not clear. It also could be that participants learned better skills for coping with stress and time demands as they aged. The variables collected for this study along with the sample size were not sufficient for examining these relationships further. The study design also did not allow diagnosis of high blood pressure in any single year. Guidelines recommend observing elevations in blood pressure on three separate occasions prior to making a diagnosis [10]. Finally, this study did not include an in-depth dietary assessment, which is a limitation given the known associations between diet and elevated blood pressure. This would be of particular concern if the working students ate a poorer diet than non-working students.
