We evaluated the differences in higher systolic blood pressure (SBP) between boys and girls in a longitudinal adolescent cohort. Results showed that although boys have a 19% relative increase in the risk of higher SBP annually between the ages of 13 and 17 years, girls do not experience the same risk increase. Among both boys and girls, being overweight increases the likelihood of high SBP almost 3-fold, and every 5 additional hours of "screen time" (television viewing, video game playing, Internet use) is associated with a 4% relative increase in the likelihood of higher SBP levels. For every 5 additional sports-related or other active behaviors over a 7-day recall, there was an 8% relative reduction in high SBP risk. Annual blood pressure assessment in teens that continues into adulthood may facilitate early detection of adult hypertension. Weight control, increased physical activity, and reduced screen time may reduce the likelihood of higher SBP levels developing in adolescents.
Hypertension is more prevalent among men than among women in young and middle-aged adults.1 Are there adolescent antecedents of this sex difference in adult hypertension? Higher systolic blood pressure in adolescence is associated with higher systolic blood pressure later in life; therefore, it may be a precursor of adult hypertension.2 Cross-sectional studies in adolescents demonstrate that higher systolic blood pressure is more common among boys than among girls, even with the use of sex-specific diagnostic thresholds.3,4 Potential boy-girl differences in higher systolic blood pressure risk during adolescence have not, to our knowledge, been previously examined in a multivariate longitudinal cohort analysis. We used a longitudinal adolescent cohort to evaluate sex differences in the risk of higher systolic blood pressure levels and the effect of weight and physical activity on this risk.5
Subjects and methods
The study cohort included 1,267 grade 7 students from 10 secondary schools in the Montreal area as part of the Nicotine Dependence in Teens study.6 Blood pressure, resting heart rate, weight, and height were assessed at 2-year intervals for 5 years. Physical activity and sedentary behavior were assessed by a questionnaire (Table), as were alcohol use (categorized as < or ≥ once per month in our analysis), tobacco use (categorized as yes/no for the recall period), and language spoken at home. The median household income of students by school was available from a government report.
In addition to a descriptive analysis of the study variables, we constructed multivariate models with the presence or absence of higher systolic blood pressure as the outcome variable. The multivariate statistical modeling method that we used was generalized, estimating equations for logistic regression. This method allowed us to include data from more than 1 assessment for each participant: the model "knew" that these data points were "linked" with one another.
We first constructed separate models for girls and boys. Potential determinants/covariates that we included in our models were age, overweight, physical activity, sedentary behavior, tobacco use, alcohol use, household income, linguistic background, and heart rate. We then constructed models using data from both boys and girls, and included the variable of sex in the model as a potential determinant of higher systolic blood pressure. In addition, we separately assessed interactions between sex and each of the other potential determinants/ covariates. In the final model that included all available data, we included interaction terms that were statistically significant and consistent with the sex-specific analyses.
A total of 1,267 participants were assessed in grade 7 (age 13), 954 in grade 9 (age 15), and 844 in grade 11 (age 17). Among the approximately 700 individuals who completed all 3 assessments, about half of those with high systolic blood pressure were boys at the 7th-grade and 9th-grade assessments, but by the 11th-grade assessment, boys accounted for 67% of those with higher systolic blood pressure levels. A total of 117 individuals (67 boys and 50 girls) had a high systolic blood pressure value recorded at some point during the course of the study, and approximately 60% of these youth had a high systolic blood pressure value on more than 1 occasion. Those who underwent all 3 blood pressure and anthropometric assessments were not substantively different from those who did not, and we were therefore comfortable in including all available data in our models.
In the girls-only model examining associations with higher systolic blood pressure, being overweight was more than twice as likely as the other covariates to be associated with higher systolic blood pressure, whereas physical activity participation reduced the likelihood of higher systolic blood pressure, with a 5-activity increment on 7-day recall associated with an almost 20% reduction in the likelihood of higher systolic blood pressure. In the boys-only model, each 1-year increment in age was associated with approximately a 20% increased likelihood of higher systolic blood pressure. Being overweight was associated with a 3-fold risk increase. Sedentary behavior was associated with increased risk of higher systolic blood pressure in the boys-only model: each 5-hour increment per week in "screen time" (television viewing, video game playing, Internet use) was associated with a 6% increased likelihood of higher systolic blood pressure.
When data from boys and girls were combined, we confirmed a significant interaction between sex and age, as suggested by the sex-specific models. Interactions between sex and physical activity, sex and being overweight, and sex and sedentary behavior, however, were not statistically significant. In the adjusted analysis, boys were progressively more likely than girls to have higher systolic blood pressure levels as they progressed through adolescence, with an approximately 2-fold risk increase by grade 9 and a 3-fold risk increase by grade 11, even after accounting for overweight, sedentary behavior, physical activity, resting heart rate, tobacco and alcohol use, household income, and first language. Using the model that included both boys and girls and an interaction term for sex and age, being overweight was associated with a 2.6-fold greater likelihood of higher systolic blood pressure; an increment of 5 hours per week of screen time was associated with a 4% increased likelihood of higher systolic blood pressure; an increment of 5 physical activities over the 7-day recall period was associated with an 8% reduction in the likelihood of higher systolic blood pressure.
Our analysis of a longitudinal adolescent cohort confirms the presence of a sex difference in the risk of higher systolic blood pressure levels during adolescence. Further, the findings indicate that the risk difference between boys and girls increases as they get older. This difference could not be explained by sex differences in the effect of weight, physical activity, sedentary behavior, or alcohol and tobacco use. What actually accounts for this sex difference, however, remains unclear.
The emergence of a sex difference in higher systolic blood pressure levels during adolescence and the erosion—and perhaps even reversal—of a sex difference in hypertension prevalence previously reported among older adults7 suggests that sex steroids or their receptors may affect blood pressure. Supporting this possibility, follow-up data from the Medical Research Council National Survey of Health and Development Cohort suggests that early puberty may increase the risk of adult hypertension among boys but not among girls.8 Unlike girls and young women who may seek medical attention for contraception, pregnancy, and Pap testing, older adolescent boys and younger men are less likely to consult a physician or nurse.9 Detection of higher blood pressure levels in boys during adolescence may be a signal to strongly recommend at least annual assessment so that adult hypertension can be diagnosed early.
Consistent with several previous studies, we identified being overweight as a critical determinant of higher blood pressure levels in both boys and girls. This underscores the importance of counseling teens and their parents to make prudent dietary choices, with avoidance of energy-dense alternatives. A key finding of our analysis was that independent of weight, engaging less frequently in physically active behaviors and clocking more hours of screen time were identified as having independent and adverse effects on the risk of systolic blood pressure. For every additional 5 active behaviors that teenagers engaged in for 5 minutes or more over the 7-day recall period, they experienced an 8% relative reduction in the likelihood of higher systolic blood pressure levels. For every 5 additional hours of screen time per week, they experienced a 4% relative risk increase for higher systolic blood pressure levels. The importance of physical activity and physical activity education in children and teenagers is receiving increasing public attention. Our results suggest that moderating screen time may be of equal importance. A previous clinical trial demonstrated that an intervention that reduced television viewing time in children had a favorable effect on body mass index.10 Our results support the possibility that such an intervention could also have a beneficial effect on blood pressure levels.
The higher prevalence of hypertension in men appears to have roots in adolescence, with teen boys increasingly likely to develop higher systolic blood pressure levels as they get older. Detection of higher systolic blood pressure levels in teens suggests the need for continued blood pressure surveillance for the early detection of adult hypertension. The importance of prudent dietary habits and weight control must be emphasized. In addition, increased exercise and participation in sports along with reduction in screen time may have beneficial effects on blood pressure and ultimately on cardiovascular health.
Dr Kaberi Dasgupta is a physician scientist who holds a Canadian Institutes of Health Research New Investigator Award. She gratefully acknowledges the contributions of the coauthors of the original study report: Jennifer O'Loughlin, Shunfu Chen, Igor Karp, Gilles Paradis, Johanne Tremblay, Pavel Hamet, and Louise Pilote. Funding for the collection of data used in this analysis was through an operating grant awarded to Dr Jennifer O'Loughlin by the National Cancer Institute of Canada, with funds from the Canadian Cancer Society. Funding for the analysis (programmer) was through an interdisciplinary capacity enhancement grant funded by the Canadian Institutes of Health Research and the Heart and Stroke Foundation of Canada awarded to Dr Louise Pilote and GENESIS (Gender and Sex Determinants of Cardiovascular Disease), a pan-Canadian research group concerned with sex differences in the development, symptoms, and treatment of cardiovascular disease.