TL;DR
This study uses advanced statistical modeling on a large health check-up dataset from South China to reveal nuanced age-related trends in high blood pressure (BP) risk, differentiating between systolic (SBP) and diastolic (DBP) pressure and by sex. The findings suggest that SBP risk increases more rapidly in women during their early adult life and highlight a potential need for earlier hypertension screening in China, particularly for individuals over 40.
What Problem Does This Paper Address?
High blood pressure (hypertension) is a major contributor to mortality and cardiovascular diseases globally. While it's generally understood that blood pressure increases with age, the precise nature of this relationship, especially how it differs for systolic versus diastolic pressure and between men and women, remains underexplored. Previous studies often simplified age into discrete groups, potentially masking finer trends, and lacked sophisticated modeling to capture complex, non-linear relationships. This paper aims to provide a more detailed understanding of age-related BP risk patterns and how these trends have evolved over time in specific age groups, which is crucial for effective prevention strategies.
Methods
The researchers employed Generalized Additive Models (GAMs), specifically Generalized Additive Models with interaction terms (GA2M), to model the relationship between age and the risk of high SBP/DBP. GAMs are powerful because they allow for flexible, non-linear relationships between predictor variables (like age) and the outcome (risk of high BP), unlike traditional linear models that assume a straight-line relationship. This approach is particularly useful for capturing complex biological processes where risk doesn't necessarily increase uniformly. They also used standard logistic regression for comparison and evaluated model performance using the Receiver Operating Characteristic Area Under the Curve (ROCAUC). To assess temporal trends in high BP prevalence within different age groups, they used variance analysis and the Mann–Kendall test.
Datasets
The study utilized health check-up data from Guangdong Second Provincial General Hospital in South China. The dataset comprises an 8-year observational period (January 1, 2011, to December 31, 2018), with data from January 1, 2015, to December 31, 2018, used for the main analysis and earlier data for model validation. A total of 71,468 participants aged over 18 with complete information on weight, height, age, gender, glucose, triglyceride, total cholesterol, systolic blood pressure (SBP), and diastolic blood pressure (DBP) were included. This is a large, real-world dataset from routine health examinations.
Key Insights and Ideas
- Sex-Specific SBP Trajectories: The risk of high SBP doesn't follow the same path for men and women. Women experience a more rapid increase in SBP risk earlier in life (around age 42) compared to men, who show a more linear rise with age.
- DBP's U-Shaped Trend: Unlike SBP, the risk of high DBP shows an inverse U-shaped trend, increasing initially and then declining after middle age (around 50-65 years).
- Transition Points Matter: Identifying "transition points" where the risk associated with factors like age, BMI, glucose, and triglycerides changes significantly can inform targeted interventions. For instance, a BMI of 20-25 kg/m² appears to be a threshold for SBP risk.
- Temporal Trends in Middle Age: The study observed a significant increasing trend in high SBP risk among middle-aged individuals (40-59 for men, 30-39 for women) over the past decade, suggesting a potential shift towards younger onset of hypertension.
- Implications for Screening: The findings suggest that current hypertension screening guidelines in China, which often focus on individuals over 45, might need to be re-evaluated, with a potential need for earlier screening (after age 40) due to the observed age-related risk progression.
Key Results
- Prevalence: Out of 71,468 participants, 12.05% had high SBP and 6.78% had high DBP.
- SBP Risk: Risk of high SBP continuously rose from age 35 to 79. For men, this rise was linear with age; for women, it was non-linear, with a steeper increase in early adulthood.
- DBP Risk: Risk of high DBP increased concurrently with SBP but declined after age 50-65.
- Temporal Trends: An increasing trend of high SBP risk was observed in men aged 40-49 and 50-59 (p < 0.003) and women aged 30-39 (p = 0.008) during the study period. High DBP risk decreased in men aged 30-39 (p = 0.004) and women aged 40-49 (p = 0.005).
- Model Performance: The GA2M model achieved a ROCAUC of 0.837, slightly outperforming the standard logistic regression model (0.803).
Limitations and Future Work
- Single Center Data: The study relied on data from a single center, which may limit the generalizability of findings to other populations.
- One-Off BP Measurement: Blood pressure measurements were taken once per participant, whereas multiple measurements over different days are recommended for more robust BP assessment.
- Future Work: The authors suggest that future research should investigate the generalizability of these findings to other populations and explore the underlying reasons for the observed gender discrepancies in BP trends, potentially involving hormonal, lifestyle, and social factors. They also highlight the need for improved hypertension awareness, prevention, and treatment strategies, particularly for middle-aged adults in China.
Summary updated