ABSTRACT
Background
Evidence regarding the impact of age at hypertension onset on brain atrophy, independent of hypertension duration, remains limited. This study investigated whether the association between hypertension and brain volume differs by age at hypertension diagnosis.
Methods
In a multicenter, community‐based cohort study (initiated in 2006), 948 participants were included: 117 early‐onset hypertensives (diagnosed ≤40 years), 354 late‐onset hypertensives (diagnosed >40 years), and 477 non‐hypertensive controls, matched through propensity score matching. Neuroimaging data have been collected since 2020 to assess brain volume. The associations between early‐onset and late‐onset hypertension with brain volume were evaluated using the voxel‐wise and generalized linear models.
Results
Compared to controls, early‐onset hypertension was associated with lower volumes of cerebral parenchyma (β = −0.302; 95% confidence interval [CI], −0.541 to −0.063) and gray matter (β = −0.338; 95% CI, −0.590 to −0.087). The frontal, occipital, and temporal lobes showed the most prominent volume loss. Compared to participants with late‐onset hypertension, early‐onset hypertensives exhibited more pronounced brain volume reductions, especially in the frontal lobe. In contrast, compared with the matched non‐hypertensive controls, the pattern of brain volume reduction in the late‐onset hypertension group was similar to that in the early‐onset hypertension group at the voxel level but did not reach statistical significance after full adjustment for covariates.
Conclusions
There is a significant correlation between early‐onset hypertension and brain atrophy. It is crucial to manage blood pressure at a young age. For patients diagnosed with hypertension before the age of 40, it is recommended to strengthen brain health monitoring while undergoing antihypertensive treatment.
Keywords: brain health, cohort study, hypertension, neuroimaging
1. Introduction
Hypertension is the leading risk factor for global mortality, causing 10.8 million deaths in 2019, accounting for 19.2% of the total global deaths [1, 2]. It also ranks as a major risk factor for the global disease burden among people over 50 years old, and the secondary risk factor among those aged 25–49 years [1]. Notably, the so called aging diseases are increasingly prevalent in younger populations, with hypertension being a prime example. But young people often overlook the health hazards of hypertension. A study showed that the awareness, treatment, and control rates of hypertension were substantially lower among younger adults (31.7%, 24.5%, and 9.9% for those aged 35–44 years) compared to older generation (58.6%, 52.8%, and 18.4% for those aged 65–74 years) [3]. Therefore, attention should be paid to the younger generation of hypertension.
The association between hypertension and cardiovascular disease has been clearly confirmed, with two‐thirds of strokes and nearly half of coronary heart diseases attributed to hypertension [4]. In the Framingham study and the study of Coronary Artery Risk Development in Young Adults (CARDIA), the correlation between early‐onset hypertension and organ damage [5, 6] and cardiovascular disease mortality [3] mediated by hypertension was greater than that of late‐onset hypertension. Wang et al. [7] also demonstrated that compared to late‐onset hypertension, early‐onset hypertension was significantly associated with a higher risk of all‐cause mortality. It can be seen that the impact of early‐onset hypertension on the risk of vascular disease cannot be ignored. In addition to obvious vascular events in clinical practice, early exposure to hypertension can also cause damage to the structure of the brain. With the widespread application of magnetic resonance imaging (MRI) in research, a growing number of studies have examined the association between hypertension and brain volume, but the results are inconsistent [8, 9, 10]. A recent report suggested that hypertension in the 15–24 years prior to neuroimaging examination was most correlated with current brain volume [8]. Data from the UK Biobank showed a positive correlation between blood pressure (BP) in midlife and the severity of white matter hyperintensity (WMH), particularly in patients with hypertension diagnosed before the age of 50 [9]. Hypertension from early to middle adulthood seemed to be associated with an increase in WMH and reduced brain volume between the ages of 69 and 71 [10]. These studies suggest that age plays an important role in the association between hypertension and brain volume, but whether age at the time of diagnosing hypertension is related to brain volume remains to be explored. Although prior studies have explored the association between BP exposure and brain structure, few have evaluated how the age at hypertension diagnosis (rather than general BP elevation) relates to brain structural changes, particularly in large community‐based cohorts of the Chinese population. It is crucial for our understanding of the risk factors and progression of neurodegenerative diseases to check the degree of any possible association between hypertension and brain volume based on the age of onset. Therefore, we explored the relationship between early‐onset hypertension and late‐onset hypertension with brain imaging in a large community‐based Chinese cohort.
In this study, we evaluated whether there were differences in brain volume between early‐onset and late‐onset hypertensive populations. We hypothesized that early‐onset hypertension was associated with an increased risk of brain atrophy.
2. Methods
2.1. Study Subjects
Participants were derived from the Kailuan Study (KLS), a multicenter, long‐term follow‐up for 18 years, community‐based cohort study conducted on an adult population in the Kailuan region of Hebei Province, China. The participants mainly include employees of Kailuan Company Limited and local residents. Participants aged from 18 to 98 years were recruited from June 2006 [11, 12, 13]. Since 2020, the Multimodality Medical Imaging Study Based on Kailuan Research (META‐KLS), a subproject of KLS, has been recruiting participants to undergo brain MRI scans for neuroimaging analysis. The META‐KLS protocol has been previously reported [14]. In short, between 2006 and 2018, demographic questionnaires, clinical information, and laboratory test results were prospectively collected every 2 years. In addition, these participants have undergone a neuroimaging examination since 2020. As of September 2022, this study included 1195 participants.
Subsequently, participants who met all of the following inclusion criteria were included: (1) completed ≥3 follow‐up visits and collected clinical data; (2) received a brain MRI examination between 2020 and 2022; and (3) had no clinically diagnosed or self‐reported stroke, dementia, or neurological or psychiatric disorders. The exclusion criteria were as follows: (1) incomplete demographic data, (2) poor quality of neuroimaging or incomplete neuroimaging data, and (3) diagnosed with neurodevelopmental abnormalities or a history of cancer.
Finally, this study included a total of 948 participants.
2.2. Measurements of Clinical Features
Demographic questionnaires, clinical information, and laboratory examinations were prospectively collected every 2 years from 2006 to 2018, according to the standardized protocols from 11 local hospitals. Measurements of clinical features have been described in previously published protocols [11, 12, 13]. Clinical history was evaluated by trained doctors during their visits to the hospital.
2.3. Neuroimaging Data Acquisition
The neuroimaging data acquisition has been described in a previously published protocol [14, 15]. From 2020 to 2022, all participants underwent a 3.0T MRI (General Electric 750 W, Milwaukee, WI, USA) to obtain neuroimaging data. The scanning sequences included three‐dimensional brain volume (3D‐BRAVO) based on high‐resolution T1‐weighted imaging (T1WI) for brain macroscopic structure volume analysis, diffusion tensor imaging (DTI) for brain microscopic structure integrity analysis, 3D fluid‐attenuation reversal recovery (FLAIR) for WMH analysis, and T2‐weighted imaging (T2WI) and diffusion‐weighted imaging (DWI) for examining neurological dysplasia and stroke. The parameter settings for the corresponding sequences are reported in Table S1.
2.4. Neuroimaging Data Processing
The processing of neuroimaging data has been described in previously published protocol [14, 15]. The data were preprocessed using a standardized and validated pipeline. Neuroimaging features included the relative volume of brain macrostructure and brain microstructural integrity.
3D‐BRAVO‐T1WI was used to calculate the brain macroscopic structure volume. The brain MRI images were quantitatively analyzed for total brain volume (TBV), gray matter (GM) volume, white matter (WM) volume, and cerebrospinal fluid (CSF) volume, as well as the total intracranial volume (TIV). TBV was defined as the sum of the volumes of WM and GM. TIV was estimated as the sum of the volumes of TBV and CSF. The volumes were quantified using an automatic pipeline based on the Statistical Parametric Mapping software (Wellcome Trust Centre for Neuroimaging). We also calculated the hippocampal volume according to the AAL_90 atlas. The voxel levels of macrostructural volume were analyzed in subregions among groups using the FMRIB Software Library (FSL) v6.0.
2.5. Definition of Hypertension and Age of Onset of Hypertension
According to the Seventh Report of the Joint National Committee recommendation [16], hypertension was defined as systolic BP (SBP) ≥140 mmHg, diastolic BP (DBP) ≥90 mmHg, or the use of antihypertensive drugs. The age of onset of hypertension was defined as the age at which the first examination met the criteria for hypertension or self‐reported age. Participants with hypertension were categorized into early‐onset hypertension group (Group E, diagnosed with hypertension before the age of 40) and late‐onset hypertension group (Group L, diagnosed with hypertension after the age of 40) according to the age of onset to be consistent with the relevant literature [17, 18].
2.6. Statistical Analysis
The Kolmogorov–Smirnov test was performed to evaluate whether continuous variables were normally distributed. Continuous variables with a normal distribution were expressed as mean (standard deviation, SD), whereas non‐normal variables were described as median (interquartile range, IQR). Categorical variables were presented as absolute frequencies (relative percentages). To reduce the interference of confounding factors, propensity score matching (PSM) method (matching age and gender) was used among participants without hypertension to match the corresponding control group (Groups EC and LC) in a 1:1 ratio for Groups E and L. Comparisons between hypertensive participants and the control group were based on independent sample t‐test, Mann–Whitney U test, or chi‐square test.
The associations between early‐onset and late‐onset hypertension with the MRI markers of brain volume were evaluated using the generalized linear model, with corresponding non‐hypertensive participants as the reference group. Model 1 did not correct for any covariates. Model 2 was adjusted according to age, gender, smoking status, alcohol consumption, physical activity, history of diabetes, total cholesterol, triglycerides, high‐density lipoprotein, and low‐density lipoprotein. Model 3 was additionally adjusted for average BP level and duration of hypertension on the basis of Model 2.
The differences in GM volume were analyzed with a general linear model using FSL v6.0 at the global brain voxel level. The interaction effect of early‐onset and late‐onset hypertension and the presence or the absence of hypertension on brain volume was analyzed by a subtraction model. Age, gender, duration of hypertension, mean SBP, mean DBP, and TIV were used as covariates. Multiple comparison correction was performed using the Gaussian random field (GRF) method. Regions with P voxel < 0.005 and P cluster < 0.05 were considered significant.
All statistical analyses were performed using IBM SPSS Statistics 27.0 (IBM Corp., Armonk, NY, USA) and R 4.2.1 (R Development Core Team).
3. Results
A total of 948 participants aged between 25 and 75 were included for this study. The median (IQR) age was 56.00 (48.25, 65.00), and 473 (49.89%) participants were females. Table 1 summarizes the demographic, clinical characteristics, and neuroimaging metrics of included participants. Group E comprised 117 patients, and Group L comprised 354 patients. However, the remaining 714 participants without hypertension underwent 1:1 PSM, resulting in 117 matched patients in Group EC and 360 matched patients in Group LC. Compared with Group E, participants in Group L exhibited lower DBP (p < 0.05) but comparable SBP (p > 0.05), together with older age, an adverse cardiovascular risk profile, smaller brain volume, and larger CSF volume (all p < 0.05). Tables 2 and 3 show the demographic, clinical characteristics, and neuroimaging metrics of Groups E and EC and Groups L and LC, respectively. Compared with Group EC, Group E had higher SBP, DBP, and TG, a smaller brain volume, and a larger volume of CSF (all p < 0.05) (Table 2). Compared with Group LC, Group L was more likely to have smoking and drinking habits, an adverse cardiovascular risk profile, a smaller brain volume, and a larger volume of CSF (all p < 0.05) (Table 3).
TABLE 1.
Demographic, clinical characteristics, and neuroimaging metrics of participants.
| All (n = 948) | Group E (n = 117) | Group L (n = 354) | p a | |
|---|---|---|---|---|
| Age b , years | 56.00 (48.25, 65.00) | 43.00 (39.00, 53.00) | 60.00 (55.00, 68.00) | <0.001 |
| Female, no. (%) | 473 (49.89) | 40 (34.19) | 122 (34.46) | 0.957 |
| Smoking status, no. (%) | 510 (47.49) | 57 (48.72) | 188 (53.11) | 0.410 |
| Alcohol consumption, no. (%) | 693 (64.53) | 94 (80.34) | 233 (65.82) | 0.003 |
| Physical activity, no. (%) | 0.004 | |||
| Sometimes or seldom | 546 (50.84) | 45 (38.46) | 190 (53.67) | |
| Usually | 528 (49.16) | 72 (61.54) | 164 (46.33) | |
| History of diabetes, no. (%) | 208 (19.37) | 19 (16.24) | 96 (27.12) | 0.018 |
| Blood pressure, mmHg | ||||
| Systolic | 125.51 (117.59, 134.47) | 131.78 (123.78, 140.29) | 132.86 (125.56, 140.92) | 0.295 |
| Diastolic | 81.28 ± 8.07 | 86.99 ± 8.20 | 84.98 ± 6.63 | 0.008 |
| Total cholesterol, mmol/L | 4.90 (4.45, 5.46) | 4.80 (4.45, 5.39) | 5.04 (4.52, 5.53) | 0.023 |
| Triglyceride, mmol/L | 1.7 (1.06, 2.07) | 1.63 (1.19, 2.50) | 1.64 (1.24, 2.33) | 0.791 |
| High‐density lipoprotein, mmol/L | 1.44 (1.28, 1.65) | 1.38 (1.20, 1.54) | 1.42 (1.25, 1.61) | 0.074 |
| Low‐density lipoprotein, mmol/L | 2.65 (2.25, 3.04) | 2.58 ± 0.51 | 2.75 (2.35, 3.12) | 0.006 |
| Neuroimaging features | ||||
| TIV, mL | 1502.58 (1403.89, 1601.34) | 1552.52 ± 158.01 | 1512.49 ± 139.36 | 0.010 |
| Relative brain macrostructural volume, % of TIV | ||||
| Cerebral parenchyma | 73.37 (70.35, 75.93) | 74.04 ± 3.58 | 71.34 ± 3.96 | <0.001 |
| Gray matter | 33.31 ± 2.11 | 40.29 ± 2.34 | 38.65 ± 2.41 | <0.001 |
| White matter | 33.46 (32.01, 34.81) | 33.93 (32.79, 35.18) | 32.70 ± 2.16 | <0.001 |
| Cerebrospinal fluid | 26.46 (24.00, 29.43) | 25.81 ± 3.52 | 28.43 ± 3.84 | <0.001 |
| Hippocampus | 0.24 (0.23, 0.26) | 0.24 ± 0.02 | 0.24 (0.22, 0.25) | <0.001 |
Note: Group E, early‐onset hypertension group; Group EC, control group of early‐onset hypertension. Values are presented as mean (SD), median (IQR), or no. (%).
Abbreviation: TIV, total intracranial volume.
Comparisons between Groups E and L were based on independent t‐test, Mann–Whitney U test, or chi‐square test.
Age (in years) was calculated at the time of MRI acquisition.
TABLE 2.
Demographic, clinical characteristics, and neuroimaging metrics of participants in Groups E and EC.
| Group E (n = 117) | Group EC (n = 117) | p a | |
|---|---|---|---|
| Age b , years | 43.00 (39.00, 53.00) | 43.00 (39.00, 53.00) | 0.882 |
| Age of onset, years | 36.00 (31.00, 38.00) | — | |
| Female, no. (%) | 40 (34.19) | 41 (35.04) | 0.891 |
| Smoking status, no. (%) | 57 (48.72) | 58 (49.57) | 0.896 |
| Alcohol consumption, no. (%) | 94 (80.34) | 83 (70.94) | 0.094 |
| Physical activity, no. (%) | 0.234 | ||
| Sometimes or seldom | 45 (38.46) | 54 (46.15) | |
| Usually | 72 (61.54) | 63 (53.85) | |
| History of diabetes, no. (%) | 19 (16.24) | 11 (9.40) | 0.118 |
| Blood pressure, mmHg | |||
| Systolic | 131.78 (123.78, 140.29) | 119.67 ± 8.87 | <0.001 |
| Diastolic | 86.99 ± 8.20 | 77.37 ± 6.30 | <0.001 |
| Total cholesterol, mmol/L | 4.80 (4.45, 5.39) | 4.77 (4.35, 5.27) | 0.320 |
| Triglyceride, mmol/L | 1.63 (1.19, 2.50) | 1.40 (1.05, 2.06) | 0.015 |
| High‐density lipoprotein, mmol/L | 1.38 (1.20, 1.54) | 1.36 (1.22, 1.65) | 0.349 |
| Low‐density lipoprotein, mmol/L | 2.58 ± 0.51 | 2.58 ± 0.56 | 0.927 |
| Neuroimaging features | |||
| TIV, mL | 1552.52 ± 158.01 | 1525.00 ± 128.85 | 0.146 |
| Relative brain macrostructural volume, % of TIV | |||
| Cerebral parenchyma | 74.04 ± 3.58 | 75.53 ± 3.32 | 0.001 |
| Gray matter | 40.29 ± 2.34 | 41.01 ± 2.52 | 0.024 |
| White matter | 33.93 (32.79, 35.18) | 34.81 (33.49, 35.75) | 0.003 |
| Cerebrospinal fluid | 25.81 ± 3.52 | 24.36 ± 3.29 | 0.001 |
| Hippocampus | 0.24 ± 0.02 | 0.25 ± 0.02 | 0.125 |
Note: Group E, early‐onset hypertension group; Group EC, control group of early‐onset hypertension. Values are presented as mean (SD), median (IQR), or no. (%).
Abbreviation: TIV, total intracranial volume.
aComparisons between Groups E and EC were based on independent t‐test, Mann–Whitney U test, or chi‐square test.
bAge (in years) was calculated at the time of MRI acquisition.
TABLE 3.
Demographic, clinical characteristics, and neuroimaging metrics of participants in Groups L and LC.
| Group L (n = 354) | Group LC (n = 360) | p a | |
|---|---|---|---|
| Age b , years | 60.00 (55.00, 68.00) | 57.00 (51.00, 65.00) | <0.001 |
| Age of onset, years | 53.00 (48.00, 59.00) | — | |
| Female, no. (%) | 122 (34.46) | 198 (55.00) | <0.001 |
| Smoking status, no. (%) | 188 (53.11) | 139 (38.61) | <0.001 |
| Alcohol consumption, no. (%) | 233 (65.82) | 200 (55.56) | 0.005 |
| Physical activity, no. (%) | 0.811 | ||
| Sometimes or seldom | 190 (53.67) | 190 (52.78) | |
| Usually | 164 (46.33) | 170 (47.22) | |
| History of diabetes, no. (%) | 96 (27.12) | 54 (15.00) | <0.001 |
| Blood pressure, mmHg | |||
| Systolic | 132.86 (125.56, 140.92) | 119.39 (112.32, 125.47) | <0.001 |
| Diastolic | 84.98 ± 6.63 | 77.05 ± 6.85 | <0.001 |
| Total cholesterol, mmol/L | 5.04 (4.52, 5.53) | 4.87 (4.36, 5.47) | 0.009 |
| Triglyceride, mmol/L | 1.64 (1.24, 2.33) | 1.27 (0.92, 1.83) | <0.001 |
| High‐density lipoprotein, mmol/L | 1.42 (1.25, 1.61) | 1.53 (1.34, 1.71) | <0.001 |
| Low‐density lipoprotein, mmol/L | 2.75 (2.35, 3.12) | 2.63 ± 0.65 | 0.002 |
| Neuroimaging features | |||
| TIV, mL | 1512.49 ± 139.36 | 1463.88 (1380.43, 1559.71) | <0.001 |
| Relative brain macrostructural volume, % of TIV | |||
| Cerebral parenchyma | 71.34 ± 3.96 | 73.41 ± 3.73 | <0.001 |
| Gray matter | 38.65 ± 2.41 | 40.02 ± 2.42 | <0.001 |
| White matter | 32.70 ± 2.16 | 33.56 (32.10, 34.73) | <0.001 |
| Cerebrospinal fluid | 28.43 ± 3.84 | 26.45 ± 3.68 | <0.001 |
| Hippocampus | 0.24 (0.22, 0.25) | 0.24 ± 0.02 | <0.001 |
Note: Group L, late‐onset hypertension group; Group LC, control group of late‐onset hypertension. Values are presented as mean (SD), median (IQR), or no. (%). Bold values indicates that p value 〈0.05.
Abbreviation: TIV, total intracranial volume.
aComparisons between Groups L and LC were based on independent t‐test, Mann–Whitney U test, or chi‐square test.
bAge (in years) was calculated at the time of MRI acquisition.
Table 4 presents the multivariate‐adjusted associations between early‐onset hypertension and brain health. After full adjustments for various confounding factors, early‐onset hypertension was found to be associated with lower relative volumes of cerebral parenchyma (β = −0.302; 95% confidence interval [CI], −0.541 to −0.063), GM (β = −0.338; 95% CI, −0.590 to −0.087), and higher relative volumes of CSF (β = 0.311; 95% CI, 0.068–0.554). Table 5 presents the multivariate‐adjusted associations between late‐onset hypertension and brain health. Before adjustments for any variables, the crude analyses showed that late‐onset hypertension was associated with lower relative volumes of cerebral parenchyma (β = −0.498; 95% CI, −0.633 to −0.362), GM (β = −0.503; 95% CI, −0.632 to −0.373), WM (β = −0.328; 95% CI, −0.473 to −0.183), and hippocampus (β = −0.373; 95% CI, −0.520 to −0.225), and higher relative volumes of CSF (β = 0.486; 95% CI, 0.351–0.622). After full adjustments, late‐onset hypertension was no longer found to be associated with any brain volume variables.
TABLE 4.
Association of early‐onset hypertension with brain macrostructural volume in Groups E and EC.
| Brain macrostructural volume (in z‐score) | Group E | Group EC |
|---|---|---|
| n = 117 | n = 117 | |
| Cerebral parenchyma | ||
| Model 1 | −0.360 (95% CI, −0.572 to −0.148) | Reference |
| Model 2 | −0.369 (95% CI, −0.530 to −0.208) | Reference |
| Model 3 | −0.302 (95% CI, −0.541 to −0.063) | Reference |
| Gray matter | ||
| Model 1 | −0.264 (95% CI, −0.490 to −0.037) | Reference |
| Model 2 | −0.273 (95% CI, −0.443 to −0.103) | Reference |
| Model 3 | −0.338 (95% CI, −0.590 to −0.087) | Reference |
| White matter | ||
| Model 1 | −0.367 (95% CI, −0.596 to −0.139) | Reference |
| Model 2 | −0.373 (95% CI, −0.585 to −0.161) | Reference |
| Model 3 | −0.155 (95% CI, −0.468 to 0.157) | Reference |
| Cerebrospinal fluid | ||
| Model 1 | 0.356 (95% CI, 0.142–0.570) | Reference |
| Model 2 | 0.365 (95% CI, 0.202–0.529) | Reference |
| Model 3 | 0.311 (95% CI, 0.068–0.554) | Reference |
| Hippocampus | ||
| Model 1 | −0.174 (95% CI, −0.393 to 0.046) | Reference |
| Model 2 | −0.151 (95% CI, −0.345 to 0.043) | Reference |
| Model 3 | 0.095 (95% CI, −0.191 to 0.380) | Reference |
Note: Group E, early‐onset hypertension group; Group EC, control group of early‐onset hypertension. Model 1 did not correct for any covariates. Model 2 was adjusted for age, gender, smoking status, alcohol consumption, physical activity, history of diabetes, total cholesterol, triglyceride, high‐density lipoprotein, and low‐density lipoprotein. Model 3 was additionally adjusted for average blood pressure level and duration of hypertension on the basis of Model 2. Results are presented as β (95% CI). Bold values indicated that p value <0.05.
Abbreviation: TIV, total intracranial volume.
TABLE 5.
Association of late‐onset hypertension with brain macrostructural volume in Groups L and LC.
| Brain macrostructural volume (in z‐score) | Group L | Group LC |
|---|---|---|
| n = 354 | n = 360 | |
| Cerebral parenchyma | ||
| Model 1 | −0.498 (95% CI, −0.633 to −0.362) | Reference |
| Model 2 | −0.167 (95% CI, −0.263 to −0.071) | Reference |
| Model 3 | −0.066 (95% CI, −0.194 to 0.061) | Reference |
| Gray matter | ||
| Model 1 | −0.503 (95% CI, −0.632 to −0.373) | Reference |
| Model 2 | −0.182 (95% CI, −0.276 to −0.087) | Reference |
| Model 3 | −0.090 (95% CI, −0.215 to 0.035) | Reference |
| White matter | ||
| Model 1 | −0.328 (95% CI, −0.473 to −0.183) | Reference |
| Model 2 | −0.093 (95% CI, −0.222 to 0.036) | Reference |
| Model 3 | −0.014 (95% CI, −0.185 to 0.158) | Reference |
| Cerebrospinal fluid | ||
| Model 1 | 0.486 (95% CI, 0.351–0.622) | Reference |
| Model 2 | 0.154 (95% CI, 0.058–0.251) | Reference |
| Model 3 | 0.056 (95% CI, −0.072 to 0.184) | Reference |
| Hippocampus | ||
| Model 1 | −0.373 (95% CI, −0.520 to −0.225) | Reference |
| Model 2 | −0.116 (95% CI, −0.239 to 0.007) | Reference |
| Model 3 | −0.034 (95% CI, −0.198 to 0.130) | Reference |
Note: Group L, late‐onset hypertension group; Group LC, control group of late‐onset hypertension. Model 1 did not correct for any covariates. Model 2 was adjusted for age, gender, smoking status, alcohol consumption, physical activity, history of diabetes, total cholesterol, triglyceride, high‐density lipoprotein, and low‐density lipoprotein. Model 3 was additionally adjusted for average blood pressure level and duration of hypertension based on Model 2. Results are presented as β (95% CI). Bold values indicated that p value <0.05.
Abbreviation: TIV, total intracranial volume.
Voxel‐wise analysis showed that compared with those in Group EC, participants in Group E had significant GM atrophy, mainly occurring in the occipital lobe, frontal lobe, and temporal lobe (Figure 1A; Table S2). The interaction effect analysis showed that the impact of early‐onset hypertension on brain volume was mainly reflected in the reduction of frontal lobe volume (Figure 1B; Table S3). Compared with those in Group LC, participants in Group L had significant GM atrophy, mainly occurring in the frontal lobe, temporal lobe, occipital lobe, and paracentral lobule (Figure 1C; Table S4).
FIGURE 1.
Association of early‐onset and late‐onset hypertension with neuroimaging features. Regions with P voxel < 0.005 and P cluster < 0.05 are reported, with age, gender, duration of hypertension, mean systolic blood pressure, mean diastolic blood pressure, and TIV as covariates. (A) Results are based on a comparison of Group E versus Group EC at the voxel level. Cold color represents reduced gray matter volume in Group E relative to Group EC, with atrophy predominantly in the occipital lobe, frontal lobe, and temporal lobe. (B) Results are based on the interaction effect of early‐onset and late‐onset hypertension on brain volume at the voxel level. Cold color represents reduced gray matter volume in Group E relative to Group L, with atrophy predominantly in the frontal lobe. (C) Results are based on a comparison of Group L versus Group LC at the voxel level. Cold color represents reduced gray matter volume in Group L relative to Group LC, with atrophy predominantly in the frontal lobe, temporal lobe, occipital lobe, and paracentral lobule. L, left; R, right.
4. Discussion
The main objective of this study was to investigate the impact of early‐onset hypertension on neuroimaging features of brain volume. Although previous studies have confirmed that hypertension is associated with brain structural damage, few studies have systematically explored whether the age at hypertension onset affects the association with brain volume atrophy. On the basis of the large sample longitudinal cohort in META‐KLS, we found that compared to participants with late‐onset hypertension, those with early‐onset hypertension exhibited more pronounced reductions in brain volume, alongside a greater increase in CSF volume. We also observed that early‐onset hypertension was robustly correlated with brain volume atrophy, and this association persisted after comprehensive adjustment for confounding factors, suggesting that early‐onset hypertension may be a key contributing factor. In contrast, the observed association between late‐onset hypertension and brain volume appeared to be potentially related to unhealthy lifestyle factors and cardiovascular risk factors, as these confounders fully explained the observed brain volume differences after multivariable adjustment. In summary, our findings emphasized the detrimental effects of early‐onset hypertension on brain health.
Several studies have compared the consequences of early‐onset hypertension with late‐onset hypertension. A multicenter cohort analysis (n = 10,313) in 1987 [19] showed that compared with normotensive individuals of the same age group, patients with hypertension onset at 40–49 years of age have significantly higher cardiovascular disease risk than patients with hypertension onset at 60–65 years of age. However, this early study lacked further adjustment for confounding factors. Analyses from the Framingham Heart Study [20] and the CARDIA study [5] found that compared with the normotensive population, individuals with early‐onset hypertension had higher odds of cardiovascular death and target organ damage than individuals with late‐onset hypertension. However, these studies did not examine changes in brain structure. As mentioned previously, there is a lack of prospective studies comprehensively evaluating the relationship between age of onset of hypertension and brain health, which has important clinical and public health implications. There is a need for bridging the disconnect between policies that address hypertension treatment among older adults while overlooking the negative brain health of early‐onset hypertension among young adults.
In this study, Group L was older and exhibited smaller brain volumes and larger CSF volume than that of Group E. In order to further disentangle whether the relatively small brain volume in Group L was caused by aging or the onset of hypertension, we subsequently conducted PSM and voxel‐based interaction analysis. Regardless of the age of diagnosis of hypertension, participants with hypertension exhibited lower brain volume and higher CSF volume compared to the corresponding controls with normal BP. Specifically, compared with Group LC, Group L was more likely to have smoking and drinking habits and an adverse cardiovascular risk profile. This was consistent with the epidemiological characteristics of this population in the real world. Without adjusting for any variables, late‐onset hypertension was associated with decreased brain volume. However, after adjusting for potential confounding variables, late‐onset hypertension was no longer associated with any brain volume variables. This suggested that this association may be potentially related to unhealthy lifestyle factors and cardiovascular risk factors in individuals with late‐onset hypertension, as these confounders explained the observed brain volume differences after full adjustments. In contrast, early‐onset hypertension presented a different situation. Even after comprehensive adjustment for various factors, early‐onset hypertension remained negatively correlated with brain volume. This further indicated that early‐onset hypertension was robustly correlated with brain volume atrophy, and this association persisted after adjusting for confounding factors, suggesting that early‐onset hypertension may be a key contributing factor. Recently, a CARDIA study [21] found that early‐onset hypertension is associated with cognitive impairment, but not with changes in the macroscopic structure of the brain. This discrepancy can be attributed to several differences in study design, population characteristics, and methodologies. A critical distinction lies in the age threshold for early‐onset hypertension. The CARDIA study defined it as a diagnosis of ≤35 years old, whereas our study adopted a diagnosis of ≤40 years old. The difference in age cutoff may lead to inclusion in different subgroups, which may not be sufficient to cause detectable macroscopic structural changes. There are also significant differences in demographic data. The CARDIA study was characterized by a multiethnic cohort in the United States, whereas this study focused on a Chinese community–based population. The racial/ethnic differences in genetic susceptibility to hypertension‐related brain injury, as well as environmental factors such as dietary patterns and healthcare pathways, may be the basis for different findings. Additionally, the CARDIA study used the 2017 ACC/AHA hypertension definition (SBP ≥ 130 mmHg/DBP ≥ 80 mmHg), which is more stringent than the JNC 7 criteria (SBP ≥ 140 mmHg/DBP ≥ 90 mmHg) employed in our study. Finally, there are also significant differences in neuroimaging methods. The CARDIA study utilized 3.0T MRI with conventional T1‐weighted sequences and global brain volume analysis, while our study used 3.0T MRI with high‐resolution 3D‐BRAVO‐T1‐weighted sequences and voxel‐wise analysis to detect regional structural changes. We also systematically adjusted the TIV to account for individual differences in head size.
Our research also conducted voxel‐level analysis of neuroimaging features, which provided objective and quantitative biomarkers to reflect brain health status [22]. This voxel‐level analysis revealed that the brain regions particularly affected were mainly distributed in the frontal, temporal, and occipital lobes [23, 24, 25]. Moreover, early‐onset and late‐onset hypertension shared some overlapping brain regions, but they also exhibited distinct patterns of affected areas. These differences may reflect different mechanisms underlying the observed structural differences [26]. Although the deterioration of GM structure may reflect neurodegenerative processes such as inflammation, amyloid beta protein deposition, and oxidative stress [27], this study excluded the influence of confounding factors that may be caused by age through PSM and interaction analysis.
The mechanisms underlying the associations of poor brain structure in participants with early‐onset hypertension remain unclear. A continuous supply of oxygen and glucose from blood to the brain is fundamental for maintaining brain health. Long‐term exposure to hypertension may lead to increased arterial stiffness and reduced compliance, which can reduce cerebral blood flow and increase cerebrovascular reactivity [28, 29, 30], thus resulting in brain damage. Early‐onset hypertension may better reflect total exposure to elevated BP throughout the lifespan, thus causing potential damage to the brain for a longer time. This may partly explain why there was a stronger association of early‐onset hypertension with smaller brain volumes in our study.
Although young adults <40 years of age have a relatively low incidence of hypertension [31], most of these young hypertensive patients are unaware of their poor brain health, which would still translate into a great disease burden [32]. Compared to the older generation, the awareness, treatment, and control rates of hypertension among young people were much lower (31.7%, 24.5%, and 9.9% for those aged 35–44 and 58.6%, 52.8%, and 18.4% for those aged 65–74, respectively). Our study suggests that the consideration of hypertension onset age will provide novel and preventive information. Self‐awareness of the condition is the first step to modify behavior and lifestyle changes. Early‐onset hypertensive patients need more stringent antihypertensive therapy as early as possible, and more intensive lifestyle interventions should also be actively encouraged, including weight loss, limiting salt intake, smoking and alcohol cessation, and regular physical exercise [33].
Our study has several limitations. First, the cohort was from a Chinese community, which limited the promotion to other ethnic or geographical populations. Second, the cross‐sectional analysis was conducted on the age at hypertension diagnosis and brain volume, which cannot refer to causal relationships. Further longitudinal studies are needed to explore the association between age at diagnosis of hypertension and changes in brain volume over time. Third, there may be recall bias in the diagnosis of hypertension, which may lead to an underestimation of the prevalence of hypertension. Fourth, this study focused solely on brain structural and volumetric metrics. We did not assess functional brain measures (e.g., functional MRI‐derived functional connectivity) or WM integrity (e.g., DTI‐derived fractional anisotropy and mean diffusivity)—key indicators that could provide complementary insights into hypertension‐related brain dysfunction, which may be more sensitive than structural volume changes in younger populations. Further studies are needed to address this gap in subsequent work by integrating functional MRI and DTI‐derived metrics to comprehensively evaluate structural, functional, and microstructural brain changes in early‐onset hypertensive populations. Finally, although neuroimaging indicators are robust, they cannot fully capture the functional or molecular changes behind structural abnormalities.
5. Conclusions
This study suggests that age at which hypertension is diagnosed plays an important role in the association between hypertension and brain volume. Compared with late‐onset hypertension, early‐onset hypertension is associated with smaller brain volume. These findings underscore the urgency of early BP management and targeted brain health protection for young hypertensive populations. For individuals diagnosed with hypertension before the age of 40, comprehensive management should integrate rigorous antihypertensive therapy and regular brain health monitoring to enable timely diagnosis and intervention before irreversible brain structural changes occur.
Funding
This study was supported by Grant nos. 62522119 and 82502453 from the National Natural Science Foundation of China; Grant nos. 7242267 and L242024 from Beijing Natural Science Foundation; Grant no. B2408 from Capital Medical University; and Grant no. KM202410025017 from the R&D Program of Beijing Municipal Education Commission.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Supporting Table 1: Neuroimaging acquisition parameters.
Supporting Table 2: GMV differences between group E and group EC.
Supporting Table 3: GMV differences after interaction between group E ‐ group EC and group L ‐ group LC.
Supporting Table 4: GMV differences between group L and group LC.
Acknowledgments
We thank all the participants for making this study possible and all the colleagues who contributed to this research.
Contributor Information
Han Lv, Email: chrislvhan@126.com.
Yuntao Wu, Email: wyt0086@163.com.
Zhenchang Wang, Email: cjr.wzhch@vip.163.com.
Data Availability Statement
The data in the current study are available from the corresponding author on reasonable request.
References
- 1. GBD 2019 Risk Factors Collaborators , “Global Burden of 87 Risk Factors in 204 Countries and Territories, 1990–2019: A Systematic Analysis for the Global Burden of Disease Study 2019,” Lancet 396, no. 10258 (2020): 1223–1249, 10.1016/S0140-6736(20)30752-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Chen Y., Hu S., Li Y., et al., “Systematic Review of Hypertension Clinical Practice Guidelines Based on the Burden of Disease: A Global Perspective,” Journal of Evidence‐Based Medicine 7, no. 1 (2014): 52–59, 10.1111/jebm.12082. [DOI] [PubMed] [Google Scholar]
- 3. Wang Z., Chen Z., Zhang L., et al., “Status of Hypertension in China: Results From the China Hypertension Survey, 2012–2015,” Circulation 137, no. 22 (2018): 2344–2356, 10.1161/CIRCULATIONAHA.117.032380. [DOI] [PubMed] [Google Scholar]
- 4. Roth G. A., Mensah G. A., Johnson C. O., et al., “Global Burden of Cardiovascular Diseases and Risk Factors, 1990–2019: Update From the GBD 2019 Study,” Journal of the American College of Cardiology 76, no. 25 (2020): 2982–3021, 10.1016/j.jacc.2020.11.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Suvila K., McCabe E. L., Lehtonen A., et al., “Early Onset Hypertension Is Associated With Hypertensive End‐Organ Damage Already by MidLife,” Hypertension 74, no. 2 (2019): 305–312, 10.1161/HYPERTENSIONAHA.119.13069. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Suvila K., McCabe E. L., Lima J. A. C., et al., “Self‐Reported Age of Hypertension Onset and Hypertension‐Mediated Organ Damage in Middle‐Aged Individuals,” American Journal of Hypertension 33, no. 7 (2020): 644–651, 10.1093/ajh/hpaa055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Wang C., Yuan Y., Zheng M., et al., “Association of Age of Onset of Hypertension With Cardiovascular Diseases and Mortality,” Journal of the American College of Cardiology 75, no. 23 (2020): 2921–2930, 10.1016/j.jacc.2020.04.038. [DOI] [PubMed] [Google Scholar]
- 8. Power M. C., Schneider A. L. C., Wruck L., et al., “Life‐Course Blood Pressure in Relation to Brain Volumes,” Alzheimers Dement 12, no. 8 (2016): 890–899, 10.1016/j.jalz.2016.03.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Wartolowska K. A. and Webb A. J. S., “Midlife Blood Pressure Is Associated With the Severity of White Matter Hyperintensities: Analysis of the UK Biobank Cohort Study,” European Heart Journal 42, no. 7 (2021): 750–757, 10.1093/eurheartj/ehaa756. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Lane C. A., Barnes J., Nicholas J. M., et al., “Associations Between Blood Pressure Across Adulthood and Late‐Life Brain Structure and Pathology in the Neuroscience Substudy of the 1946 British Birth Cohort (Insight 46): An Epidemiological Study,” Lancet Neurology 18, no. 10 (2019): 942–952, 10.1016/S1474-4422(19)30228-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Zhang Q., Zhou Y., Gao X., et al., “Ideal Cardiovascular Health Metrics and the Risks of Ischemic and Intracerebral Hemorrhagic Stroke,” Stroke; A Journal of Cerebral Circulation 44, no. 9 (2013): 2451–2456, 10.1161/STROKEAHA.113.678839. [DOI] [PubMed] [Google Scholar]
- 12. Wu S., An S., and Li W., “Association of Trajectory of Cardiovascular Health Score and Incident Cardiovascular Disease,” JAMA Network Open 2, no. 5 (2019): e194758, 10.1001/jamanetworkopen.2019.4758. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Yu Y., Dong Z., Li Y., et al., “The Cardiovascular and Cerebrovascular Health in North China From 2006 to 2011: Results From the KaiLuan Study,” Frontiers in Cardiovascular Medicine 8 (2021): 683416, 10.3389/fcvm.2021.683416. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Sun J., Hui Y., Li J., et al., “Protocol for Multi‐Modality MEdical Imaging sTudy bAsed on KaiLuan Study (META‐KLS): Rationale, Design and Database Building,” BMJ Open 13, no. 2 (2023): e067283, 10.1136/bmjopen-2022-067283. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Sun J., Zeng N., Hui Y., et al., “Association of Variability in Body Size With Neuroimaging Metrics of Brain Health: A Population‐Based Cohort Study,” Lancet Regional Health – Western Pacific 44 (2024): 101015, 10.1016/j.lanwpc.2024.101015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Chobanian A. V., Bakris G. L., Black H. R., et al., “The Seventh Report of the Joint National Committee on Prevention Detection, Evaluation, and Treatment of High Blood Pressure: The JNC 7 Report,” JAMA 289, no. 19 (2003): 2560–2572. [DOI] [PubMed] [Google Scholar]
- 17. Gilsanz P., Mayeda E. R., Glymour M. M., et al., “Female Sex, Early‐Onset Hypertension, and Risk of Dementia,” Neurology 89, no. 18 (2017): 1886–1893, 10.1212/WNL.0000000000004602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Manghat N. E., Robinson E., and Mitrousi K., “Cerebrovascular Variants and the Role of the Selfish Brain in Young‐Onset Hypertension,” Hypertension 79, no. 6 (2022): 1265–1274, 10.1161/HYPERTENSIONAHA.121.18612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Buck C., Baker P., Bass M., et al., “The Prognosis of Hypertension According to Age at Onset,” Hypertension 9, no. 2 (1987): 204–208. [DOI] [PubMed] [Google Scholar]
- 20. Niiranen T. J., McCabe E. L., Larson M. G., et al., “Heritability and Risks Associated With Early Onset Hypertension: Multigenerational, Prospective Analysis in the Framingham Heart Study,” BMJ 357 (2017): j1949, 10.1136/bmj.j1949. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Suvila K., Lima J. A. C., Yano Y., et al., “Early‐But Not Late‐Onset Hypertension Is Related to Midlife Cognitive Function,” Hypertension 77, no. 3 (2021): 972–979, 10.1161/HYPERTENSIONAHA.120.16556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Lv H., Zeng N., Li M., et al., “Association Between Body Mass Index and Brain Health in Adults: A 16‐Year Population‐Based Cohort and Mendelian Randomization Study,” Health Data Science 4 (2024): 0087, 10.34133/hds.0087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Li X., Liang Y., Chen Y., et al., “Disrupted Frontoparietal Network Mediates White Matter Structure Dysfunction Associated With Cognitive Decline in Hypertension Patients,” Journal of Neuroscience 35, no. 27 (2015): 10015–10024, 10.1523/JNEUROSCI.5113-14.2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Li X., Wang W., and Wang A., “Vulnerability of the Frontal and Parietal Regions in Hypertensive Patients During Working Memory Task,” Journal of Hypertension 35, no. 5 (2017): 1044–1051, 10.1097/HJH.0000000000001250. [DOI] [PubMed] [Google Scholar]
- 25. Gonzales M. M., Ajilore O., Charlton R. C., et al., “Divergent Influences of Cardiovascular Disease Risk Factor Domains on Cognition and Gray and White Matter Morphology,” Psychosomatic Medicine 79, no. 5 (2017): 541–548, 10.1097/PSY.0000000000000448. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Haines H. T., Suri S., Patel R., et al., “Association of Early Life Cardiovascular Risk Factors With Grey Matter Structure in Young Adults in the United Kingdom: The ALSPAC Study,” EBioMedicine 110 (2024): 105490, 10.1016/j.ebiom.2024.105490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Qiu C. and Fratiglioni L., “A Major Role for Cardiovascular Burden in Age‐Related Cognitive Decline,” Nature Reviews Cardiology 12, no. 5 (2015): 267–277, 10.1038/nrcardio.2014.223. [DOI] [PubMed] [Google Scholar]
- 28. Boutouyrie P., Chowienczyk P., Humphrey J. D., et al., “Arterial Stiffness and Cardiovascular Risk in Hypertension,” Circulation Research 128, no. 7 (2021): 864–886, 10.1161/CIRCRESAHA.121.318061. [DOI] [PubMed] [Google Scholar]
- 29. Sforza M., Bianchini E., Alivernini D., et al., “The Impact of Cerebral Vasomotor Reactivity on Cerebrovascular Diseases and Cognitive Impairment,” Journal of Neural Transmission (Vienna) 129, no. 11 (2022): 1321–1330, 10.1007/s00702-022-02546-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Jefferson A. L., “Midlife Consequences of Cumulative Blood Pressure Exposure: Importance of a Lifespan Approach,” Circulation 141, no. 9 (2020): 725–727, 10.1161/CIRCULATIONAHA.120.044447. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Joffres M., Falaschetti E., Gillespie C., et al., “Hypertension Prevalence, Awareness, Treatment and Control in National Surveys From England, the USA and Canada, and Correlation With Stroke and Ischaemic Heart Disease Mortality: A Cross‐Sectional Study,” BMJ Open 3, no. 8 (2013): e003423, 10.1136/bmjopen-2013-003423. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Zhu H., Li J., Li L., et al., “Prevalence and Cardio‐Renal Comorbidities of Masked Hypertension: A Meta‐Analysis,” Journal of Evidence‐Based Medicine 17, no. 4 (2024): 833–842, 10.1111/jebm.12672. [DOI] [PubMed] [Google Scholar]
- 33. Sun J., Wang L., Gao Y., et al., “Discovery of High‐Risk Clinical Factors That Accelerate Brain Aging in Adults: A Population‐Based Machine Learning Study,” Research (Washington DC) 7 (2024): 0500, 10.34133/research.0500. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Table 1: Neuroimaging acquisition parameters.
Supporting Table 2: GMV differences between group E and group EC.
Supporting Table 3: GMV differences after interaction between group E ‐ group EC and group L ‐ group LC.
Supporting Table 4: GMV differences between group L and group LC.
Data Availability Statement
The data in the current study are available from the corresponding author on reasonable request.