Screening for secondary hypertension (HTN) is recommended for early-onset HTN.
However, there have been few studies on secondary HTN in young adults. We aimed
to investigate the prevalence and risk factors for secondary HTN in young male
military personnel. In this retrospective cross-sectional study,
hypertensive men (age, 19-29 years) were identified using the electronic medical
records (EMR) database between 2011 and 2017. Among them, patients with secondary
HTN were confirmed through a review of the EMR. Using clinical characteristics
and laboratory findings, independent predictors associated with secondary HTN
were identified by binary logistic regression analysis. Secondary HTN was
confirmed in 140 of 6373 participants (2.2%). Overall, the most common causes
were polycystic kidney disease (n = 47, 0.74%) and renal parenchymal diseases (n
= 24, 0.38%). The independent predictors of secondary HTN were abnormal thyroid
function test (TFT) (odds ratio [OR]: 9.50, 95% confidence interval [CI]:
Secondary hypertension (HTN) is defined as HTN with an identifiable cause, which
may be treatable with specific interventions. If the underlying cause is
identified and treated before the age of 30-40 years, blood pressure (BP) is
likely to normalize. However, if the cause is treated later in life, HTN may
persist because long-standing high BP results in damage to the arterial system
and other organs (Streeten et al., 1990). Therefore, the guidelines from
North America and Europe suggest screening for secondary HTN in patients with
The prevalence of secondary HTN in hypertensive patients is reportedly 5%-15%
(Anderson et al., 1994; Danielson and Dammström, 2009; Omura et al., 2004; Sinclair, 1987). Unlike the prevalence of primary HTN, which increases with
age, (Chow, 2013; Jago, 2006; Zhou et al., 2017) the prevalence of secondary
HTN is the highest in preschool children (51%), (Gupta-Malhotra et al., 2015) the lowest in younger adults aged
In South Korea, men aged 18-35 years are obligated to complete a 2-year military
service. All recruits undergo medical checkups to confirm eligibility at the time
of enlistment. Men are exempted from military service if their height is
There have been many studies on secondary HTN so far, but little is known about
secondary HTN in young adults aged
Patient flowchart. searching, assessment, and exclusion. HTN, hypertension. Among the 6586 cases, there were 638 cases of ‘possible’ secondary HTN with a diagnostic code associated with secondary HTN. After excluding 213 cases according to the exclusion criteria, 140 cases were confirmed to have secondary HTN through a meticulous review of EMR. In contrast, the remaining 285 cases were of primary HTN, despite having diagnostic codes associated with secondary HTN.
This retrospective cross-sectional study was conducted to investigate the prevalence and predictors of secondary HTN in young male military personnel. This study used data from the Defense Medical Information System (DEMIS) database between January 2011 and June 2017.
Patients who had a diagnosis of primary and secondary HTN were identified from the DEMIS database. The DEMIS database includes electronic medical records (EMR), imaging data, and laboratory data from 19 military hospitals and about 1200 medical corps in South Korea from 1997 till date. The diagnoses were coded according to the International Classification of Diseases 10th revision (ICD-10).
Collecting primary HTN cases is relatively easy and simple using ICD-10 code ’I10’; however, this is not the case for secondary HTN. Although ICD-10 code ’I15’ is assigned to secondary HTN, a significant number of cases with a secondary cause only have a diagnostic code for the secondary cause without ICD-10 code ’I15’, at the physician’s discretion. Therefore, we first collected all possible cases of secondary HTN with ICD-10 code ’I15’ or other codes for causative diseases (Table S2 in the Supplementary Material). Subsequently, through meticulous reviews of EMR, we eliminated inappropriate cases according to the exclusion criteria. Exclusion criteria were designed to remove cases that failed to meet the diagnostic criteria, those with an obvious preceding cause such as trauma or sepsis, and those that failed to meet the diagnostic criteria for HTN (Fig. 1) (Mancia et al., 2013). Subsequently, patient characteristics including clinical and demographic information and laboratory and imaging data were extracted from DEMIS directly into the electronic database by the author.
HTN was defined as a systolic BP
To confirm the diagnosis of primary aldosteronism (PA), test values of plasma
aldosterone concentration (PAC), plasma renin activity (PRA), and plasma ARR were
obtained. Patients taking any antihypertensive medication had to undergo a
medication washout period before undergoing screening tests to
avoid interference with the
renin-angiotensin-aldosterone system (RAAS). Dihydropyridine calcium channel
The diagnostic criteria for subclinical hypercortisolism remain controversial
and uncertain. Considering the sensitivity and specificity of previous studies,
we diagnosed subclinical hypercortisolism if the 24-h UFC was
For autosomal dominant polycystic kidney disease (ADPKD), the presence of three
or more renal cysts (unilateral or bilateral) was considered sufficient for
diagnosis in patients with a family history. Without a family history, the
All urine or blood samples for hormone tests except TFT were sent from 19 military hospitals across the country to Green Cross Laboratories, Korea’s leading clinical laboratory, for analyses. PAC, PRA, and 24-h UFC were assessed using radioimmunoassay. Plasma metanephrine and normetanephrine levels were analyzed using liquid chromatography coupled with tandem quadrupole mass spectrometry, whereas 24-h urine metanephrine, normetanephrine, catecholamines, and vanillylmandelic acid levels were measured using high-performance liquid chromatography.
The number of patients needed to obtain a confidence level of 95% and such that
the real value is within
Data are presented as median (interquartile range) for continuous variables with
a non-normal distribution and as frequency (percentage) for categorical
variables. The Shapiro-Wilk normality test was used to assess normal
distribution. The Wilcoxon rank sum test with continuity correction was used to
compare continuous variables with a non-normal distribution. The chi-square test
We imputed missing data using multivariate imputation via the chained equations
package in R program under missing-at-random assumptions. Binary logistic
regression analysis was performed to identify independent variables associated
with an increased risk for the diagnosis of secondary HTN. We included
overweight/obese, family history of HTN, hypokalemia, hematuria, proteinuria,
ALT, prediabetes/diabetes, TFT, and severe HTN in binary form in the first model.
Multicollinearity of independent variables was checked using the variance
inflation factor (VIF). A VIF value
Approval for the study was obtained from the institutional review board of the Armed Forces Medical Command (institutional review board approval number: AFMC-16057-IRB-16-045). The requirement for written informed consent was waived because the data were anonymous and retrospectively analyzed. The study was designed in accordance with the ethical guidelines of the Declaration of Helsinki.
|Cause of secondary HTN||N (% of total HTN)|
|Autosomal dominant polycystic kidney disease||47 (0.74)|
|Renal parenchymal disease||24 (0.38)|
|Chronic glomerulonephritis||19 (0.30)|
|Atrophic kidney||2 (0.03)|
|Solitary kidney||2 (0.03)|
|Tubulointerstitial nephritis||1 (0.02)|
|Subclinical hypercortisolism||12 (0.19)|
|Without adrenal adenoma||9 (0.14)|
|With adrenal adenoma||3 (0.05)|
|Renal artery stenosis||9 (0.14)|
|Fibromuscular dysplasia||3 (0.05)|
|Takayasu’s arteritis||2 (0.03)|
|Cushing’s syndrome||5 (0.08)|
|Primary aldosteronism||3 (0.05)|
|Aldosterone-producing adenoma||2 (0.03)|
|Idiopathic hyperaldosteronism||1 (0.02)|
|Coarctation of the aorta||1 (0.02)|
|Total HTN||6373 (100)|
|Clinical characteristics||Secondary HTN (n = 140)||Primary HTN (n = 285)||P-value|
|Age [range], years||21.0 [20.0; 22.0]||20.0 [20.0; 21.0]||0.018|
|Body mass index (BMI) [range], kg/m
||24.5 [22.3; 27.2]||26.0 [23.7; 29.1]||0.001|
||2 (1.6%)||0 (0.0%)|
||43 (33.9%)||42 (18.8%)|
||82 (64.6%)||182 (81.2%)|
|Smoking, number (%)||0.289|
|Never||65 (55.1%)||109 (52.9%)|
|Past||15 (12.7%)||17 (8.3%)|
|Current||38 (32.2%)||80 (38.8%)|
|History of HTN in first-degree relatives, number (%)||66 (49.6%)||138 (57.3%)||0.19|
|Headache, number (%)||39 (29.8%)||73 (32.3%)||0.705|
|BUN [range], mg/dL||13.8 [11.4; 16.5]||13.2 [11.6; 14.7]||0.018|
|Creatinine [range], mg/dL||0.9 [0.8; 1.1]||1.0 [0.9; 1.1]||0.005|
|Serum potassium, no. (%)||0.009|
||9 (6.5%)||3 (1.1%)|
||125 (89.9%)||260 (95.2%)|
||5 (3.6%)||10 (3.7%)|
|Hematuria, number (%)|
|Absent (-)||92 (68.7%)||241 (94.1%)|
|Trace (+/-)||16 (11.9%)||7 (2.7%)|
|Mild (1+)||4 (3.0%)||5 (2.0%)|
|Moderate to severe (
||22 (16.4%)||3 (1.2%)|
|Proteinuria, number (%)|
|Absent (-)||86 (64.2%)||242 (94.2%)|
|Trace (+/-)||25 (18.7%)||13 (5.1%)|
|Mild (1+)||7 (5.2%)||1 (0.4%)|
|Moderate to severe (
||16 (11.9%)||1 (0.4%)|
|Alanine aminotransferase (ALT) [range], IU/L||22.0 [16.0; 35.0]||25.0 [17.0; 44.0]||0.045|
|Total cholesterol [range], mg/dL||170.0 [145.0; 197.0]||176.0 [154.0; 198.0]||0.066|
|Diabetes, number (%)||0.096|
|Non-diabetic||117 (84.8%)||200 (91.7%)|
|Prediabetic||17 (12.3%)||16 (7.3%)|
|Diabetic||4 (2.9%)||2 (0.9%)|
|Thyroid function test (TFT), number (%)|
|Normal||67 (63.2%)||243 (93.8%)|
|Subclinical hypothyroidism (high TSH)||1 (0.9%)||6 (2.3%)|
|Subclinical hyperthyroidism (low TSH)||5 (4.7%)||8 (3.1%)|
|Hypothyroidism||4 (3.8%)||0 (0.0%)|
|Hyperthyroidism||23 (21.7%)||0 (0.0%)|
|Other abnormal TFT||6 (5.7%)||2 (0.8%)|
|Office heart rate [range], rate/min||82.0 [74.0; 94.5]||82.0 [72.0; 95.0]||0.745|
|Classification of office BP* - no. (%)|
|High normal (
||6 (4.3%)||5 (1.8%)|
|Grade 1 (
||67 (47.9%)||166 (58.9%)|
|Grade 2 (
||41 (29.3%)||92 (32.6%)|
|Grade 3 (
||26 (18.6%)||19 (6.7%)|
|Isolated systolic HTN||51 (36.4%)||121 (43.1%)||0.231|
|Hypertension (HTN), blood urea nitrogen (BUN),
thyroid-stimulating hormone (TSH), blood pressure (BP).
*Office BP and heart rate values are the average of measurements
before taking antihypertensive drugs. Hypertension was defined and classified
according to the 2018 European Society of Hypertension/European Society of
High normal: systolic 130-139 mmHg and/or diastolic 85-89
Grade 1: systolic 140-159 mmHg and/or diastolic 90-99 mmHg.
Grade 2: systolic |
|Logistic regression||Univariate analysis||Multivariable analysis|
|Variables||Unadjusted OR||95% CI||P-value||VIF||Adjusted OR||95% CI||P-value|
|Thyroid function test (TFT), hypertension (HTN), odds ratio
(OR), confidence interval (CI), variance inflation factor (VIF).
*Severe HTN was defined as systolic |
We initially collected 6586 cases of primary and ‘possible’ secondary HTN from the DEMIS database using ICD-10 diagnostic codes assigned to primary HTN (I10), secondary HTN (I15), and various diagnoses related to secondary HTN (Table S2 in Supplementary Material). Among the 6586 cases, there were 638 cases of ‘possible’ secondary HTN with a diagnostic code associated with secondary HTN. Of the 638 cases, 425 cases of ‘possible’ secondary HTN were identified and analyzed after 213 cases were excluded according to the exclusion criteria. Finally, 140 cases (2.2% of all 6364 hypertensive cases) were confirmed to have secondary HTN through a meticulous review of EMR. In contrast, the remaining 285 cases were of primary HTN, despite having diagnostic codes associated with secondary HTN (Fig. 1).
Hormone tests for hyperaldosteronism, Cushing’s syndrome, pheochromocytoma, and thyroid disease screening were performed in 4311/6373 (67.6%) hypertensive patients. Imaging studies such as abdominal CT or ultrasonography were conducted for 3817/6373 (59.9%) patients to eliminate the presence of intra-abdominal mass, ADPKD, renal atrophy, or renal artery stenosis (Table S3 in the Supplementary Materials).
The most common causes of HTN were ADPKD (n = 47/6373, 0.74%) and renal parenchymal diseases (n = 24/6373, 0.38%), followed by hyperthyroidism (n = 23/6373, 0.36%) and subclinical hypercortisolism (n = 12/6373, 0.19%). Other endocrinologic causes, including pheochromocytoma/paraganglioma (n = 10/6373, 0.14%) and PA (n = 3/6373, 0.05%), were very rare (Table 1).
The demographic, clinical, and laboratory characteristics of patients with primary HTN (n = 285) and secondary HTN (n = 140) are summarized in Table 2. The median (interquartile range) age of patients with primary and secondary HTN was 20.0 (20.0-21.0) years and 21.0 (20.0-22.0) years, respectively. Both groups had few comorbidities, such as chronic kidney disease, diabetes, and hypercholesterolemia.
Patients with secondary HTN were more likely to be hypokalemic (6.5%, n =
9/139, vs. 1.1%, n = 3/273, P = 0.009), and to have hematuria (
Among the nine patients with hypokalemia in the secondary HTN group, there were three patients with renal artery stenosis, two patients with ADPKD, and one patient with subclinical hypercortisolism in addition to three patients with PA.
In multiple logistic regression analyses, six major clinical parameters were
identified as predictors of secondary HTN (Table 3): abnormal TFT (OR: 9.50,
95% CI 4.84-19.45, P
The VIF values for predictive factors were
In this study, we found that secondary HTN was very rare (2.2%, n = 140/6373)
among hypertensive military males aged
A previous study conducted from 1974 to 1991 in the United States reported that
secondary causes were noted in 5.6% of patients aged 18-29 years who were
referred for severe HTN (Cohen, 2017). Another study conducted from 1995 to
1999 in Japan demonstrated that the prevalence of secondary HTN was 9.1% among
1020 hypertensive patients in an outpatient clinical setting (Omura et al., 2004). However, the study did not describe in detail the age-related prevalence
of secondary HTN. To the best of our knowledge, there has been no previous study
on the prevalence of secondary HTN in young adults
The main reasons for the lower prevalence of secondary HTN in this study could be explained by the following two reasons. First, most of the 1200 medical corps in South Korea are in charge of primary care; however, they do not have antihypertensive drugs and cannot perform screening tests for secondary HTN. Therefore, in most cases where hypertensive patients are identified by a primary physician in the medical corps, patients are referred to a nearby military hospital for screening and treatment. This peculiarity of military medical systems makes the denominator large in calculating the prevalence of secondary HTN, hence the low prevalence. Second, as new recruits undergo multiple tests at the time of enlistment, a significant number of cases of secondary HTN would have been newly detected, and the affected individuals would be exempted from military service (Table S1 in the Supplementary Appendix). This could have reduced the number of patients with secondary HTN and increased the total number of hypertensive patients, resulting in a lower prevalence of secondary HTN.
Previous studies have reported that renal disease was the most common cause of
secondary HTN (Anderson et al., 1994; The Japanese Society of Hypertension, 2014). However, most of them did not specify ADPKD as an important single
disease entity. ADPKD is the most common hereditary kidney disease and accounts
for about 10% of end-stage renal disease (ESRD) cases requiring renal
replacement therapy (Spithoven et al., 2014). Early detection of ADPKD and
rigorous BP control are associated with improvement in clinical markers and delay
in the onset of ESRD (Schrier et al., 2003, 2014).
Surprisingly, however, a previous study on HTN in young adults with ADPKD
revealed that only 7% of the subjects had been screened for ADPKD (Kelleher et al., 2004). Normal findings in routine laboratory tests are not a basis for
excluding ADPKD, and ultrasound has a low diagnostic sensitivity for ADPKD
screening in individuals aged
Recent studies have shown that PA is more prevalent in patients with severe or
resistant HTN than previously thought (Fagugli and Taglioni, 2011). According
to a recent systematic review, however, there were wide variations in prevalence
and large heterogeneity in primary care settings (3.2%-12.7%, I
In the present study, of the 6373 hypertensive patients, only 5 patients
(0.08%) had an ARR
Subclinical hypercortisolism had not been determined as a cause of HTN in previous reports until its prevalence was reported to be 1.0% of all hypertensive patients in 2004 (Omura et al., 2004). Although the diagnostic criteria are complex and controversial, subclinical hypercortisolism should be considered as a cause of resistant HTN, as it is associated with increased comorbidities and mortality (Martins et al., 2012). In this study, subclinical hypercortisolism was the fourth most common cause of secondary HTN (0.19%).
Multiple logistic regression analysis revealed five major predictive factors for secondary HTN. Abnormal TFT results had the strongest association with secondary HTN (Table 3). TFT is important for the diagnosis of hyperthyroidism or hypothyroidism as a cause of HTN, and other abnormal TFT findings were also more prevalent in secondary HTN. Further studies will be required to ascertain the correlation between other abnormal TFT results and secondary causes of HTN.
Abnormal findings in urine dipstick tests for blood and protein can occur transiently and functionally for various reasons, especially in young patients. In this study, however, these were much more prevalent in patients with secondary HTN, mainly due to renal parenchymal diseases and ADPKD. Of note, even trace amounts of proteinuria or hematuria were suggestive of secondary HTN.
Being overweight/obese (BMI
Although the sensitivity and PPV of the model were unsatisfactory, the specificity (90.5%) and NPV (82.2%) were satisfactory with an accuracy of 80.5%. In other words, if hypertensive patients do not have any predictive factors, it is likely that they will not need to undergo further tests for secondary HTN.
This study has some limitations. First, as this was a retrospective study, there was a lack of unique, predefined criteria for screening secondary HTN. Second, 24-h ABPM was performed in only 31.6% of all hypertensive patients. Thus, it is possible that the total number of hypertensive patients was exaggerated by white coat HTN. Third, the EMR of only 638 patients with ‘possible’ secondary HTN were reviewed among the 6373 hypertensive patients. Fourth, obstructive sleep apnea, one of the most common causes of secondary HTN (Pedrosa et al., 2011), was not evaluated in most patients. As obesity is a typical finding in patients with obstructive sleep apnea, if it was screened as a cause of HTN in this study, excessive weight might have been a positive predictor rather than a negative predictor of secondary HTN. Lastly, the existence of selection bias should be considered when interpreting the results of this study, as the study population consisted of relatively healthy young male military personnel.
Despite these limitations, our study has several strengths. First, this is the largest prevalence study on secondary HTN among young hypertensive men. The study population comprised a certain, well-defined, homogeneous demographic group. In South Korea, nearly all men are obligated to serve in the army in their 20 s, and most of them are examined in military hospitals when suspected of having HTN. Second, as medical services at military hospitals are free of charge, individuals can undergo screening tests for secondary HTN without concerns about the cost. This was key to enabling screening for secondary HTN in up to two-thirds of general hypertensive patients. Third, all urine or blood specimens for hormone tests were analyzed at a single institution, thus increasing confidence in the test results. Therefore, our findings could be applied to an unselected hypertensive population in their early 20 s without severe comorbidities, especially in military personnel.
Secondary HTN is rare in relatively healthy hypertensive patients aged
Conceptualization: Kihyun Kim, Se-Joong Rim Data collection and analysis: Kihyun Kim Methodology: Eui-Young Choi, Hyuck Moon Kwon Draft: Kihyun Kim Revision: Jong-Youn Kim, Eui-Young Choi, Hyuck Moon Kwon, Se-Joong Rim.
The Korean Military Medical Research Project funded by the Republic of Korea (ROK) Ministry of National Defense (ROK-MND-2016-KMMRP-033) supported this work.
The authors declare no conflicts of interest.