METHODS
Study population
The study population consisted of participants of the Seventh Tromsø Study (Tromsø7) in Norway and the Know Your Heart (KYH) study in Russia. Both studies were cross-sectional and population-based. Tromsø7 was conducted in the Tromsø municipality between March 2015 and October 2016, and KYH from 2015 to 2018 in Arkhangelsk and Novosibirsk. The two studies were conducted in parallel and included questionnaires, health examinations, and biological sample collection. During the development phases, several aspects of data collection between the studies were harmonized, including the echocardiography protocols. Echocardiograms were performed in 2340 participants from Tromsø7 and 4521 from Russia.
In Tromsø7, the inclusion age range was 40 years and older, without an upper limit, whereas KYH included participants aged 35–69 years. For the present study, a stratified random selection of participants was made, giving three equal-sized age groups (40–49, 50–59, and 60–69 years old). As shown in Figure 1, the sample comprised participants from Tromsø7 (50%), Arkhangelsk (25%) and Novosibirsk (25%). A total of 1194 participants from KYH (594 women and 600 men) and 1013 participants from Tromsø 7 (553 women and 460 men), aged 40–69 years, were selected.
Definition of study groups
Figure 1 displays a flow chart of in- and exclusion of study participants. Individuals with cardiac disease that could impact myocardial function were excluded from the study population. Therefore, we excluded subjects with valvular heart disease (aortic insufficiency grade 3 and 4); aortic valve mean pressure gradient >25 mmHg; mitral insufficiency degree 3 and 4; moderate and high-grade mitral stenosis; history, or objective indicators of previous coronary artery disease (classes 1.1 – 1.2.7. of Minnesota Code) or myocardial diseases; ECG with QRS > 130 ms, EF < 45%.
Hypertension was defined as either increased systolic or diastolic BP during the visit or the current use of antihypertensive medication. The study population was divided into groups using the following criteria: systolic BP ≤ 140 mmHg and diastolic BP ≤ 90 mmHg without antihypertensive medication (group A, normotensive); systolic BP ≤ 140 mmHg and diastolic BP ≤ 90 mmHg with antihypertensive medication (group B, controlled hypertension); systolic BP 140–159 mmHg and/or diastolic BP >90 mmHg (group C), and systolic BP ≥ 160 mmHg (group D). Groups C and D were defined as hypertensive irrespective of the antihypertensive medication.
Participants were classified as taking antihypertensive medications when self-reported to be currently taking renin-angiotensin system drugs, beta-blockers, calcium antagonists, or diuretics.
Data collection and echocardiography in KYH and Tromsø7
The transthoracic echocardiography was performed in the left lateral decubitus position using commercially available GE Healthcare systems: Vivid q equipped with a 1.5–3.6 MHz sector matrix transducer in KYH study and high-end machine E9 with single crystal matrix sector probe of 1.5–4.6 MHz in Tromsø7. From parasternal and apical views, two-dimensional (2D) grayscale images and pulsed, continuous, and color Doppler data were acquired. Both studies included apical four-chamber (4CH) and two-chamber (2CH) views, while apical long-axis views (APLAX) were only acquired in Tromsø7. 2D-images were obtained at a frame rate of at least 50 fps. The commercial software EchoPAC (v.203, GE-Vingmed AS, Horten, Norway) was used for strain analyses. Intra- and inter-observer variability for conventional echocardiographic measures were regularly assessed within both KYH and Tromsø7 reading laboratories and compared between laboratories. Conventional echocardiography included left ventricular systolic and diastolic volume (LV ESV; LV EDV), stroke volume (SV), ejection fraction (EF), and left atrial volume (LAV) measurements using the Simpson biplane method. Doppler-derived measurements included mitral valve (MV) E, A, E/A ratio, and deceleration time (MV DT). The M-mode was used to estimate the septal wall thickness and myocardial mass. SV, myocardial mass, and left atrial (LA) volume were indexed by body surface area (BSA).
Strain and strain-rate analysis
A single reader (M.K.) analyzed all strain data using the Q-analysis function of EchoPAC. For all analyses, the peak R was set as the endpoint of end-diastole. Aortic valve closure was defined as a transaortic CW Doppler signal. The region of interest (ROI) was defined by manual definition of the subendocardial border and adjustment of the ROI width. Automated tracking was visually controlled, and suboptimal tracking results were repeated a maximum of three times. From the 12 (or 18) segments, the following segmental values were extracted from the strain analysis: segmental end-systolic (ES) subendocardial (endo), mid-myocardial (myo), and epicardial strain (epi). Furthermore, the peak systolic SR (SR S), peak SR E and SR A, and respective peak global S/SR values for two or three apical views were extracted using the software. Post-systolic stretching (PSS) was calculated as the difference between the ES strain and peak diastolic strain. Segments were classified as PSS present or absent by defining cutoff values (-1 to -3% and > -3%).
S/SR from APLAX views was analyzed in 176 Tromsø7 participants. Based on previous analyses of this study population, we assigned segmental groups with similar values, that is basal septal, basal, mid, and apical segments.
Artefact reading
Artifact detection was used to identify distorted echocardiographic records that were to be excluded from analysis. The identification of strain-curve artifacts was based on artifact detection by visual assessment of the strain curves and was described in detail in a previous publication (12) on the same study population. Strain curve artifacts were subjectively assessed using these previously described features and classified as (1) “blunted curves” (a reduced or even positive strain in the start of the cycle, mirrored by a similar curve-formation at the end of systole), (2) “diastolic mismatch” (late diastolic strain curve significantly deviating towards 0 or positive values compared to the late diastolic strain-curves of other segments), and (3) “floating” (segmental strain curves with several negative and positive peaks without correspondence with timing or configuration of other segments). Segments with curve artifacts or apical foreshortening were discarded. As shown in Figure 2, apical foreshortening and curve artifacts reduced the number of echocardiograms included in the final analysis.
Blood pressure (BP) measurements
In KYH and Tromsø7, blood pressure was measured in the sitting position in a quiet room, either before or after the echocardiogram, but not on the same day. The measurements were performed three times at 1-minute intervals after an initial 2-minute seated rest. The results of the second and third measurements were averaged and used for the analysis. All study participants were asked about the medications they were currently taking, and the data were coded using the Anatomical Therapeutic Chemical (ATC) classification system. Any medications within ATC classes C02, C03, C07, C08, or C09 were regarded as antihypertensives.
Statistical analyses
If not stated otherwise, continuous variables are presented as mean ± standard deviation (SD). Variables with skewed distributions are presented as medians with quartiles (Q1–Q3). Categorical characteristics are presented as absolute numbers and proportions (%). Between-group differences in continuous variables were tested using one-way analysis of variance (ANOVA) with Bonferroni post hoc tests, and a χ2 test was used for group comparisons of categorical variables. Multiple linear regressions were used to assess associations between hypertension status and echocardiographic parameters with adjustments for potential confounders (age, sex, height, body mass index (BMI), heart rate (HR), atrial fibrillation (AF), and KYH/Tromsø study population). All multiple regression models were run on the sample with complete data for all covariates (n  = 1707). Statistical significance was set at p < 0.05. SPSS v28.0 (IBM Corp.) was used for statistical analyses.
Intra-and interobserver-variability
The intra- and inter-observer variabilities have been previously reported (12). For intra-and inter-observer variability in strain and SR measurements, the same observer repeatedly analyzed 135 randomly selected echocardiographic records comprising 1620 segments within 6–12 months from the initial analysis. The same data were reanalyzed by a second experienced observer. Intra- and inter-observer values were calculated as Bland–Altman limits of agreement for segments and discarding segments with curve-artifacts.