The Effects of a Simulated Workday of Prolonged Sitting on Seated vs. Supine Blood Pressure and Pulse Wave Velocity in Adults with Overweight/Obesity and Elevated Blood Pressure

Participants

As previously described, twenty-five working-aged adults were enrolled in the study. The inclusion criteria in this study were: a) 20 – 65 years old, b) SBP between 120–159 mmHg or diastolic blood pressure (DBP) between 80–99 mmHg, which were classified as prehypertension or Stage I hypertension according to the JNC7 guidelines at the time when the study was conducted (27), and c) overweight or stage I/II obesity with a body mass index (BMI) of 25.0 to less than 40.0 kg/m². Participants were excluded for the following: a) SBP ≥ 160 mmHg or DBP ≥ 100 mmHg, b) atrial fibrillation or a cardiovascular event in the past 6 months, c) using antihypertensive, glucose lowering, or any other drugs known to influence cardiovascular responses, d) current treatment for cancer or other serious medical condition, e) current smoker, f) pregnant in the past 6 months or breastfeeding in the past 3 months, g) performing moderate-to-vigorous exercise ≥ 90 minutes per week over the past 3 months, or h) currently enrolled in a weight loss program. The study was approved by the Institutional Review Board at the University of Pittsburgh. All participants provided informed consent prior to participating in any research procedures.

Experimental procedures

Participants took part in a simulated workday of prolonged sitting. The experimental visit lasted approximately 10 hours and included morning (7:15 AM), mid-day (12:05 PM), and afternoon (4:45 PM) cardiovascular testing, standard breakfast and lunch meals, and morning and afternoon sessions of prolonged sitting (3 hours and 40 minutes each) (shown in Fig. 1). During the seated periods, participants sat continuously at a standard desk with feet on the floor and performed office work, with the exception of necessary restroom breaks (mean=3.7 ± 1.3 breaks).

Participants reported to our laboratory between 07:00–07:30 AM. Upon arriving, adherence to pre-visit instructions including abstention from food, nicotine, and alcohol for 12 hours and abstention from exercise and caffeine for 24 hours, was verbally confirmed. Next, following a 10-minute rest in a seated posture, morning BP and PWV measurements were completed. Next, participants rested in a supine posture for 10 minutes and then cardiovascular measurements were repeated. Though we considered counterbalancing the order of BP and PWV measurements across the seated and supine postures, a systematic order (seated then supine) was purposefully selected. This choice allowed us to capture the effects of prolonged sitting in the seated posture during the protocol, avoiding any potential effects due to postural interruption. Thereafter, a standardized breakfast was distributed to participants followed by a restroom break if needed. The standardized meals (breakfast and lunch) were designed to each fulfill 30% of participants’ daily caloric need during low activity level (28). These meals comprised of approximately 55–60% carbohydrates, 10–15% protein, and 25–30% fat.

Then, participants begun their morning prolonged sitting period (3 hours and 40 minutes) at a standard desk equipped with a phone and a computer with internet connection. During the seated period, participants were required to complete personal, job-related, desk-based work, with no specific restrictions. Participants were also allowed to use their personal phone or the restroom as needed. During the final 30 minutes (beginning at 3 hours and 10 minutes into the seated work period), participants were required to stay seated to limit influence of movement on cardiovascular measurements at the end of the work period.

Following the morning working session, mid-day measurements of BP and PWV were completed with the same protocol as the morning measurements. Then, participants consumed a standardized lunch followed by a restroom break if needed. Thereafter, participants completed the afternoon prolonged sitting session (3 hours and 40 minutes) followed by BP and PWV measurements with the same protocol as the morning period.

Measurements

Participants’ demographic characteristics were self-reported on standard questionnaires. Body height was measured in duplicate by a wall-mounted stadiometer (Perspective Enterprises, Portage, MI) whereas body weight was measured in duplicate by a digital scale (WB-110A, Tanita, Japan). Duplicate measures were averaged.

Blood pressure

BP was measured by the HEM-705 oscillometric device (Omron Healthcare, Inc., Lake Forest, Illinois, USA) which has previously been validated (29, 30). During eligibility screening, arm circumference was measured by a Gulick tape measure to determine the proper cuff size for each participant. Prior to BP measurement, participants sat in a clinical chair for at least ten minutes with back supported, feet on the floor, and arms supported at heart level (26). Then, BP was measured in both arms; a second measure of BP was taken in the arm that had the higher initial pressure, with one-minute intervals between all measurements. The average of the two measurements was used for eligibility. Thereafter, the same arm and cuff size were used for measurement during experimental visits. During experimental morning, mid-day, and afternoon assessments, BP was measured twice each in a seated posture and in a supine posture. The measurements were completed after at least 10 minutes of assuming each posture. The average of both measurements was used in the analysis. Mean arterial pressure (MAP) was calculated using the following formula: MAP = DBP + 1/3 [SBP – DBP]. For descriptive purposes, HR was measured concurrently by the oscillometric BP monitor and averaged across repeated measures for analysis.

Pulse wave velocity

Following the current guideline (24), carotid-ankle (caPWV), carotid-radial (crPWV), and carotid-femoral (cfPWV) PWV were measured using tonometry via the Complior Analyse (ALAM Medical, Vincennes, France). Arterial path lengths were measured as follows: cfPWV was estimated as the sternal notch-to-femoral artery distance minus the sternal notch-to-carotid artery distance; crPWV was measured as the direct distance between the carotid and radial arteries; caPWV was measured as the distance between the carotid and posterior tibial arteries. Sensors were simultaneously placed at the carotid, radial, femoral (supine only), and posterior tibialis arterial sites. Three high quality PWV scans (10 waveforms each) were obtained for each site. PWV was calculated for each scan as the difference in path length divided by the difference in time for the pulse wave to reach each site; calculated PWVs for each site were averaged. The acceptable coefficient of variation within-scan was set as less than 10% across waveforms. Scans that resulted in coefficients of variation over 10% were rejected and were not averaged. Following 10 minutes in a seated posture, caPWV and crPWV were measured from each participant. Of note, the femoral artery was not accessible when seated and, thus, cfPWV was not measured in the seated posture. Next, participants transitioned themselves several feet to a supine posture on an exam table, rested for 10 minutes, and then were measured for caPWV, crPWV, and cfPWV. A recent study from our laboratory examining the properties of crPWV and caPWV in seated and supine postures found these measures had acceptable-to-good reliability (ICC ranging from 0.636 to 0.759) but poor-to-good validity (r ranging from 0.227 to 0.812) for crPWV and caPWV across postures (31).

Blood pressure and pulse wave velocity responses across the simulated workday

Table 2 shows BP and PWV responses to the simulated workday of prolonged sitting across the entire experimental visit by measurement posture. For the currently recommended BP measurement posture, seated SBP, DBP and MAP significantly increased over time with small-to-moderate effect sizes. Though not currently recommended, supine BP was also examined and reported in Table 2. Only supine SBP significantly increased over time with a small-to-moderate effect size, while DBP and MAP did not. Additionally, seated and supine HR did not significantly change over time with both having small effect sizes. The exploratory analysis by sex revealed several significant sex-by-time interactions with prolonged sitting. Females had greater responses than males in seated DBP and MAP with prolonged sitting (interaction p<0.05); the effects of prolonged sitting on seated SBP and all supine BP were similar in females and males (Supplemental Table 1).

For the currently recommended PWV measurement posture, Table 2 shows that supine cfPWV and caPWV significantly increased over time with small-to-moderate effect sizes. However, supine crPWV did not significantly change over time with small effect size. Though not currently recommended, seated PWV was also measured for caPWV and crPWV. Neither seated caPWV nor crPWV significantly changed over time, with both having small effect sizes. The exploratory analysis by sex revealed no significant sex-by-time interactions with prolonged sitting. However, increases in supine cfPWV, caPWV, and crPWV were more apparent in males (all p<0.05) vs. females (all p>0.05) (Supplemental Table 1).

The influence of posture on responses across the simulated workday

Trajectories of BP and HR across the simulated workday by posture are shown in Figure 2. Differences in the effects of time across measurement postures were small and nonsignificant for SBP, DBP, MAP and HR (interaction β all p>0.05, reported in Fig. 2). However, there were moderate posture effects where SBP, DBP, MAP and HR were significantly higher when seated compared to supine (all p<0.05, shown in Fig. 2). When combined across postures, there were overall small-to-moderate significant time effects of prolonged sitting for SBP and MAP, but not significant for DBP or HR (shown in Fig. 2)

Trajectories of cfPWV, crPWV, and caPWV across the simulated workday by posture are shown in Figure 3. Because cfPWV was not measured in a seated posture, only the supine trajectory can be reported and was found to increase with a moderate effect size across the simulated workday (shown in Fig. 3). The difference in the effect of time across measurement postures was again trivial for crPWV and caPWV (interaction β had p>0.05). There were small and large posture main effects, respectively, for crPWV (d=0.30) and caPWV (d=1.13) where PWVs were significantly higher in the seated compared to the supine posture. In contrast, considering seated and supine together, crPWV and caPWV did not change significantly over time across the simulated seated workday with both having small effect sizes (shown in Fig. 3.).

Due to the potential influence of sex on cardiovascular responses to prolonged sitting, analyses were repeated with the addition of sex interaction terms; all were not significant (p>0.05; shown in Supplemental Fig. 1 - 2). Furthermore, the analyses of PWV responses were repeated adjusting for SBP and HR. The adjusted models revealed similar PWV responses to prolonged sitting and across measurement postures (shown in Supplemental Table 2).

Effects of prolonged sitting during a simulated workday

As hypothesized, prolonged sitting resulted in an increase in seated BP among overweight/obese individuals with elevated BP. While some studies have found comparable adverse effects following 3, 7 or 8 hours of prolonged sitting (9–12), others suggest no change in BP with 3 or 7 hours of prolonged sitting (8, 35). Speculatively, several factors such as postural positioning when measuring BP, allowance of restroom breaks during prolonged sitting, and participants’ clinical characteristics might have contributed to these inconsistent findings. However, no clear pattern emerges across these studies that explains why some studies find a deleterious effect of prolonged sitting on BP and some studies do not. Thus, further research into the mechanisms through which prolonged sitting may increase BP could be helpful for elucidating the true effects of prolonged sitting on acute BP. Importantly, when evaluating the effects of prolonged sitting on cardiovascular outcomes, studies should take into consideration these factors that can differently affect cardiovascular outcomes and clearly report prolonged sitting measurement protocols to facilitate comparison across the existing literature.

When measured using current recommendations of a supine posture (24), a simulated workday of prolonged sitting significantly increased supine cfPWV (0.52 ± 0.20 m/s) and caPWV (0.38 ± 0.16 m/s) over time in participants with elevated BP and overweight/obesity. We are aware of only two other studies that assessed the effect of 3 hours of prolonged sitting on cfPWV (8, 13). In both studies, the investigators also reported significant elevation in cfPWV following prolonged sitting. Chronic exposure to prolonged sitting and resulting acute elevations in PWV, similar to the increases observed in the Credeur, et al. (8), Evans, et al. (13), and our study, could be detrimental to cardiovascular health long term. In the current study, though cfPWV (central) and caPWV (peripheral lower extremity) increased in response to prolonged sitting when measured in a supine posture, crPWV (the peripheral upper extremity) did not significantly increase. This could reflect that the upper limbs had continued muscular action during the simulated workday (i.e., typing or writing) that could have maintained function in these vessels. This contrasts exposure to different mechanisms in the lower extremity during prolonged sitting (gravity, tortuous vessels, lack of muscular pump) that lead to blood pooling in the legs and vascular dysfunction (36, 37).

Though limited by small sample size (n=9 females, n=16 males), our data suggest that physiological sex differences could potentially be a source of variation in cardiovascular responses to prolonged sitting. Although Vranish, et al. (10) found no interaction by sex in BP or HR responses to 3 hours of prolonged sitting, we detected significant interaction by sex in seated DBP and MAP responses (females > males) to a simulated workday of prolonged sitting. Further, though no significant interaction by sex was observed, increases in supine PWV with prolonged sitting were more apparent (all p<0.05) in males than females. This finding was similar to a 3-hour prolonged sitting study by Credeur, et al. (8) in 13 males and 7 females. Further investigation of sex differences in cardiovascular responses to prolonged sitting is warranted.

Effects of posture on cardiovascular measurements during prolonged sitting experiments

Postural positioning for cardiovascular measurements during and following prolonged sitting has gained attention as a potentially important methodological consideration (8, 38). Postural positioning, specifically transitioning from seated to supine to perform cardiovascular measurements, may be enough of a stimulus to obscure or attenuate some of the effects of prolonged sitting. Thus, we compared BP and PWV responses to a simulated workday of prolonged sitting while maintaining a seated posture and then repeated assessments after a transition and 10-minute rest in a supine posture. Though prolonged sitting over time increased SBP, the magnitude of the effects were larger in the seated vs. supine measurements for all BP measures (SBP: d=0.43 vs. 0.28, DBP: d=0.30 vs. 0.13, MAP: d=0.40 vs. 0.18, respectively). As described below, the greater magnitude of seated compared to supine BP responses could be due to physiological mechanisms that promote greater seated, compared to supine, cardiovascular responses to the simulated workday of prolonged sitting.

Although, central PWV (cfPWV) was not measured in the seated posture due to the lack of access to femoral artery, peripheral PWV (crPWV and caPWV) was obtained in both postures. crPWV responses to the simulated workday did not change significantly over time when measured in either a seated or supine posture. Furthermore, caPWV had a sim ilar magnitude of increase in both seated and supine postures, but the increase was only statistically significant during the supine measurement. It is possible that the difference in statistical significance for caPWV responses is attributable to greater variability and measurement error in the seated posture, as has been previously reported.(38). Moreover, the smaller overall increases in seated compared to supine PWV could potentially be due to a ceiling effect, where high values of PWV are less likely to increase further (39, 40). Nevertheless, we are unaware of other studies comparing PWV responses to prolonged sitting measured in different postures; future studies are needed to explore the mechanisms that may explain differences in PWV responses in seated versus a supine posture.

Plausible mechanisms

Several mechanisms, such as alpha (mainly found in vessels) and beta (mainly found in the heart) sympathetic overactivation, which can lead to increased BP, PWV and HR (41, 42), have been proposed to explain the effects of prolonged sitting on cardiovascular function (43). One theory suggests that assuming a quiet, seated posture where the trunk and lower extremity are vertically positioned on a chair and floor, respectively, results in blood pooling in the lower extremity due to gravitational force and diminished muscular pump. This leads to a decrease in venous return (preload), ventricular filling, cardiac output and mean arterial pressure (22). To compensate, neural mechanisms, specifically alpha and beta receptors, are activated leading to decreases in venous compliance and increases in systemic vascular resistance and heart rate (22, 44, 45). Over a period, these mechanisms may be overactivated which may lead to increased BP, HR and, potentially, PWV. In this study, we observed increases in seated SBP, DBP and MAP, which may imply increased activation of alpha sympathetic receptors. We also noted that seated HR did not significantly change and, if anything, tended to be lower at the time of cardiovascular assessments following the simulated workday of prolonged sitting. This HR response could potentially indicate that beta sympathetic receptors did not change; nevertheless, because HR is regulated by both branches of the autonomic nervous system (sympathetic and parasympathetic) which can function reciprocally or non-reciprocally (46), caution should be taken when interpreting HR response in this study. It is possible that the beta sympathetic activity has increased but is accompanied with equivalent increases in parasympathetic outflow to the heart, leading to unchanged HR. However, without more direct measurements of autonomic nervous system activity, this hypothesis cannot be confirmed from our study and warrants further investigation.

We hypothesized that these mechanisms and cardiovascular responses would be different when measured in a supine compared to seated posture. We expected that interrupting prolonged sitting by transitioning to a supine posture could interfere with these mechanisms or promote counterpart mechanisms. For example, in a supine posture, the whole body is vertically positioned on a bed or table and gravitational forces are distributed across the whole body. This is in contrast to the gravitational force during sitting that would be more strongly exerted on the lower extremity. Consequently, transitioning from a seated to supine posture for vascular assessments could promote recirculation and interfere with pooled blood in the lower extremity that is observed during prolonged sitting. In that case, the neural responses, specifically alpha and beta sympathetic activity, would be different. In fact, several studies have demonstrated greater vascular (alpha) sympathetic activity in muscles when seated compared to supine (47, 48). While we did not test this mechanism explicitly, our findings are consistent with this hypothesis. Though no significant interactions were observed as mentioned earlier, the magnitude of the effects of prolonged sitting on SBP, DBP and MAP were about twofold greater when measured in the seated compared to supine posture. This could be due to greater sympathetic activation when seated compared to supine as observed in other studies (47, 48).

Limitations and strengths

There were limitations that should be considered when interpreting our findings. Our sample included both sex groups (16 M; 9 F) and a wide range of age (42±12 years old) to increase the generalizability of the findings. Yet, this may have introduced some variability and our small sample size in this secondary analysis may also have led to a lack of statistical power.

The lack of a comparison control condition for evaluating overall responses to prolonged sitting is another limitation. In acute physical activity or other behavioral research, sitting is often the control condition. However, due to recent growing evidence showing the potential detrimental effects, prolonged sitting may now be considered as the experimental condition. Further, because this study aimed to evaluate the effects of a simulated workday of prolonged sitting, participants were only instructed to remain seated without any further restrictions on their lower limb motion or mode (i.e., not instructed to avoid fidgeting or leg crossing), which could have affected cardiovascular responses (37, 49).

The postures, by logistical limitation of needing to maintain the prolonged seated posture, could not be counterbalanced. All measurements were always completed in the same order, seated followed by the supine measurement, and this could have introduced systematic bias. During this postural transitioning (from seated to supine), participants took a couple of steps from the chair to the bed for assessment. This transitioning, while likely a part of many other study protocols like ours, could have interfered with the true effects of prolonged sitting on the cardiovascular outcomes beyond the simple postural change. Lastly, measurement error was more likely to exist during the seated caPWV measurement due to the small size and anatomical position of the vessel which creates more technical challenges. This was reflected in the standard errors when comparing seated vs. supine caPWV.

Though we did not evaluate the postmenopausal status or control for menstrual cycle phase for female participants as this is not typically included as a recommended practice for the measurement of BP and PWV (24, 26), these could have affected cardiovascular responses to the simulated workday of prolonged sitting (50–52) and are thus a limitation of our study.

The study also has several strengths that are worth mentioning. The study evaluated cardiovascular responses to a simulated workday (> 7 hours) of prolonged sitting which mimicked a typical workday and contrasts other studies that have utilized shorter duration (≤ 3 hours) of prolonged sitting (8, 10, 45, 53). This study also compared cardiovascular responses to prolonged sitting when measured in seated and supine postures, informing protocol design for future prolonged sitting studies. Most of the previous studies evaluated cardiovascular responses to prolonged sitting in only one posture (9, 44, 45, 53). Further, our study had direct observation of posture and activity, dietary intake was standardized, and we used gold standard, objective assessment of cardiovascular responses.