Orthostatic hypotension (OH), a sustained fall in blood pressure (BP) upon standing, is among the most debilitating manifestations of autonomic dysfunction in Parkinson’s disease (PD). Classical OH is defined as a sustained reduction of at least 20 mmHg of systolic blood pressure (SBP) or 10 mmHg of diastolic blood pressure (DBP) within 3 min of standing or the head-up tilt-table testing (Freeman et al., 2011). Currently, the reported incidence of PD-OH varies from 11.1 to 51.6% (Hommel et al., 2016; Li et al., 2019; Quarracino et al., 2020). The prevalence of OH in PD increases with age and disease duration (Li et al., 2019). OH can impair the blood supply to organs above the heart, such as the brain, resulting in symptoms related to tissue hypoperfusion. The characteristic symptoms include lightheadedness, vertigo, presyncope, and syncope. Symptomatic OH increases the likelihood of falls and has been proven as an independent risk factor of mortality (Masaki et al., 1998; McDonald et al., 2017). There is a significant linear association between the change in systolic BP from supine position to standing and the 4-year mortality rates (Masaki et al., 1998). Furthermore, daily activities and quality of life of patients with PD-OH are significantly compromised relative to patients with PD without OH (Li et al., 2019).

Several confounding variables may influence the extent to which orthostatic BP falls, such as gender, age, hydration, blood glucose, deconditioning, and anemia (Freeman et al., 2011; Perez-Lloret et al., 2012; Li et al., 2019). OH is also independently associated with PD-related parameters, such as the PD phenotype of postural instability and gait disorders (PIGD), lower Mini-Mental State Examination score, longer follow-up time, and higher levodopa equivalent dosage (Hiorth et al., 2019).

However, there have been few studies investigating the longitudinal orthostatic BP changes of patients with PD. BP trajectory could facilitate the prediction of hypotension and hence may be useful in OH prevention. This study was based on the data from the Parkinson’s Progression Markers Initiative (PPMI). The aim of this study was to identify latent subgroups of orthostatic BP trajectories and to investigate the associated factors that influence the different trajectory types. The analysis of orthostatic BP change was based on SBP because the systolic criteria seem to be sufficient to identify 95% of subjects with OH (Fedorowski et al., 2017). In addition, the absolute magnitude of changes in SBP is larger and easier to measure than DBP and as such is more accurate.

We identified subgroups of ΔSBP trajectory in patients with PD based on a longitudinal analysis of PPMI data. The best fit was identified for a three-class model, reflecting three subgroups of the ΔSBP change. Class 1 was characterized by rapidly increasing ΔSBP. Class 2 displayed a lower level of ΔSBP throughout the follow-up, while Class 3 was characterized by a greater ΔSBP at baseline, which kept relatively stable over time. The main predictors of rapidly increasing ΔSBP in PD were male sex, lower supine DBP, and lower level of total protein at baseline.

Orthostatic hypotension is associated with target organ damage and typically manifests with recurrent syncope and orthostatic intolerance. Previous research has demonstrated that OH increases the risk of falls in patients with PD and is associated with cognitive impairment and greater deterioration in daily living activities, regardless of whether or not OH is symptomatic (Merola et al., 2016, 2018; Romagnolo et al., 2019). Our results show the differences in OH trajectories of patients with PD and in particular highlight those that are worth clinical attention, especially male patients with PD and lower supine DBP and lower total proteins. These patients tend to develop OH and are more prone to falls. As such, this patient group may greatly benefit from an early management strategy that includes advice such as changing positions slowly, particularly upon standing.

In previous studies, there are contradictory results about the association between disease duration and OH. Wüllner et al. (2007) revealed a significant correlation of OH and clinical course through logistic regression analyses. Li et al. (2019) found a difference in disease duration between patients with PD with or without OH; however, disease duration was not found to be a statistically significant influencing factor in PD-OH, with a p-value of 0.718 in further logistic regression. In a longitudinal study conducted by Jost and Augustis (2015), there was a lack of a correlation between the severity of OH in PD and disease duration. Nevertheless, in clinical practice, a subset of patients will develop OH during the course of PD. Our findings indicate that, for the majority of patients with PD, ΔSBP remains stable over the disease course. However, in 2.7% of patients, there was a rapidly increasing ΔSBP over the disease course, which may be very difficult to detect in an analysis of the total population.

Clinical variables have been shown to play different roles in cross-sectional studies of PD-OH, as well as in the longitudinal trajectory of PD-OH. In the previous studies, PD-OH has been significantly related to older age, lower body mass index (BMI), more severe motor symptoms, higher LEDD, and higher blood glucose (Perez-Lloret et al., 2012; Nakamura et al., 2017; Li et al., 2019). Similarly, our results showed that older patients with higher MDS-UPDRS-III scores, indicating severe motor symptoms, tend to have higher ΔSBP. In Nakamura et al. (2017), it was emphasized that weight, rather than height, drove the association between BMI and OH. In contrast, however, we found the height of Class 2 to be significantly shorter than that of Class 1 and Class 3, and no association of weight or BMI with OH was identified. Campese et al. (2021) also reported a negative association between classic OH and shorter stature in male patients with PD and considered that the non-neurogenic mechanisms may prevent shorter individuals from BP declines on standing. Considering that some PD medication may be risk factors for OH, we compared the baseline LEDD and the average LEDD during follow-up years among the three classes. No significant difference was found between groups, indicating that LEDD does not change the natural trajectory of ΔSBP. In addition, non-motor scores, BP at baseline, and some laboratory parameters were also different among classes. For example, the patients in the high-stable class had more severe RBD and olfactory dysfunction than those in the low-stable class. It seemed that some patients with PD had remarkable non-motor symptoms at the early stage of PD, indicating the presence of a particular internal phenotypic cluster. It was reported that the combination of olfactory dysfunction, OH, and RBD was associated with a malignant phenotype of PD characterized by more rapid progression of cognitive deficits and postural instability, which could result from the pathological expansion of PD from the olfactory pathway and along fiber tracts in the brain stem simultaneously (Pilotto et al., 2019; Nomura et al., 2020).

In PD, gender is significantly related to multiple motor and non-motor symptoms. In some previous studies, no correlations were identified between OH and gender in patients with PD (Wüllner et al., 2007; Bae et al., 2011). OH was reported for 10% of women and 11% of men in Wüllner et al. (2007), without significant correlation of OH with gender through logistic regression. In contrast, Velseboer et al. (2017) found male sex was associated with the presence of OH in PD. We also found that male sex was one of the predictors of increasing ΔSBP. This finding in patients with PD is distinct relative to the general population, where it is reported that the incidence of OH is higher in women than men (Fedorowski et al., 2009), due to higher estrogen levels, a more active parasympathetic system, lower sympathetic activity, and an attenuated venous distension reflex in women (Cheng et al., 2011; Mansur et al., 2019). Estrogen decreases BP by lowering the plasma renin activity, whereas testosterone decreases BP by increasing the plasma renin activity (Reckelhoff et al., 1998). However, PD is a neurodegenerative disorder with an average onset of approximately 60 years (Obeso et al., 2000), and the median age of patients with PD in this study was 63 years. This stage of life is associated with decreased levels of estrogen and testosterone in women and men, respectively. In addition, older women show a higher sympathetic and lower parasympathetic activity at rest compared with age-matched men (Sachse et al., 2019), providing a possible explanation as to why male patients with PD have a relatively higher incidence of OH.

Another key predictor of increasing OH incidence is supine DBP. Although a decline in diastolic BP during orthostasis may be less relevant in the clinical diagnosis of OH, its potential impact on long-term prognosis requires attention. The diastolic component of BP reflects the real pressure of coronary perfusion as the coronary flow is substantially reduced during systole (Smit et al., 1999). Decreases in DBP may lead to periodic myocardial hypoperfusion and increased orthostatic change in BP. One study has shown that DBP was inversely associated with the risk of myocardial infarction (Fedorowski et al., 2014). Moreover, OH with severe DBP decline is a powerful independent predictor of mortality (Lagro et al., 2012).

Malnutrition is also an important risk factor for OH (Kocyigit et al., 2021). It has been reported that the total protein level is associated with nutritional status in patients with PD (Yang et al., 2020). Kocyigit et al. (2018) revealed that both malnutrition and malnutrition risk were correlated with systolic, but not diastolic, OH. Malnutrition may lead to a reduction in muscle mass and muscle tone (Reijnierse et al., 2015) and bring about impairment in peripheral vasoconstriction, which further causes systolic OH. Our results show that total protein was one of the predictors of increasing ΔSBP. Lower levels of total protein indicated a higher risk for OH. As another parameter for nutritional status, albumin also tended to decrease in Class 1 but without significant difference among groups.

The strengths of our study include a large well-characterized sample, careful screening, comprehensive data collection and processing, and advanced analysis of longitudinal information. However, some methodological limitations ought to be recognized when interpreting the findings. First, the analysis was based on observational data collected at multiple centers, and patients had various confounding factors. Second, there were many other drugs except for PD medication related to OH not included in the analysis. Third, the participants in PPMI were mostly patients with de novo PD with short follow-up durations, so, the data for late-stage PD was insufficient. Fourth, the conditions for BP measurement were not rigorously defined. For example, room temperature during BP measurement was not uniform across different centers. Fifth, supine BP was measured after 1–3 min of quiet rest, which was shorter than the standard procedure requirement of 5 min (Lahrmann et al., 2006). Inadequate rest may affect the point measurements but influence the trajectory analysis mildly.

This study provides new information on the longitudinal development of ΔSBP in patients with PD with distinct trajectories of rapidly increasing, low-stable, and high-stable class. Male sex, lower supine DBP, and lower total proteins helped to identify the class with increasing risk for OH. Further research is needed to discover the biological mechanisms that explain these subgroups and guide future preventive interventions for OH.

Data used in the preparation of this article were obtained from the Parkinson’s Progression Markers Initiative (PPMI) database (www.ppmi-info.org/access-dataspecimens/download-data). For up-to-date information on the study, visit ppmi-info.org.

Each participating PPMI site received ethical approval before study initiation and written informed consent was obtained from all participants.

KC and YaZ designed the study. KC and KD collected the data. KD and YiZ performed the statistical analysis. KC and YG analyzed the results and drafted the manuscript. All authors contributed to the manuscript revision and read and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.