The gravitational hypothesis of blood flow distribution in the lung has been a cornerstone of pulmonary physiology, influencing both the interpretation and direction of studies related to pulmonary function for the past four decades. The model has been taught to generations of students because of its reliance on readily understood physical principles that yield elegant explanations of lung function. Modern studies, however, using high-resolution methods and experiments performed in microgravity provide observations that cannot be explained by the gravitational model alone. A new fractal model explains these anomalies, providing insights and new directions for investigation.
The opponents have constrained the discussion to human physiology by reasoning that there are differences between quadrupeds and bipeds (18). The differences are real, with respect to vascular compliance and muscularity, but of course they do not necessarily mean that observations in animals are not relevant to human physiology. The first study by Banister and Torrance (3) proposing a gravitational model and many other studies validating that model were performed in animals (e.g., Refs. 19, 28). The restriction to human lungs is immaterial as there are a number of modern studies demonstrating that gravity is not the dominant factor determining the distribution of pulmonary blood flow in humans (17, 20, 22–24).
The foundation for this debate must be set on two central issues. First, is the understanding that the spatial resolution of the methods used to measure pulmonary blood flow are central to interpreting the data. Perfusion distribution in the lung is composed of variability along a vertical axis and within horizontal (isogravitational) planes. The methods used to form the basis of the gravitational model used external scintillation counters that measured a mean flow value within horizontal planes (2, 27). None of these studies could therefore detect blood flow heterogeneity within these planes. By mathematically averaging flows within horizontal planes, the spatial resolution of modern studies can be reduced to that available in the first studies reporting a vertical gradient in pulmonary blood flow. When the spatial resolution is reduced in this manner, the distribution of blood flow obtained with modern methods appears virtually identical to the original studies of West and colleagues (19, Fig. 1) However, when blood flow distributions are viewed at a higher resolution, isogravtitational heterogeneity is observed and the relative influence of gravity on regional blood flow becomes secondary (Fig. 2). At even higher spatial resolution, the heterogeneity of perfusion within horizontal planes increases further (16), while the vertical gradient remains unchanged.
Prior to the 1990s, all human studies used data acquisition systems that did not permit the measurement of perfusion heterogeneity within horizontal planes. Hence, only the variability in blood flow over a vertical distance was seen and gravity was reasonably proposed as the sole mechanism accounting for this observation. However, with the realization that isogravitational perfusion heterogeneity exists, these older data can no longer be used as evidence that gravity is the major determinant of pulmonary blood flow distribution.
The second issue is that the lung is an elastic structure and gravity causes lung parenchyma to stretch at the top and compress at the bottom. Consequently, blood flow will appear to be greater toward the lung bases because of the increased density of blood vessels within the parenchyma (17). The pioneering studies that formed the basis of the gravitational model did not account for tissue compression down the lung. Subsequent studies correcting for this tissue redistribution (e.g., Ref. 1), confirmed a vertical gradient of perfusion, but found less difference between the top and bottom of the lung. This Point:Counterpoint is focused on the effects of gravity on blood flow distributions within the lung parenchyma (independent of tissue redistribution).
Every human study that has used high spatial resolution methods and corrected for the vertical compression of lung tissue, has found that gravity is not the major determinant of regional blood flow. While they have all reported a small vertical distribution of blood flow there is also considerable perfusion heterogeneity within horizontal planes. Using electron beam computed tomography, Jones and colleagues (20) found that vertical height accounted for 21–41% of perfusion distribution in supine and prone humans. Hopkins et al. (17) used magnetic resonance imaging (MRI) in supine humans and found that between 4 and 13% of perfusion heterogeneity was explained by height up the lung. Unfortunately, modern imaging methods do not allow humans to be studied in the upright posture where gravity has a greater influence. Bipedal primates can be studied in multiple postures, including upright, using the microsphere method (9) and it was estimated that 7, 5, and 25% of perfusion heterogeneity was due to gravity in the supine, prone, and upright postures, respectively. Modeling studies, using high resolution measurements of the human pulmonary vascular tree, also predict that gravity is unlikely to play a major role in determining regional blood flow (4).
Indirect measures can be used in upright humans to assess blood flow distributions. While there is no spatial information in these measures, temporal differences in exhaled gas profiles are attributed to topological differences in the lung. Under conditions of sustained microgravity, Prisk et al. (23) estimated that perfusion heterogeneity was reduced to ∼60% of that on earth, indicating that the majority of perfusion heterogeneity persists in the absence of gravity. Montmerle and coworkers (22) used similar methods to explore perfusion heterogeneity during transient microgravity in humans and concluded that more than one-half of indexes of perfusion heterogeneity at 1 G are caused by nongravitational mechanisms.
Despite the large degree of perfusion heterogeneity within horizontal planes, it is still possible that gravity could be the major determinant of pulmonary blood flow if this heterogeneity was a random process. We clearly demonstrated that the observed isogravitational heterogeneity is not random but rather that neighboring lung regions have similar magnitudes of flow (8) and that this pattern remains similar down to the level of gas exchange (11). We also demonstrated that high-flow pieces remain high flow and low-flow regions remain low flow regardless of posture (12) or gravitational influence (13). In addition, we showed that patterns of perfusion distribution are stable over days (14) and with growth from birth to near maturity (10). These observations have lead to an alternative model based on fractals (15, 25, 26) that incorporates the geometry of the vascular tree as an important determinate of pulmonary blood flow distribution. This asymmetrically branching model not only explains isogravitational heterogeneity and spatial correlation of perfusion but also provides a mechanism by which the geometry of the vascular tree can be generated from the genetic code. We most recently reported that two-thirds of the variability in perfusion distribution is genetically determined (7).
Not surprisingly, there has been resistance to the displacement of such a long-held concept as the gravitational model of pulmonary perfusion. Paradoxically, this skepticism plays an important role in scientific progress as it ensures that scientists will only accept a model once it has been thoroughly tested and can then be confident of its fidelity when accepted (21). Common ground between the gravitational and fractal models is developing. Recent studies are using both models to interpret data (17, 20). Review articles acknowledge the importance of vascular structure (6) and pulmonary textbooks (5) are teaching the fractal model. We hope that the fractal model will be questioned by future observations and new models proposed, so as to continually advance our understanding of basic pulmonary physiology.
- Copyright © 2008 the American Physiological Society