| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
PRESIDENTIAL ADDRESS |
From the Department of Psychiatry, University of California, San Diego, CA.
Address reprint requests to: Dr. Joel E. Dimsdale at UCSD, 9500 Gilman Drive, La Jolla, CA 92093-0804. Email: Jdimsdale{at}ucsd.edu
ABSTRACT
Ethnicity impacts the course of illness and medical treatment. There are enormous ethnic differences in care delivered to patients who are treated for pain. However, in addition to these social forces, there are enormous ethnic differences in physiological response to stimuli as diverse as diet, exposure to agonist infusions, or habitual patterns of response to stressors. The author’s clinical research studies on this topic for the last 20 years are reviewed in this article.
Key Words: ethnicity • pain • race • hypertension • sodium • life stress
Abbreviations: BP = blood pressure; PCA = patient-controlled analgesia; PRN = as needed medication; SNS = sympathetic nervous system.
I am deeply honored and yet also humbled to write this Presidential Paper. My topic concerns the legacy of our past because that past experience illumines and shadows the future. With your forbearance, I would like to apply this question to a difficult subject, the impact of ethnicity on health. Because this is such an explosive topic, I shall be reasoning from two rather different perspectives: empirical research and a perspective drawn from literature, history, and art. Rather than rely on "usual scholarly sources," I shall also be citing Homer, Melville, Falkner, the native American author William Least Heat-Moon, J. M. W. Turner, and Casper David Friedrich. Medicine could do far worse than rely on such sources for guidance on this contentious topic.
I have entitled this paper somewhat provocatively, "Ethnicity—stalked by the past." Ethnicity is important for four reasons. First, the demography of the United States is extraordinarily diverse. I grew up in a predominantly white community in Sioux City, Iowa, where social demarcations reflected religion, and ethnicity was not particularly relevant. That sort of ethnically homogenous community is about as relevant today as a sliderule. Through immigration, differential birthrates, and seasonal migrating workers, there is not a community in the United States where ethnicity is so unidimensional. Currently, 10% of US residents are foreign-born, and 14% of residents >5 years old speak a language other than English at home (1).
A second reason for studying ethnicity is that we sicken and die from different illnesses. Figure 1 summarizes Centers for Disease Control (CDC) data that demonstrate the remarkable influence of ethnicity on health (2). There are significant ethnic disparities in death rates from all major disease groups. To use a computer metaphor, our culture has a very strong "program" determining the impacts of ethnicity. One of the more challenging questions is deciphering the code of that program. How is it encoded and decoded so consistently?
|
The fourth reason for studying ethnicity is that it gives one a unique view on pathophysiology. Drugs such as antihypertensive agents are not equally efficacious in people of different ethnicities, and such differences may reveal fundamental insight in illness pathophysiology (5). Not just prescription drugs but abusable drugs as well have different pharmacokinetics in different groups. For instance, blacks manifest a higher nicotine intake per cigarette and also a slower clearance of cotinine (a nicotine metabolite) in response to cigarette smoking (6).
Ethnicity basically devolves to phenotype. Paul Baker (7) reminds us that phenotype represents a composite of culture, genotype, and the physical environment; when we talk about ethnicity, all of those factors come to play. But there are peculiarities in self-report of phenotype that make ethnic studies somewhat challenging. There are many individuals whose ethnic phenotype is puzzling even to themselves. For instance, United States Bureau of Census data suggest that approximately one third of the households self-report a different ethnicity during successive census samplings, even when the same occupants live in the same dwelling (8)! Thus, this issue of phenotypic recognition is not so simple. It is all the more remarkable that ethnicity, however shakily recognized and defined, exerts such a powerful influence on health.
At its heart, ethnicity involves the legacy of the past, and that legacy comes from many sources of memory. There is memory in our genes. William Least Heat-Moon recognizes this legacy by speaking "of genetic inclinations whereby ancestors seem to stalk our blood" (9). Memory also comes, of course, from our personal life experiences and our families’ experiences. As Falkner points out "the past is not dead; the past is not even past." Thus, there are reverberating levels of history and ethnicity that influence health on multiple levels.
I would like to discuss this area using two rather different approaches to studying ethnicity and health. The approaches generate rather different sorts of conclusions. The first study points to the enormous social footprint of ethnicity on health care and the second points just as strongly to the massive biologic footprint of ethnicity on health.
PAIN AND ETHNICITY
There is a sizable literature concerning pain and ethnicity. If one examines pain imposition in a laboratory setting, there are not strong, consistent ethnic differences in this context (10). On the other hand, in clinical settings, one cannot help but be impressed by differences in responses to pain. There are stoics and hysterics, and observers have long felt that these expressive styles are not randomly distributed across the globe.
POSTOPERATIVE PAIN
We were interested in examining the effect of ethnicity on postoperative management of surgical pain and also on patient controlled analgesia. The postoperative pain study entailed merging a number of databases and identifying a consecutive series of patients. We studied 250 consecutive patients treated with open reduction and fixation of a limb fracture (11). We obtained the operative log from the operating room. From this database, we obtained data pertaining to the complexity of the surgery—variables such as operative time and units of blood transfused. We also measured the postoperative analgesics received in the first 7 days and converted them into morphine equivalents. This information came from the hospital pharmacy computer. Finally, we retrieved information about age, gender, and social class from the hospital admissions computer. Although not a high-tech problem, the merging of these various databases presented not insignificant challenges.
Unlike simple casting for a fracture, open reduction and fixation is a complex procedure attended with considerable postoperative pain. In this study of 250 patients, we saw some startling differences in the amount of narcotics that patients received. As summarized in Figure 2, whites received significantly more narcotics than did Hispanics or blacks. Furthermore, this difference was not attributable to injury severity, social class, or any of the other logical confounds we could think of. We were left with the rather disturbing picture that PRN medication was being provided rather differently to patients of different ethnicities.
|
The standard postoperative analgesia is administered PRN. With a PRN medication regimen, there is a dialog between two actors. The patient has to request medication for pain and the nurse has to respond and agree that the medication is indicated. We wondered who was determining this ethnic variation in postoperative medication. That information was not accessible from the data sets. However, a similar study with PCA seemed to offer promise at disentangling this complex behavior. With PCA, the physician programs a small pump to administer a given dose of analgesia within a specified amount of time, whenever the patient pushes the button on the pump. The pump does not allow an excessive analgesia dose to be administered, and, more to our point, the nurse is out of the picture.
In this second study, we recruited 454 patients who required PCA. We obtained self-ratings of pain; we measured the analgesia that was self-administered; we measured the analgesia prescribed; and we controlled for the likely sorts of confounds. As Figure 3 demonstrates, again we saw striking ethnic differences in analgesia (12). The Hispanics and Asians received less medication. The curious thing is that the patients rated their pain similarly, and they pushed the PCA pump button similarly. The issue was that the pumps were programmed differently by the physician from the beginning to deliver different amounts of medicine per button press. This behavior was unintended on the doctors’ part, and yet it was apparent on a sizable sample of >400 people. There are similar studies performed in recent years, which show the remarkable effect of ethnicity on physicians’ and nurses’ behavior in determining response to patients’ pain (13, 14).
|
HYPERTENSION IN BLACKS AND WHITES
One of the most visible instances of the impact of ethnicity on health involves the noteworthy difference in prevalence and severity of hypertension in blacks and whites. Accounting for this disparity in rates has been impossible. All of the risk factors for hypertension have been proposed as culprits—genetics, nutrition, life experience, and personality. Some evidence can be marshaled in favor of each of these areas of risk transmission, and indeed it is probable that no one of these areas accounts for the phenomenon in isolation. It is known that black men and women have two to four times higher rates of hypertension than do white men and women (15). It is also known that blacks have a higher degree of end-organ disease associated with their hypertension than do whites. Aside from this epidemiological knowledge, there is a fog of uncertainty about explaining the observation. Because these black/white differences are generally not observed in rural Africa today (16), one wonders what is contained in the legacy of slavery?
SALT AND SLAVERY
As a beginning, one must suspect nutritional factors, particularly sodium. In this and many of the following analyses, we contrasted observations obtained during a high-salt diet in hospitalized subjects with observations obtained during a low-salt diet. The high-salt diet provided 200 mEq of sodium per day, and the low-salt diet provided a very small amount, 10 mEq of sodium per day. We used a sequence-controlled crossover study, where approximately 60 patients were studied after consuming each diet for a number of days. Potassium intake and calories were held constant, and dietary compliance was monitored at the clinical research center. We examined four subgroups of patients—blacks and whites, normotensives and hypertensives. All hypertensives had been tapered off their medication for weeks in advance of their participation. We contrasted the resting blood pressure on the two diets. The ÄBP (high salt minus low salt) was calculated for each of our four patient subgroups. This ÄBP is one way of defining salt sensitivity. What we observed was that salt sensitivity was found in only one group of individuals—black hypertensives. Despite the widely held view that sodium is deleterious for blood pressure, we and others have observed that sodium sensitivity is quite variable (17).
How could this peculiar salt sensitivity come about? Is this salt sensitivity a clue about genetics? Although we are not currently involved in genetic studies in this area, many other groups are exploring this topic, but so far there is little consensus (18, 19).
There are some unique historical sources that may provide a clue to the apparent higher rates of salt sensitivity in black hypertensives. To locate these sources, one needs to consult different libraries, journals, and texts from those found in our medical libraries. The history of slaving may provide some insight. Twelve million black people were kidnapped and transported to the West in the course of 300 years. There were three separate phases of the transport of slaves. The initial passage was from individuals’ villages to the slave port of embarkation. The second or "middle passage" was the trans-Atlantic voyage, and the final passage was from the Western arrival port to the plantation. The overall mortality rate on the middle passage averaged 30% (20) and reflected infection and suicide, but primarily it reflected dehydration from heat and diarrheal illnesses. The dead and dying were thrown cavalierly overboard while still in chains. The English painter Turner (1775–1851) captured the horror of the slave ships in his graphic painting "Slavers Throwing Overboard the Dead and Dying" (Figure 4). The slaves died most commonly from dehydration due to dysentery. Clarence Grim has speculated (21) that perhaps the 30% mortality experience constituted a selective pressure in favor of individuals who were better able to retain sodium, and who thus would be less likely to sustain lethal electrolyte abnormalities. This is sheer speculation because, to date, the frequency of alleles—at least for the renin-angiotensin system—is similar in blacks from West Africa and the United States (22), but the hypothesis remains an intriguing one. Herman Melville (1819–1891) talks about this extraordinary death rate in Moby Dick. "Sharks are the invariable outriders of all slave ships crossing the Atlantic. Systematically trotting alongside to be handy in case a parcel is to be carried anywhere, or a dead slave to be decently buried" (23). In this sense, the sharks and slavers had a commensal relationship throughout the slaving years.
|
STRESS AND PHYSIOLOGY
I have discussed thus far some of our data on nutrition and blood pressure and hinted at some of the genetic implications of slavery. I would now like to consider some of the psychological issues that may be relevant to understanding the ethnic differences in blood pressure. Much of this discussion centers on sympathetic nervous system responses to stressors, but I will also be discussing epistemological considerations that surface whenever one examines reactivity across different cultures.
There is a sizable literature that compares stressor reactivity in blacks and whites (cf. Refs. 25 and 26). As with every other research area, there is a spread of findings that reflects different experimental design in terms of subjects, tasks, measures, and analyses, and of course uncertainty itself. However, I believe that most readers of this literature would agree that black individuals hyper-respond to stressors, and that, when the underlying hemodynamics are studied through techniques such as impedance cardiography, this hyper-reactivity reflects primarily increased total peripheral resistance as opposed to cardiac output responses to stressors.
The problem with this kind of study (and with my breezy summary of the area) is that it skirts the issue of the embedded meaning or nuance of the stressful task itself. Some investigators discuss reactivity findings as if the BP response solely reflects underlying individual differences in the subject’s physiology (eg, "blacks hyper-respond"). However, that sort of conclusion rests on a shaky foundation. How are tasks construed by different individuals? Do they really "mean" the same thing to individuals from a different culture? Perhaps if blacks hyper-respond to the tasks, this hyper-response merely reflects that the task is more evocative to them. If I ask a black person and a white person to talk about an imagined scenario of being falsely accused of shoplifting, the task may not be the same; it may be viewed in a different manner. This is again a problem of culture, interpretation, and communication.
One way of depicting how cultures "frame" a task differently is to examine how different individuals translate the same text. The analogy may seem peculiar, but I would argue that this "translation problem" is akin to the problem of experimental subjects interpreting and responding to task instructions in a psychophysiology laboratory. The first memorable lines of the Iliad come across very differently, depending on the translation (see Table 1; Refs. 27 and 28). Homer’s Greek certainly did not change; however the text, as translated by Lattimore vs. Fagles, is decidedly different. Over the span of decades and centuries, we "see" different nuances in the original stimulus of the Greek. In a way, this different vision is a core problem in any psychophysiology task, particularly when one contrasts subjects of different ethnic groups. Is the task truly "perceived" the same by the subjects or is it "translated" differently?
|
Years ago, I was examining the BP response of a college student in response to a standardized math challenge task. After the study was completed, I accidentally left the BP monitor on while he discussed a troubling event in his life. As the student talked about his hopes to get into medical school and his frustrations with the process, the BP monitor began to alarm. My response was to question the validity of the monitor; the student was talking quietly but earnestly about something very important to him, but in no way could I imagine that his BP was elevated so profoundly as to trigger the monitor’s default alarms. The machine was accurate. We subsequently studied many blacks and whites and learned that the amplitude of BP response to this faintly structured task was extraordinarily high, and far greater than the reactivity to highly structured tasks like a math challenge, but equivalent across both ethnic groups (29, 30). Thus, we did not see evidence that there is an ethnic difference in the amplitude of BP response to individually tailored stressors.
Another way of examining black/white differences in stress reactivity involves pharmacological manipulation. The advantage here is that such approaches presumably bypass psychological filters and directly assess physiological systems’ responses to known doses or stimuli. Over the years, we have infused various agonists and observed the end organ response. This approach allows one to precisely perturb the system with a known amount of a substance while measuring the sensitivity of the body to that substance. Thus, rather than elicit an SNS response with a math challenge, we administer an SNS agonist and measure the response. The BP response to exponentially increasing amounts of infused norepinephrine is reasonably linear. We calculated the slope of this dose response curve in patients studied once on a high salt diet and once on a low salt diet and then calculated the Äslope (high salt minus low salt). If that Ä value is positive, it implies that the individual’s end organ response to norepinephrine exhibits salt sensitivity (ie, the BP increases much more sharply on the high salt than the low salt admission). Figure 5 provides a thought-provoking perspective on salt sensitivity. I have tried to portray this information like peeling an onion. The simplest "outer" peel (left panel) suggests that there is salt sensitivity for everyone. However, when one examines ethnic groups separately, there are surprising differences in salt sensitivity. Similarly, when one contrasts normotensives with hypertensives, there is a striking difference in salt sensitivity. Probably the most striking panel of all is that portrayed on the right panel of the figure. Patients who were black and hypertensive manifest exquisite salt sensitivity; whereas those patients who were white and hypertensive had responses that are apparently indifferent to salt loading (31).
|
BP is one way of measuring vasodilation. Blacks are comparatively resistant to the vasodilating properties of isoproterenol (Figure 6) (33). This observation is consistent with observations from the reactivity literature suggesting that blacks respond to stressors with more of a vasoconstrictive response. They have a blunted vasodilating response to ß-adrenergic stimulation (Figure 6).
|
EMOTIONAL EXPRESSION
In the 1930s, the most prominent voice in psychosomatic medicine was that of Franz Alexander. Although his specificity hypothesis has been rebutted, there are aspects of it that survive in contemporary research. The link between withheld anger and increased BP has been repeatedly discussed. One of the major problems is definitional—ie, what is "withheld anger"? Using the approach of Harburg (36), we examined how anger expression was related to adrenergic receptor physiology. As Figure 7 shows, individuals with increased amounts of withheld anger manifest increased ß-adrenergic receptor sensitivity (37).
|
"LETTING GO"
All of us face daily challenges and insults. Psychosomatic medicine has generally focused on the responses to these stimuli as opposed to the recovery from these challenges. There are two gross measures of such recovery that we have used in our studies of ethnicity and BP.
The original way of treating hypertension, before medication, was "environment-ectomy." When individuals are taken out of their stressful environment and placed in a hospital setting, their BP invariably falls. We found that the BP of hospitalized blacks dropped much less than the BP of whites (39). It is as if there is a carryover or spillover of accumulated frustrations, which even limit the efficacy of removal from the environment for blacks.
An alternative way of examining recovery (or "letting go") would be to measure BP responses during sleep. Although sleep studies featured prominently in psychosomatic research decades ago, our field has relatively ignored this topic in favor of daytime studies. For a number of reasons, attention to this "other third" of the day may be particularly interesting for studies of ethnicity and BP. One such reason involves scrutiny of blood pressure dipping. When one sleeps, BP normally declines or "dips" by about 10%. There are certain pathological conditions characterized by a more restricted or blunted amount of nocturnal dipping. It is curious that this relative nondipping of BP has been reported for blacks. There has been a peppering of such reports over the past decade, and we reasoned that the time was ripe for a meta-analysis. Indeed, when one examines the literature, there is significantly less nocturnal BP dipping in blacks (40). The "meaning" of this observation is still unclear. Perhaps it reflects the possibility that blacks have a greater prevalence of undiagnosed obstructive sleep apnea. The link with sleep apnea is an interesting possibility. First, there are data suggesting that blacks have a higher prevalence of obstructive sleep apnea (41). Second, the repeated episodes of arousal from sleep and desaturation during apneic episodes further augment sympathetic nervous system activity (42). Perhaps the explanation is simpler and reflects disturbed sleep due to housing conditions, life stress, etc. It is unclear which set of variables is best at explaining the phenomenon. The topic cries out for additional study.
CONCLUSION
I have tried to summarize many of the approaches my colleagues and I have used in our studies of ethnicity and health. Low social class, invidious life treatment, and differences in physiological vulnerability all determine risk trajectories across ethnic groups. Every one of us carries legacies (bequests and burdens), which are encoded in our genes, immune system, our group’s experience, and our own individual life experiences. These "memories" must be considered in understanding different susceptibilities to illness. As I have argued elsewhere (43), there is a fog of uncertainty surrounding scientific studies. We catch glimpses of truth through fogbanks. Ethnicity, if ignored, tends to obscure any vision. Ethnicity, if glorified, lists toward racism. On the other hand, by seeking to understand the implications of ethnicity, one has a chance of piercing through the fog and discerning reality. The most productive insights will come from studies that are sensitive to the broad nuances of ethnicity’s determinants.
ACKNOWLEDGMENTS
The work reported in this paper has been supported, in part, by Grants HL36005 and RR 00827 from the National Institutes of Health. The other source of support for this work have been less concrete, but no less tangible—my colleagues and mentors who have shaped this work, my students who asked provocative questions, and the patients who cheerfully volunteered for our studies. The author thanks Thomas Wise, MD, Action Editor for this manuscript, and the anonymous reviewers for their comments.
Received for publication July 2, 1999.
Revision received August 30, 1999.
REFERENCES
2c10-adrenergic receptor genotype. Hypertension 1996; 27: 1210–5.This article has been cited by other articles:
 |
|
 |
|
  C. J. Hilmert, C. D. Schetter, T. P. Dominguez, C. Abdou, C. J. Hobel, L. Glynn, and C. Sandman Stress and Blood Pressure During Pregnancy: Racial Differences and Associations With Birthweight Psychosom Med, January 1, 2008; 70(1): 57 - 64. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Coussons-Read, M. L. Okun, M. P. Schmitt, and S. Giese Prenatal Stress Alters Cytokine Levels in a Manner That May Endanger Human Pregnancy Psychosom Med, July 1, 2005; 67(4): 625 - 631. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. L. Coker, S. Bond, M. M. Madeleine, K. Luchok, and L. Pirisi Psychosocial Stress and Cervical Neoplasia Risk Psychosom Med, July 1, 2003; 65(4): 644 - 651. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. S. Kaufman, J. E. Dimsdale, L. Gleiberman, and J. E. Dimsdale No More ""Slavery Hypothesis"" Yarns Response Psychosom Med, March 1, 2001; 63(2): 324 - 325. [Full Text] [PDF]
|
|
|
|
|
|
L. Gleiberman and J. E. Dimsdale Salt, Hypertension, Evolution Response Psychosom Med, March 1, 2001; 63(2): 325 - 327. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
