Experiments at The State University of New York Health Science Center

continued

CURRENT GRANT PROPOSAL
Rosenblum's current NIMH grant, entitled Experimental Studies of Susceptibility to Panic,²⁰ extends from early 1988 to August 1998. Funding for the 1993-1994 grant period was $205,99720 and for the 1994-1995 grant period is $261,630.41 Although his laboratory is located at SUNY, Rosenblum has been collaborating with clinical psychobiological and psychopharmacological investigators from The New York State Psychiatric Institute at Columbia University, also in New York City.
Their collaborative research purports to address a human condition called panic disorder (PD) by inducing a condition in bonnet macaque monkeys they call "acute endogenous distress" (AED). Rosenblum describes AED as a pattern of emotional responses in monkeys originally discovered by injecting them with sodium lactate.²⁰ The experiments funded by the grant are designed for "neurochemical and behavioral investigation of the provocation of anxiety symptoms in non-human primates which reliably mirror significant aspects of human PD."²⁰ In naming the monkey and human conditions differently (AED and PD respectively) Rosenblum has avoided the error^19,42 he and other monkey researchers had repeatedly committed of inappropriately assigning the name of the human disorder (depression) to the experimentally induced monkey disorder. But, by not employing the identical name, there is tacit admission that the two conditions are not sufficiently similar to warrant the same name, casting doubt on the validity of the model. Although mere similarities between the human disease and the animal condition may suffice at times to create a strong "model," in this case the similarities are too disparate.
In order to determine whether or not (monkey) AED truly models (human) PD, it is helpful to utilize the guidelines established by monkey researcher William McKinney and former NIMH Chief William Bunney, who have suggested that the animal model and human condition have similar, if not identical, (1) causes, (2) symptoms, (3) biological mechanisms, and (4) cures.⁷ Relating the characteristics of PD that Rosenblum has purportedly studied in monkeys to these criteria can be accomplished by using the following outline:

A. Causes

1) employment of agents capable of provoking both PD and AED
2) insecure early attachment as psychological predisposing factors to PD
3) inadequate social and cognitive mastery skills as concomitant factors in PD
4) recent stressful events as sensitizing precipitants to PD

B. Symptoms

1) greater incidence of PD in females
2) diagnostic criteria of PD
3) cognitive dysfunction resulting from PD

C. Biological Mechanisms

1) dysregulation of the locus ceruleus and related neurochemicals in PD

D. Cures

1) treatment of PD

ANALYSIS OF PROTOCOL ACCORDING TO ESTABLISHED GUIDELINES

A1 Employment of Agents Capable of Provoking Both PD and AED
PD occurs spontaneously in humans, without chemical induction. The three variations of human PD are: (l) unexpected [no situational trigger]; (2) situationally bound [occurs immediately on exposure to or anticipation of a trigger]; or (3) situationally predisposed [may or may not occur following a trigger, and then not immediately].²⁹ AED, on the other hand, is non-spontaneous, non-situational, and chemically induced by sodium lactate.⁴³ Since sodium lactate stimulates normal monkeys, but only stimulates humans with PD, it appears to have different effects in PD and AED. Also, Yohimbine, an alpha-2 adrenergic antagonist drug clinically used to treat sexual dysfunction and used by Rosenblum's research group to precipitate AED, seems to work differently in humans and monkeys, tending to produce anxiety more in humans.^44,45 Also, Yohimbine can apparently provoke AED in normal animals,⁴⁴ but only precipitates panic attacks in PD patients. Drugs, therefore, seems to have different effects in AED and PD.
Because of species differences in absorption, metabolism, distribution, and excretion, the different actions of sodium lactate and Yohimbine in humans and monkeys are not surprising. Indeed, many drugs can induce different effects in dissimilar primate species. Aspirin causes birth defects in monkeys⁴⁶ but not in humans. Benoxaphen, a non-steroidal anti-inflammatory drug, showed no toxicity in rhesus monkey tests, but produced liver failure and cutaneous photosensitivity in humans and was withdrawn in 1982 after 3500 cases of severe reaction and 61 deaths were documented in the UK alone.⁴⁷ More recently, toxicity for the anti-hepatitis drug fialuridine (FIAU) was not noted in hepatitis-infected monkeys, but, in 1993, clinical trials in 15 patients revealed severe hepatotoxicity in seven, with five dying and another two surviving only through liver transplants.⁴⁸ Drugs show disparate effects even in different species of monkeys. For example, atropine administration prevents eye elongation from eyelid suturing in Macaca arctoides, but not Macaca mulatta.⁴⁹ Finally, non-human primate tests may screen out useful drugs. One such drug was the injectable contraceptive Depo-Provera, which was banned in the US in the early 1970s because it caused cancer in baboons. In 1993, however, the FDA lifted the ban after 20 years of human experience in those countries not prohibiting its use disclosed that it did not cause cancer in humans.⁵⁰

A2 Insecure Early Attachment as Psychological Predisposing Factors to PD
Rosenblum discusses the clinical finding that insecure social attachments predispose to PD, citing many studies in the clinical literature relating PD to childhood separation anxiety disorder,²⁰ an association first proposed in modern clinical psychiatry by Donald Klein in 1964.⁵¹ Rosenblum has concluded, after 12 years of study, that the effect on human infants of inadequate mothering can be modelled in monkeys by the effect on infants of unpredictable foraging demands placed on the mothers,²⁰ an important factor in the development of AED as a model of PD. Yet, three recent clinical reports, from different research teams, relating childhood psychological factors to adult PD, barely mention animal research, and ignore completely Rosenblum's monkey work.^52-54
The University of Cincinnati team's 1993 paper studying separation anxiety in panic disorder⁵² and the Columbia University team's 1994 paper--written by some of Rosenblum's own co-authors--studying childhood separation anxiety disorder in patients with adult anxiety disorders (62% of whom had PD)⁵³ does not contain a single citation of animal research. The Cornell University team's 1993 paper discussing aspects of inhibited behavior within the framework of a psychodynamic model of panic disorder,⁵⁴ contains a single reference to an animal research finding of "socially submissive" monkeys, which, the authors claim, is "perhaps" the equivalent of the human characteristic of shyness that is documented in three cited clinical papers. From the work of these PD clinical investigators, it appears that maternal deprivation studies in general, and Rosenblum's maternal foraging studies in particular, have not been instrumental in gaining understanding into childhood psychological roots of panic disorder.
In fact, the monkey studies have only dramatized human maternal deprivation findings described by Spitz in 1946⁵⁵ and Bowlby in 1952,⁵⁶ and re-examined twice by Bowlby in 197357 and again in 1988,⁵⁸ without a single reference to monkey maternal deprivation experiments. Despite this, monkey researchers continue to write about the "phenomenon of insecure early atachment" as if only monkey data exist and only monkey data can illuminate this clinical question. Their complete disregard for the significance of the clinical literature, and their repeated calling for still more monkey experiments in areas already understood through human findings, are difficult to understand considering the experimenters' professed motives of better understanding of humans.
For example, Rosenblum depreciates the clinical literature by his 1994 assertion that "the need for stable, secure early attachment experiences as essential in reducing the likelihood of certain forms of adult pathological anxiety" will be even more dramatically accepted "should [the] pattern of findings continue in [his] future work."⁴¹ But the "dramatic" value of monkey experiments has limits, despite monkey researcher Phyllis Dolhinow's 1986 unsupported and likely incorrect assumption that maternal deprivation monkey experiments "have encouraged physicians, psychologists, and others concerned with early development to integrate biological and social behavior and to look more closely at the physiological effects of early experiences."¹⁹ In all likelihood, clinical experience alone, including studies by Bowlby and Spitz, should have been, and probably were, sufficient to draw attention to this issue and to support the need to further explore mother-child relations in humans.
It is hubris, and deceptive, to suggest that 50 years' of clinical study of the psychologically harmful effect of maternal deprivation on human infants and (later) adults must await confirmation from monkey experiments. This attitude likely derives in part from the frequent disregard or even contempt of early monkey researchers for human data. For example Harlow wrote:

The data of the human theorists did not generalize to monkeys because the human theory was false. Monkey theories basically generalized to human infants because the monkey facts were true. . . Sometimes when monkey data fail to generalize to human data the answer lies in the superiority of the monkey data and the need to revise those data that are human.⁸

Suomi added, "We are not aware of human data--quite frankly, they tend to bias our research efforts."⁹
But contradictions abound. At the same time that human data are disparaged, Harlow, Suomi, and colleagues also acknowledge their importance. For example, Suomi has written:

whether actual data obtained from nonhuman primates have added measurably to our understanding of human development is another matter. The record to date has not been particularly impressive. . .[the] few findings derived from nonhuman primates that have clearly advanced knowledge of human development. . .are relatively rare. Most monkey data that readily generalize to humans have not uncovered new facts about human behavior; rather, they have only verified principles that have already been formulated from previous human data. . . A case in point is in the area of the development of mother-infant relationships. To date the monkey data have added little to knowledge of human mother-infant interactions;¹⁰

Other similar opinions have been expressed by Harlow and colleagues:

We have a considerably more difficult establishing a strong case [for] the data obtained from depressed monkeys for clinicians currently working with depressed patients since so much monkey work to date has been based upon existing human data and theories. . .it must be admitted that the primary flow of information, to date, has been from human clinical research into the animal modelling arena."11-13

A misguided tradition that dominates contemporary biomedical research holds that human data must be confirmed by experimental animal data. But animal studies merely reprocess already known human psychological information into a "new" animal behavioral format in an attempt to lend greater scientific credibility to human data. Even experimental psychologists themselves are critical of this approach as being inherently invalid and methodologically unsound.^59,60 Were the monkey data at least explanatory, they might offer something of value. But, in truth, they only invalidly, scientistically, and unnecessarily transform ideas from one conceptual framework (human psychology) to another (animal behavior).

A3 Inadequate Social and Cognitive Mastery Skills as Concomitant Factors in PD
According to Rosenblum, in relating AED to PD, a monkey's failure at a video game is equivalent to "lack of mastery" in humans, and a monkey's experience in performing a difficult video game subsequent to an easy one is equivalent to loss of control in humans. Who can know, however, how a monkey experiences success or failure at a video game? And, prior experience performing the game may be a confounding variable. There is no reason to believe that video game performance in monkeys relates to causal factors of PD in humans.

A4 Recent Stressful Events as Sensitizing Precipitants to PD
Rosenblum believes that monkey equivalents of adverse human life events, such as social anxiety and interpersonal loss, can be created in the laboratory. Social anxiety is produced by introducing an individual "new" bonnet monkey, housed in a plexiglass encasement, into a cage containing a long-term stable bonnet monkey group.²⁰ Interpersonal loss is produced by unpredictable separations of monkeys previously pair-housed for a period of months.²⁰
Such "traumata" hardly mirror complex human stressful interpersonal life events. Also, the research data is confused by other stressors, some related to the infusion vests and indwelling catheters the monkeys are forced to wear and carry.²⁰ In his most recent refunding grant application, Rosenblum alludes to an "ambiguous" result derived from "within- and between-subject variability" relating to "problems with the comfort and ease of movement when subjects wore the pump/jacket combination, particularly given the weight of the volume of fluid to be infused," and stated his intention of ". . .redesigning the commercial jackets currently used and. . .determin[ing] separately whether. . .lighter solution volumes may be used effectively or whether new means of remote or other forms of administration will be possible. . ."⁴¹ Who can know to what extent such confounding variables effected previous data, or to what extent such factors can be or have been corrected so as not to contaminate future data?
The jackets and catheters typify stressful laboratory conditions, repeatedly shown to produce profound metabolic, immunologic, and other physiological and biochemical variations in laboratory animals, sufficient to constitute uncontrolled, confounding variables that distort data and undermine reliability of experimental results. This is especially true in Rosenblum's protocol, which specifically studies stressful life events that may precipitate AED, and suggests that the stressors experimentally introduced (and thought to be controlled)--namely, social and separation anxiety--constitute the only stressors experienced by the monkeys. But, in addition to the jackets and catheters, there are many additional sources of stress.
Studies have conclusively demonstrated that monkeys and other animals are psychologically stressed by typical laboratory-induced factors, such as caging, social disruption, isolation, restraint, methods of transport, repeated handling, noise levels, lighting, abnormalities in food or water, and routine use of cleansers, insecticides, and chemical sterilizers.^61-66 Altered plasma cortisol levels⁶² caused by such stress directly undermine Rosenblum's experiments, since he measures corticotrophin-releasing factor. Unavoidable laboratory-induced stress has also been shown to result in physiological changes,⁶³ such as pulse rates alternations,⁶⁴ self-mutilating behavior,⁶⁵ and other significant endocrinological and immunological changes.⁶⁶
Furthermore, captive-bred monkeys, usually assumed to be at least as scientifically valid as laboratory animals as those obtained in the wild, show more laboratory stress than conspecific animals from the wild. Duncan Harris of the University of London found that macaques bred in captivity appeared less well adapted to life in a cage than those who had grown up in the wild.⁶⁷ He concluded, "As well as affecting the welfare of animals used in research, the way in which primates are reared could cause biochemical changes which affect the research itself."⁶⁷

B1 Greater Incidence of PD in Females
Whether or not the incidence of AED parallels the 4:1 female predisposition found in PD is not known, for female monkeys alone serve as the initial experimental subjects.²⁰ Rosenblum plans to study male monkeys later.

B2 Diagnostic Criteria of PD
The requirement of predominently subjective symptoms rather than objective signs for the diagnosis of PD indicates the difficulty of using behavior to diagnose a subjective affective state. Behavior can "mean" different things in different people, and, even in the same person, purpose, motivation, or meaning cannot necessarily be inferred from behavior. This is even more true in animals, which are unable to communicate what is on their minds, and Rosenblum himself admits "the importance of contextual factors when studying animal behavior."²⁰ Yet, unlike the diagnosis of PD, the diagnosis of AED--also a subjective state--is based on inference from observing monkey behavior. Rosenblum knows that the diagnostic symptoms of PD (e.g., palpitations, sensation of respiratory distress, feeling of choking, chest pain, nausea or abdominal distress, dizzy feeling, feeling of dissociation, feeling of loss of control, fear of dying, numbness)²⁹ simply cannot be assessed in monkeys because these symptoms must be subjectively experienced and reported by the patient rather than observed by the clinician. The diagnosis, then, cannot, by definition, be given to non-human primates. Rosenblum avoids this problem by naming the disorder differently in monkeys--AED--but states that it is sufficiently like PD to serve as a model.
Since diagnostic requirements make PD a uniquely human condition, there is (and can be) no animal model. There is no reason to believe that the subjective experience of monkeys with AED is similar to that of humans with PD, and, as we have seen in the previous section, causes differ. Despite Rosenblum's assertion that AED has "important similarities"²⁰ to PD, it has never been validated as a useful or reliable model for PD.

B3 Cognitive Dysfunction Resulting from PD
PD adversely effects human cognitive function, and Rosenblum believes that the equivalent adverse effects in AED can be appropriately studied by examining two tasks thought to indicate cognitive function; i.e., short-term memory deficits, and "delay of reward capacity," (the degree to which monkeys can tolerate the frustration experienced from delays in their rewards for accomplishing tasks.) But the kinds of complex and subtle cognitive deficiencies demonstrated in humans on neuro-psychological subtests indicate the need for much more sophisticated methods for testing cognitive dysfunction in monkeys than the tasks used, which reveal only gross deficits. Such refined methods are, of course, impossible to utilize in monkeys. Also, sort-term memory is just one of many measures assessed on mental status exam to evaluate cognitive function. Finally, "delay of reward capacity," deriving from poor frustration tolerance, seems more an affective deficit than a cognitive one, similar to the difficulty with delayed gratification observed in human patients with impulse disorders. The monkey cognitive deficits, as determined and measured by Rosenblum, do not parallel human cognitive deficits found in PD.

C1 Dysregulation of the Locus Ceruleus and Related Neurochemicals in PD
Rosenblum uses Yohimbine challenge to assess locus ceruleus function in monkeys²⁰ since dysregulation of the locus ceruleus and related neurochemicals is thought to be involved in both AED and PD. However, as previously stated, Yohimbine may work differently in humans and monkeys. In addition to causing more anxiety in humans,^44,45 it also appears to effect primarily cognitive rather than motor function in monkeys^,68 whereas in humans it may induce hypertension, tachycardia, motor excitation, diaphoresis, vomiting, headache, flushing, penile erection, mild anti-diuresis, and mydriasis. For these reasons it is highly speculative to assert that Yohimbine in monkeys will elucidate neuroanatomical and neurochemical aspects of PD.

D1 Treatment of PD
The treatment for PD includes several modalities, including pharmacotherapy, psychotherapy, group counseling for anxiety management, and behavioral desensitization. Only pharmacotherapy, however, is available to "treat" monkeys with AED. In non-human animals it is impossible to test other methods that have had some success in the treatment of PD,⁵⁴ demonstrating yet another weakness of the animal model approach.
Pharmacological agents recommended for the treatment of AED include imipramine, clomipramine, clonidine, alprazolam, monoamine oxidase inhibitors, and fluoxetine,²⁰ all showing some efficacy in the treatment of PD, and all deriving from work in humans. Rosenblum's colleagues, in an article discussing the role of serotonin in panic-anxiety states, suggest that animal studies have played a role in the use of selective serotonin reuptake inhibitors in the treatment of PD.⁶⁹ However, the utility of such drugs (fluoxetine, paroxetine, sertraline) for such varied conditions as depression, obsessive-compulsive disorder, post-traumatic stress disorder, bulemia nervosa, menopausal syndrome, premenstrual syndrome, attention deficit disorder, and various anger dyscontrol conditions has been discovered through clinical use, and its success or failure in PD will likewise depend on the results of human trials.

CONCLUSION
The question of the relevance and importance of maternal deprivation monkey experiments has increasingly become a subject of debate, because of Harlow's questionable motives⁷⁰ and the growing recognition of conceptual and methodological flaws. Even prior to recent scientific critiques by Giannelli,¹⁶ Stephens,¹⁷ and Reines,^18,19 the early practitioners of and advocates for maternal deprivation monkey experiments expressed their own doubts at the time of publication.^11-14 More recently, McKinney wrote in 1986 that, "The 'modelling' of a specific clinical syndrome in animals will [in the future] be de-emphasized,"⁷¹ and Gluck¹⁵ has questioned the ethics of animalexperimentation in general.^72,73 Lack of clinical relevance, however, has had no noticeable effect on the research activity of Rosenblum, Sackett, Suomi, Hofer, Kraemer, Mason, and Levine, who have continued their monkey experiments. However, for reasons of economic feasibility, scientific validity, and medical accuracy, maternal deprivation research in general and Rosenblum's studies in particular should be defunded and phased out.

SUMMARY
Psychologist and monkey researcher Leonard Rosenblum assumed directorship of the Primate Behavior Laboratory at the State University of New York Health Science Center in Brooklyn, New York in 1963, and, for about three decades, has been conducting maternal deprivation experiments with bonnet and pigtail macaque monkeys. Previously, he trained with Harry Harlow, who performed the first maternal deprivation experiments at the University of Wisconsin.
From early 1988 until 1998, Rosenblum is being funded by NIMH grant MH-42545, entitled Experimental Studies of Susceptibility to Panic. Rosenblum received $205,997 in 1993-1994 grant support, and $261,630 in 1994-1995.
Rosenblum and colleagues have been attempting to investigate a spontaneous human condition called panic disorder (PD) through an induced condition in bonnet macaque monkeys they call acute endogenous distress (AED). By naming the monkey disorder AED, Rosenblum has avoided the obvious fallacy of calling a laboratory induced animal condition by the same name as a human disease. Indeed, it is impossible to diagnose PD in monkeys by established clinical psychiatric criteria. Although Rosenblum believes that AED sufficiently reflects characteristics of PD to qualify as an animal model, criteria for a valid animal model established by monkey researchers themselves argue against Rosenblum's assessment. The animal model and human condition must have similar, if not identical causes, symptoms, biological mechanisms, and cures. Yet by these criteria AED does not model PD because these two conditions show essential differences, the latter derived from a state of mind and the former induced by a chemical.
The causes, precipitants, and sequelae of PD include unstable and insecure early attachment, inadequate social and cognitive mastery skills, recent stressful events, and cognitive dysfunction. None of these are convincingly demonstrated in monkeys. Furthermore, the intentional stress of the experimental protocol confounded by many other unintentional sources of stress, which complicate the research data and make them difficult, if not impossible, to interpret.
Maternal deprivations experiments have merely dramatized findings previously known in humans, such as those described by Spitz in 1946 and Bowlby in 1952, before any of Harlow's results were ever published. Major monkey researchers like Harlow, Suomi, Kraemer, and McKinney have all acknowledged the superiority of human maternal deprivation data. Yet they cling to the highly questionable scientific tradition that human data must be confirmed by experimental animal data in order to be "proven" valid. Such animal research, in essence, merely reprocesses already known human psychological information into a "new" animal behavioral format. Were the monkey data at least explanatory, they might offer something of value. But, in truth, they only invalidly, scientistically, and unnecessarily transform ideas from one conceptual system (human psychology) to another (animal behavior).
The scientific validity of maternal deprivation monkey experiments has been increasingly scrutinized and challenged. At least two of Harlow's original colleagues have recently re-examined their value, one recommending that they be de-emphasized, and another questioning their ethical basis. Although the prevalence of maternal deprivation experimentation does not seem to have decreased over the past decade, the scientific deficiencies in the methodology call into question the value of Rosenblum's studies in particular and maternal deprivation research in general. Should they continue to be assessed as methodologically unsound, they should be defunded and phased out.

Copies of this booklet can be obtained by contacting MRMC by e-mail, or by contacting MRMC at:

MRMC
P.O. Box 201791
Cleveland, Ohio 44120
(216) 283-6702 (phone and fax)

references