A manual of standardized terminology, techniques and criteria for scoring of states of sleep and wakefulness in newborn infants

614

BOOK

Electroencephalography and Clinical Neurophysiology Elsevier Publishing Company, A m s t e r d a m - Printed in The Netherlands

REVIEWS

E d i t e d b y H . PETSCHE a n d JOHN R . HUGHES A manual of standardized terminology, techniques and criteria for scoring of states of sleep and wakefulness in newborn infants. - - T. Anders, R. Erode and A. Parmelee (Editors). (UCLA Brain Information Service, NINDS Neurological Information Network, Los Angeles, Calif., 1971, 41 p.). This manual is the infant companion to the Kales-Rechtschaffcn publication for the scoring of adult human sleep (see review, Electroenceph. olin. Neurophysiol., 1969, 26." 644). Because of the instability and ambiguity of infant behavior, the task of scoring sleep in the newborn is even more difficult than in the adult and the authors of this volume recognize and squarely face many of these difficulties. The result is a first step toward quantitative analysis of the dynamic behavior of early human life. Using this manual, a second step, based on the first, can be taken as I will attempt to show in this review. The introduction begins with Prechtl's definition of state: "Constellations of certain patterns of physiological variables and/or patterns of behavior which seem to repeat themselves and which appear to be relatively stable". "Relatively stable" is just what infant behavior is not, however, so that a method of coding, rather than scoring, states is proposed by Anders et al. Coding allows the recognition, evaluation, and weighing of individual components of a constantly changing (i.e., relatively unstable) constellation of variables. A concise outline of environmental factors that may influence infant behavior precedes the sections on recording techniques and methods for coding of observations. Behavioral observation is emphasized by the authors and it is to become an important part of the coding system. Here those who have done long-term recording will recognize the problems of flagging observer attention, of subjectivity, and of making all or none judgments on what are clearly graded phenomena. The authors might agree that if behavioral observation is really essential to a scoring system, it should be objectified, as by video-taping, so that some of these pitfalls could be avoided and so that reliability could be documented. Hand-written comments on polygraphic records are clearly inadequate even at this early stage of development (sic). Polygraphic monitoring yields records that are objective but still must be judged subjectively to be coded. The limitations mentioned above can be quickly appreciated by attempting to use the authors' criteria to code some of the illustrative records in the manual (in discussing the figures, A refers to the left and B to the right epoch). This is particularly true of the E M G records which must be scored plus (Fig. 5A and B, Fig. 6A) or minus (Fig. 8A), and the respiratory records which must be judged to be regular or irregular (Fig. 5A and B absent, Fig. llB, Fig. 13A) despite the fact that they are both. As the authors point out, both muscle tone

and respiratory regularity are continuous variables and thus could be quantified more precisely allowing sharp, unambiguous dichotomization. With refreshing candor, the editors acknowledge the failure of their committee to agree on criteria for defining and scoring states. As with most oscillating systems, the peaks and troughs of the infant sleep cycle are easier to identify than the transitions. This is because the first differential of the several variables is changing slowly in the first instance: the "state" is relatively stable. Infant sleep oscillates between active REM sleep (A) and quiet sleep (Q) which are clearly defined and illustrated by the authors. Unfortunately, an unspecified amount of time is spent betwixt and between, in "indeterminate sleep", a frightful term for an undefined state, if indeed it is a state at all. "Intermediate"' or "transitional" seem to be preferable and the authors do not explain why they rejected these alternatives. The greatest weakness of the manual is the inadequacy of the definition of this category and the resulting uncertainty of scoring. By the authors" own evaluation, 7 of 17 illustrated sleep epochs are ambiguous. In assessing the magnitude of the problem, it would be helpful to know what proportion of total recording time is spent in "indeterminate sleep". Perhaps a general principle emerges here: that standardization depends upon arbitrary decisions that cannot easily be made by committees. To test further the usefulness of the manual and to attempt to obviate the inherent difficulties of the nonquantitative and ambiguous coding system, I followed the authors' suggestion and developed a scoring method which is simple and reliable. There are six variables, five of which are judged to be discontinuous (i.e., digital): Presence (1) or absence (0) of: (a) facial movement; (b) body movement: (c) EMG suppression; (d) eye movement: and (e) irregular respiration. The sixth variable, the EEG, is continuous, but can be scored low voltage irregular (LVI) - 4; mixed (M) 3: tracO alternant (TA) = 2: and, high voltage slow (HVS) I. If one accepts the authors' coding judgments, application of these weights to the 17 epochs of sleep illustrated in Fig. 6 through 14A yields state scores which range continuously from 1 to 9. A score of I is obtained only from records called unambiguous quiet sleep (Fig. 10A and B). Scores of 8 and 9 are obtained only from unambiguous A-REM sleep (Fig. 6A, 7A and B, 13B). Scores from 2 through 7 come from the other records, all but two of which were called (among other things) indeterminate by the authors. Whether this approach is valid or not, it is unambiguous, quantitative, and reliable. Since the goal of the work is standardization, this complementary scheme is offered by an arbitrary outsider with no conceptual or terminological ax to grind. I assume only that states, by definition, are not "indeterminate". Neither, by their nature,

BOOK REVIEWS

615

are numbers. Bibliophiles will recoil from this manual as the page size, paper, and covers are odd, offensive, or both. The absence of time calibration in the records may also cause some regret, though it is not really necessary for the pattern recognition involved in this eyeball kind of analysis. In summary, this work extends the objectification of sleep behavior to the newborn period and clearly exposes many problems that remain to be solved by future work. As a state-of-the-art guide to the perplexed, this manual will be welcomed by electroencephalographers, pediatric neurologists, and developmental psychologists, as well as by students of sleep. J. ALLAN HOBSON

Harvard Medical School, Boston, Mass. 02115 (U.S.A.) Electroenceph. clin. Neurophysiol., 1972, 33." 614-615

Explorers of the brain. - - L. A. Stevens. (A. A. Knopf, New York, 1971, 348 p., $ 7.95). Those of us who have made an understanding of brain function a life-time pursuit should be grateful to Leonard Stevens for providing a remarkably thorough and detailed account of the tortuous process that led to our current state of knowledge. Faithfully recorded under eye-catching chapter headings such as " F r o m Spirits to Electricity", "Life without Mind", "The Last of the Galvanometers", "Secrets in a Sea of Detail", "Electrical Maps of the Nervous System", and "The Search for the Engram" are the observations, speculations, and conclusions that shaped the thinking of generation upon generation of scientists and clinicians. The false steps as well as the gainful ones are set forth. The historical controversies are brought into sharp focus with proponent and opponent being clearly identified. And in so doing, the author brings alive a subject that might otherwise be dull and obscure. Mr. Stevens" background makes him well suited to his task. He has had extensive experience in making scientific and technological matters palatable to a general audience. Though not a scientist himself, he is clearly at home among them, preparation for the present book having included contacting and interviewing many of the most active contemporary workers. The author's narrative touches upon some of the most recent developments in brain physiology. Here the reportorial style is especially effective, making the reader a party to scientific history in the making. A minor overemphasis upon the startling item in this section can be forgiven readily because of the general soundness of the approach. It is hoped that this book will be displayed prominently on every "'current non-fiction" shelf but it is likely to b~ sought out even when placed in the back stacks. Many will find it a valued addition to their own libraries. K. A. Kool

Departments of" Psychiatry and Neurology, University of Michigan Medical School, Ann Arbor, Mich. 48104 (U.S.A.) Electroenceph. clin. Neurophysiol., 1972, 33:615

Languages of the brain. - - K. H. Pribram (Prentice-Hall, Englewood Cliffs, N.J., 1971, 432 p., $9.50). As suggested by its title, this is a poetic book. This fact is responsible both for its major strengths and for its major weaknesses. Complex ideas are creatively juxtaposed and compressed, deftly elaborated by allusion and well chosen examples. Poetry, however, may legitimately be developed unfettered by fact, where science is more constrained. In his striving to bridge the great gulf between neurophysiology and behavior, the author has in places soared freely into speculation. He and his fellow neuroscientists can readily recognize these points of poetic departure. The less advantaged readers, to whom primarily this book is addressed, may not so easily perceive when they have left the ground. The 20 chapters of the book are divided into 4 sections : (1) basic neurophysiology; (2) psychological processes, perception and emotion; (3) motor mechanisms and reinforcement; (4) communicative processes, signs, symbols, speech and thought. It is proposed that graded responses, i.e., postsynaptic potentials (PSPs) and DC potentials, are the critical phenomena for the integrative processes underlying most neural action of behavioral or psychological importance. Such considerations lead the author into the romantic extrapolation that mnemonic mechanisms are analogous to those of holograms, the EEG somehow representing waves propagating through an ethereal neuropil. In historical perspective one can here perceive the ghost of the "reticular" theory of neural action, returned in a new form. The contrary view, "the neuron doctrine", around which 20th century neuroscience has grown, scarcely requires defense. The specific points of controversy, however, need more precise identification than they have received in the book. No one denies the integrative role of PSPs in controlling the action of the neuron in which they are generated. This has been the major thrust of neurophysiology for the past 20 years. It is an entirely different matter, however, to propose that the PSPs in one neuron have significant influence upon its neighbors. Rather, it is probable that the chemical step in synaptic transmission has evolved precisely to curtail such non-specific "crosstalk'; and in those cases where electrotonic coupling is prominent there is both morphological specialization and curtailment of integrative scale (Bennett;,Pappas: Baker and Llin~s) or pathological consequences (Rusinov and Ezrokhi). The organization of the neocortex into modality-specific columns of neurons (Mountcastle; Hubel and Wiesel; Asanuma) is not likely to be achieved by non-specific gradients of electrical fields, nor to offer any advantage were such fields to play a prominent integrative role. indeed, from what is known, it might be supposed that cortical analysis proceeds despite interfering field potentials rather than by means of them, and this has been the conclusion reached by the few experiments designed specifically to test this point (e.9., Lashley, Chow and Semmes; Sperry, Miner and Myers). Thus, the proposal that diffuse effects of PSPs and DC gradients might account for the still mysterious and exquisite subtlety of cortical processes needs something more than mere analogy with holograms to support it. Even the analogy is far fetched since the remarkable properties of holograms are wholly dependent upon two features which