James M. Nortona, Ph.D., Gretchen Sibleyb, B.A., and
Richard Broder-Oldach, B.S.

Departments of Physiologya and Osteopathic Principles and Practiceb
University of New England College of Osteopathic Medicine
11 Hill's Beach Road
Biddeford, ME 04005

Materials and Methods


Physiological rhythms within the neural, muscular, respiratory, cardiovascular, endocrine and other systems play a major role in the maintenance of stable and appropriate conditions within the internal environment of the body. Characterization of these rhythms, a major thrust of biomedical research, focusses on rhythmogenesis (the cellular mechanisms responsible for initiating and/or maintaining the rhythm), frequency (cycles per sec, per min, per year, etc.), amplitude (magnitude of force, range of concentration, amplitude of membrane potential, etc.), control of the rhythm, and the interaction of a given rhythmic process with other homeostatic or compensatory mechanisms.

The cranial rhythmic impulse (CRI) has been identified and studied by a large number of osteopathic physicians and scientists ever since the development of cranial osteopathy by Sutherland(1),(2). The CRI is generally considered to be an independent rhythm occurring at a frequency similar to that of respiration (6-14 cycles/min(3),(4)) and has been described as "resembling the respiratory excursion of the chest in minute form ..."(5). The amplitude of the CRI is variously described as "strength" or "vitality" in a rather subjective fashion; attempts to quantify CRI amplitude(6),(7),(8) have been made with varying degrees of success.

This author recently published a tissue pressure model for the CRI(9) that incorporated the cardiovascular and respiratory rhythms of both subject and examiner as determinants of the perceived CRI frequency and amplitude. According to this model, the CRI would arise in the soft tissues of the two participants (rhythmogenesis) and would have a frequency and amplitude dependent upon the complex interaction of tissue fluid pressure fluctuations and the characteristics of the mechanoreceptors in the hands of the examiner. The model therefore incorporates all the necessary characteristics of a true physiological rhythm as described above and applies these characteristics to the CRI in a testable hypothesis. The following article describes results of experiments designed to generate the kind of hard data concerning CRI frequency and amplitude required to test the validity of the tissue pressure model.


subjects: The twenty-four (24) subjects who volunteered for this study were apparently healthy students and faculty members at the College of Osteopathic Medicine of the University of New England. All subjects were informed in advance of the general nature of the project and of the nature of the physiological measurements that would be made, and gave their informed consent prior to participation in the study.

examiners: The twelve (12) examiners were osteopathic medical students, teaching fellows with the UNE/COM Osteopathic Principles and Practice Department, UNE/COM clinical faculty, and members of an informal cranial osteopathy study group which meets regularly at the university. The level of experience for the examiners in CRI palpation ranged from two to fifteen years.

measurements and procedures: Subjects were asked to lie supine on an examining table and the examiner was seated at the subject's head. Examiners were asked to monitor the subject's CRI using a standard vault hold. In order to assure that both subject and examiner were in a steady state, the period of CRI monitoring was preceded by several minutes of quiet rest, initially with no contact between examiner and subject, then with the examiner's hands on the subject's shoulders. The period of CRI monitoring was usually about two (2) minutes in duration.

A switch attached to the leg of the examining table and activated by examiner knee pressure was used to record the CRI. The examiners were asked to depress the switch during the flexion phase of the CRI cycle, thereby producing in the time/event marker channel of a chart recorder an upward deflection that corresponded to the flexion phase of the CRI as perceived by the examiner. The chart recording speed used throughout was 2 mm/sec.

Following a period of monitoring and recording the CRI, the examiners were asked to quantify the amplitude of the subject's CRI using a five-point scale (1=well below average, 2=somewhat below average, 3=average, 4=somewhat above average, 5=well above average). This figure was recorded on the permanent record of the session.

analysis of the data: The basic data obtained for each measurement session using the protocol described above consisted of CRI cycle length (measured as the time between the beginning of one flexion and the beginning of the next), duration of flexion, and CRI amplitude. Calculated values derived from these basic measurements included average values for cycle length and for duration of the flexion phase (in seconds), duration of the extension phase (in seconds, calculated as the difference between total cycle length and the duration of flexion), and CRI frequency (cycles/min, calculated as 60 ÷ the cycle length in seconds). Calculated frequencies were validated by visually counting the number of flexions (or extensions) over a period of one minute on the permanent record.

Statistical analyses (simple and multiple linear regression, curvilinear regression, analysis of variance, and Student's t-test) were performed using commercially available software and statistical programs written by this investigator following standard formulae and procedures(10).


A total of 274 CRI cycles were recorded on twenty-four (24) subjects by twelve (12) examiners; these data are summarized in Table I. The most remarkable finding from these experiments is the relatively long average cycle length and the correspondingly low average CRI frequency of 3.7 cycles/min, considerably lower than previously published values. The range of average cycle lengths among the twenty-four (24) subjects was 11.7-22.3 seconds. Flexion occupied 46.5% of the average CRI cycle, and little variation in this percentage was seen over the range of cycle lengths observed in these experiments. Amplitudes ranged from 2-4 on the arbitrary 5-point scale, with no values of "1" or "5" reported by any examiner. No significant correlation was found between CRI frequency and amplitude using standard linear regression techniques, although visual inspection of the data suggested an inverse relationship between the two. The measured duration of flexion and the calculated duration of extension were both found to be linearly correlated with CRI cycle length (flexion: r2 = 0.877, p<.001; extension: r2 = 0.907, p<.001).


methodology and experimental conditions: The knee switch designed for these experiments proved to be an effective method of allowing the examiner to record palpatory findings while continuously monitoring the CRI with both hands. The switch was silent, and the slight leg movement required to activate it was imperceptible to the subjects. Using this method to record the CRI did not require the examiners to be constantly aware of the passage of time, and allowed them to focus on their palpatory findings. All examiners questioned reported that they were able to monitor the subject's cranial rhythm in a manner consistent with their usual practice, and that there was nothing unusual about the rhythm(s) they were monitoring. However, almost all of the examiners were surprised to find that the frequencies were so much lower than expected.

relationship between CRI frequency and amplitude: Several well-described physiological rhythms demonstrate an inverse relationship between frequency and amplitude. For example, a given minute ventilation can be achieved by a wide variety of ventilatory frequencies and tidal volumes; a slower frequency requires a larger tidal volume, and vice versa. The relationship between CRI frequency and amplitude for the subjects in this study suggested a similar pattern, but no statistically significant relationship was found. The coarseness of the 5-point scale used by the examiners and the narrow range of reported values for amplitude may have obscured a potentially significant inverse relationship, and a new amplitude scale will be devised for future experiments.

CRI frequency and examiner experience: If the CRI represents a true physiological rhythm that an examiner can be trained to evaluate, then CRI frequency, as a basic property of that rhythm, should be readily palpated even by relatively unskilled examiners. Good agreement should therefore exist among examiners with respect to the determination of CRI frequency, regardless of the examiners' levels of experience in cranial techniques. The data collected during the experiments described above allowed a test of this hypothesis. One subject participated in experimental sessions with twelve (12) different examiners, ten of which examined the subject within a three-hr period on a single day. The examiners were divided into three groups based on years of experience in cranial techniques, and the results of this data analysis are shown in Table II.

Although the average cycle length as determined by the most experienced examiners was longer, and the CRI frequency lower, than the corresponding values for less experienced examiners, the differences among the three groups were not statistically significant. Groups of examiners with varying degrees of experience therefore appeared to agree on the basic CRI frequency of this one subject, supporting the hypothesis stated in the preceding paragraph. The overall variability in each of the measured and calculated variables is least among the most experienced examiners, suggesting that inter-examiner agreement increases with increasing experience, as one might expect. Assuming the most experienced examiners to be the most proficient and accurate in identifying the CRI and its components, the low frequency determined by this group and the good agreement among the examiners support the accuracy of the generally low CRI frequencies documented in this report.

calculation of extension duration vs measurement of extension duration: For a small group of subjects (n=3), an examiner was asked first to indicate the flexion phase of the CRI with the knee switch for 1-2 min, then to indicate the extension phase for a similar period. The results of these sessions are shown in Table III. The average cycle lengths (and therefore CRI frequencies) determined using either flexion or extension were found to be the same. The extension phase, when measured directly rather than calculated, was found to occupy only 45.9% of the CRI cycle. An average of 17.6% of the cycle in these subjects was neither flexion nor extension, corresponding perhaps to the "neutral zone" described previously(11). If each CRI cycle does include time within such a "neutral zone", then the period described as "extension" in the results described above obtained on the twenty-four (24) subjects should more appropriately be described as the "non-flexion" component of the CRI. Incidentally, the examiner for the sessions during which both flexion and extension were monitored reported that the extension phase was more difficult to isolate and identify than was the flexion phase and that flexion appeared more "active".

"still point": Related to the possible presence of a neutral period or position within the CRI cycle is the concept of "still points", frequently identified by examiners while monitoring the CRI of a subject. One subject in this study exhibited what the examiner described as a "still point" in the middle of an experimental session. The characteristics of this subject's CRI changed dramatically after this episode; the duration of the flexion phase and the total cycle length increased, indicating a decrease in CRI frequency. A copy of the experimental record of this event is shown in Figure 1.


The experiments described in this report were designed to provide the kind of quantitative information on the characteristics of the CRI that is not available in the literature and that is essential for the validation of any conceptual model for the origin of the cranial rhythmic impulse. The authors would appreciate feedback concerning the data itself or the methods by which the data were generated. The ultimate goal of research in this area should be to gather and evaluate information on the CRI in a manner that is both scientifically rigorous and clinically relevant to practitioners in the field.


CRI Cycle Characteristics(12)

duration of flexion (sec)      7.7 ± 1.4

duration of extension (sec)(13)      8.8 ± 1.6

CRI cycle length (sec)      16.5 ± 2.8

CRI frequency (cycles/min)(14)      3.7 ± 0.6

CRI amplitude(15)      3.0 ± 0.7


CRI Cycle Characteristics and Years of Examiner Experience(16)

  years of experience            < 5      5-10      > 10

   number of examiners             6          3           3

duration of flexion (sec)             7.1       6.9        8.5
                                                  ±2.3    ±1.8    ±0.9

duration of extension (sec)         9.5       8.8      10.4
                                               ±2.6    ±2.7    ±0.6

     CRI cycle length (sec)         16.7     15.7     18.9
                                               ±4.9    ±4.2    ±1.2

CRI frequency (cycles/min)       4.1      4.1      3.2
                                               ±1.5    ±1.1    ±0.2

             CRI amplitude             3.0      2.7      2.5
                                                 ±0.7    ±0.5    ±0.5


Comparison of CRI Cycle Characteristics
Derived from Measuring Either Flexion or Extension

CRI phase monitored              flexion           extension

duration of flexion (sec)          8.9                    n/a

duration of extension (sec)         n/a                 11.6

CRI cycle length (sec)            24.5                  25.2
                                             ±4.5                 ±5.6

CRI frequency (cycles/min)       2.5                   2.5
                                              ±0.4                 ±0.6

Legend for Figure 1

(Figure 1 not available)

A tracing of a portion of the record of a subject who demonstrated a "still point" during an experimental session. The small vertical arrows represent the onset of flexion. The distance between arrows represents cycle length. The still point was detected by the examiner at the point indicated by the asterisk (*). Four cycles with relatively short flexions are followed by the still point, followed in turn by three cycles with clearly longer flexion phases and cycle lengths. CRI frequency prior to the pause, or "still point", was 4.4 cycles/min; after, 2.5 cycles/min. The horizontal bar represents 30 seconds.


1. Sutherland, W.G. The Cranial Bowl. Mankato, MN: Free Press Company, 1939.

2. Wales, A.L. The work of William Garner Sutherland. JAOA 71:788-793, 1972.

3. Upledger, J.E. and J.D. Vredevoogd. CraniosacralTherapy, Chicago: Eastland Press, 1983, p. 6.

4. Kappler, R.E. Osteopathy in the cranial field: its history, scientific basis, and current status. Osteopathic Physician Feb 1979, pp. 13-18.

5. Magoun, H.I. (ed). Osteopathy in the Cranial Field. Kirksville, MO: The Journal Printing Company, 1976, p. 86.

6. Frymann, V.M. A study of the rhythmic motions of the living cranium. JAOA 70:928-945, 1971.

7. Adams, T., R.S. Heisey, M.C. Smith, and B.J. Briner. Parietal bone mobility in the anesthetized cat. JAOA 92:599-621, 1992.

8. Upledger, J.E. and Z. Karni. Mechano-electric patterns during craniosacral osteopathic diagnosis and treatment. JAOA 78:782-791, 1979.

9. Norton, J.M. A tissue pressure model for the palpatory perception of the cranial rhythmic impulse. JAOA 91:975-984, 1991.

10. Dowdy, S. and S. Wearden. Statistics for Research. New York: John Wiley & Sons, 1983.

11. Upledger, J.E. and J.D. Vredevoogd. op. cit., p. 7.

12. Values given are mean ± standard deviation for twenty-four (24) subjects.

13. Calculated for each subject by subtracting the duration of flexion from the total cycle length.

14. Calculated for each subject as 60 divided by the cycle length in seconds.

15. Graded by the examiner on a relative scale of 1-5 as described in the text.

16. Values are expressed as mean ± standard deviation; duration of extension, CRI frequency and CRI amplitude determined as described previously.