J Manipulative Physiol Ther 2004 (Feb); 27 (2): 109–117 ~ FULL TEXT
John Hart, DC, Edward F Owens Jr, DC
Interim Coodrinator of Research,
Sherman College of Straight Chiropractic,
Spartanburg, SC 29304, USA.
BACKGROUND: Paraspinal thermography has been used by chiropractors since 1924. One method of its interpretation is with the use of "pattern analysis"-a method that assesses temperature differentials (patterns). This, in turn, theoretically provides information about nervous system function. When a warm back is exposed to the cooler air in the examining room, the skin temperature, in general, drops but the differentials could remain fairly constant. OBJECTIVE: To determine what changes occur in paraspinal heat patterns when the back is exposed to room temperature.
STUDY DESIGN: Observational; measures repeated at 5-minute intervals for 31 minutes.
METHODS: Thirty subjects were scanned with digital infrared thermographic instrumentation every 5 minutes over a 31-minute period for a total of 7 readings. A computerized calculation of percent similarity between consecutive comparisons of the readings was then performed to determine if and when the pattern stabilized.
RESULTS: Cervical spine temperatures remained constant while lower back temperatures, in general, decreased for the entire 31-minute recording period. Although the results varied among subjects, on the average, the patterns stabilized after 16 minutes.
CONCLUSIONS: Once the patient's back is exposed to cooler room temperature, the skin temperature decreases constantly for 31 minutes; however, the pattern becomes stable after 16 minutes. Readings taken for the purpose of pattern analysis during this 16-minute period may be unreliable for some patients. Therefore, a 16-minute acclimation period is recommended. Further research is needed to not only verify this finding with the same equipment in a separate experiment but to verify it as well with other types of temperature instrumentation.
From the Full-Text Article:
On average, for all 30 subjects in this study, the pattern comparisons of the earlier-acclimated readings (between the 1-minute and 6-minute acclimated readings) were approximately 10% less similar than comparisons of the later-acclimated readings (ie, between the 16-minute and 21-minute readings). A 10% difference, particularly in the higher percentage ranges (ie, 70% to 80%), may not be noticeable with visual inspection. However, a 30% difference (as seen with Subject 7, Fig 3, A through H) was noticeable with visual inspection. In the case of the 10% difference, the computer is able to detect and document differences that may escape the human eye. There were a number of other subjects that had visually noticeable pattern changes between the 1-minute and the eventually stabilized pattern with varying percentages. The clinical significance of these percentages, as well as the significance of the pattern theory itself, needs further investigation.
A physiological test can change spontaneously, without any intervention, as seen in subject 7's comparison of the 1-minute and 31-minute readings. On average, the readings cooled down over the 31-minute study period, unlike the DeBoer et al  reliability study, which found that the temperature increased by approximately one half of 1°F between 2 scans over a 1-hour study period. The reason for this is not readily apparent. Most subjects demonstrated stabilized patterns around the 16-minute reading. A small number of subjects, however, had high percentages of similarity with the earlier acclimated readings.
A change in the characteristic shape of a thermographic reading is seen as a good sign in pattern analysis.  If there was an intervention (spinal adjustment) between subject 7's 1-minute and 16-minute readings (Fig 3, C), then such changes might be erroneously attributed to the intervention. Based on this study, the chiropractor using pattern analysis would probably obtain more accurate analyses by utilizing a 16-minute acclimation period before taking the initial thermographic reading. A change of thermal pattern is associated with nervous system function, whether the change is during acclimation or following an adjustment of the spine. However, if the pattern is stable prior to the adjustment (via acclimation), then any changes following the adjustment would be more likely attributable to the adjustment rather than to acclimation.
Although some subjects might show less change of pattern in the cervical area (ie, Fig 4, A), equilibration of cervical-only readings would probably still be an important consideration, because the shirt collar typically covers the spine to the C5 vertebral level, thus insulating the area from T1 to C5 from the cooler room temperature.
In this study, readings taken with only 1 minute of room temperature acclimation tended to show more variability than readings taken with 16 to 31 minutes of equilibration. There was an average difference of approximately 10% between the earliest comparisons of readings (1 minute and 6 minutes of acclimation) as compared with later comparisons (ie, between 16 minutes and 21 minutes of acclimation).
Some subjects in this study (such as subject 7) exhibited greater pattern changes than other cases (such as subject 8) during equilibration. Since it would be difficult to predict what patient would or would not be significantly affected by the lack of room temperature acclimation, it seems prudent to acclimate all patients to room temperature for at least 16 minutes prior to the thermographic examination.
Further research is needed to determine what percent change is clinically significant, as well as to investigate the pattern theory itself. Acclimating the patient prior to thermal scanning would seem to lessen the possibility of erroneous pattern analysis.