Assessment of withdrawal symptoms, psychological distress, and impairment of performance

• The tests were administered at the end of the second day of the caffeine and placebo periods because withdrawal symptoms are generally greatest 20 to 48 hours after the most recent consumption of caffeine.1 The tests were the Beck Depression Inventory, a 21-item questionnaire designed to assess the severity of depression11; the State-Trait Anxiety Inventory, a 40-item questionnaire designed to measure anxiety12; the revised Symptom Checklist-90 (SCL-90-R), a 90-item questionnaire designed to assess patterns of psychological symptoms 13; the Profile of Mood States, a 65-item questionnaire designed to assess mood states 14; and a 33-item caffeine-withdrawal questionnaire, a portion of which was previously reported to be sensitive in identifying caffeine withdrawal.
• After completing the questionnaires, the subjects completed a series of psychomotor-speed and cognitive-performance tests, including a digit-symbol substitution task,15 a reaction-time task,16 a number-recall task,17 a numerical-Stroop task,18 and a tapping task.

Analyses of Salivary Caffeine

• Five-milliliter samples of saliva were collected at each laboratory visit during the placebo and caffeine periods to assess compliance with the dietary restrictions.
• The results indicated that the subjects had not consumed caffeine during the placebo period; the mean values were 0.10 and 2.05 μg per milliliter during the placebo and caffeine periods, respectively. The sensitivity of the assay was 0.005 μg per milliliter.

 Statistical Analysis 

• A chi-square test with continuity correction was conducted to compare the percentage of subjects who violated the dietary restrictions by taking medications during the placebo and caffeine periods. Such a violation was defined as any use of an over-the-counter or prescription medication during a study period.

Results

• Ceasing to consume caffeine affected measures from every questionnaire completed by the subjects themselves, as well as one of the performance tasks.
• The severity of the effects that followed the cessation of caffeine consumption was assessed by comparing the percentage of subjects with abnormal responses on several measures during the base-line, placebo, and caffeine periods. The percentage of subjects who reported moderate or severe headache during the placebo period (52 percent) was significantly higher than the percentages during the base-line (2 percent) and caffeine (6 percent) periods (Fig. 1)
• On the Beck Depression Inventory, the State-Trait Anxiety Inventory, and the Profile of Mood States, the percentage of subjects with abnormal scores during the placebo period was significantly higher than the percentages during the base-line and caffeine periods (Fig. 2).
• The percentage of subjects who violated the dietary restrictions by taking medications was significantly higher during the placebo period than during the caffeine period (13 percent vs. 2 percent, P = 0.038).
• The dose of caffeine was significantly correlated with the difference in scores during the placebo and caffeine periods for only six measures: the ratings of drowsy/sleepy (r = 0.27, P = 0.033), yawning (r = 0.34, P = 0.006), and headache (r = 0.27, P = 0.033) on the withdrawal questionnaire; the SCL-90-R Positive Symptom Distress Index (r = 0.25, P = 0.049); the insomnia item on the Beck Depression Inventory (r = 0.26, P = 0.042); and the average latency on the reaction-time test (r = 0.31, P = 0.014).