Online Activity Levels Are Related to Caffeine Dependency

Introduction

A range of technologies are being developed to appreciate a technology user's affective state^1–3 as this potentially allows a customization of computer system response (e.g., warnings, advertisements, recommendations) to a computer user's requirements. Hence, it could be important to understand how substance use influences activity levels and patterns of use as emerging technologies now afford the capacity to monitor, locate, and deliver inducements ⁴ or interventions contingent upon certain behaviors.^5–7 This article therefore develops techniques to compare levels of use of a common stimulant (caffeine) with online activity, specifically relating self-reported caffeine use to online tracking statistics.

Caffeine is legally available and commonly used as a cognitive enhancer to promote alertness and mental well-being.^8,9 The antagonism of adenosine A_2A receptors in the basal ganglia appears to be responsible for caffeine's stimulatory effects. ¹⁰ Although typically considered benign,^8,11 heavier caffeine users may exhibit dependence. ¹² DSM5 has recommended further consideration and research into the effects of caffeine.^13,14

Caffeine dependency may accompany behavioral addictions. Young ¹⁵ reported clinical links between Internet addiction and caffeine use. Relationships have been found between caffeine use and heavier video gaming ¹⁶ or problematic Internet use,^17,18 although the strength of such relationships may vary with other interests such as gambling. ¹⁹ Porter et al. ²⁰ noted that higher levels of video gaming were accompanied by caffeine use and poorer work/school performance.

Although some researchers have demonstrated caffeine's stimulant properties,^21,22 others suggest that caffeine just reverses the withdrawal experienced by heavier users.^23,24 Conceivably caffeine assists alertness, but does not improve complex performance ²⁵ that may be associated with the workplace.

Smith ²⁶ compared self-reported levels of caffeine use with reaction times at the start and end of a workday of 110 employed workers. Smith ²⁶ found that workers consuming more than 220 mg of caffeine during the day showed a greater increase in alertness and a smaller slowing of reaction time over the course of the day. Smith ²⁶ also reanalyzed data on self-reported caffeine consumption, cognitive failures, and workplace accidents of 1,253 white-collar workers. Smith ²⁶ found higher caffeine use was associated with fewer cognitive lapses and a reduced risk of accidents at work and reported similarly for a nonworking sample. ²⁷ Smith's studies examined self-reported caffeine use, alertness, and inattention, but real-world decisions may also require information integration.

Streufert et al. ²⁸ examined caffeine and information integration of 25 managers using complex simulations (previously found to correlate with managerial success). After caffeine deprivation, managers engaged in tasks that monitored breadth, depth, and speed of decisions. Caffeine deprivation led managers to make fewer and slower decisions. Caffeine deprived managers exhibited less diversity of action, and their decisions exhibited less utilization of opportunity and less applied initiative. When managers were given caffeine, decisions were faster, but exhibited less breadth (fewer alternatives considered). Excess caffeine led to action without consideration. ²⁹

Given that patterns of activity are sensitive to the effects of stimulants in animals, ³⁰ we sought an analog to study the effects of caffeine upon human decision-making. As a range of electronic devices now serve as data loggers, recording the time and amounts of activity, this article explored relationships between self-reported caffeine use and online activity tracked in a cross-section of students.

Methods

Results

There was a correlation between self-reported dose of caffeine and severity of reported dependence (r = 0.468, N = 62, p < 0.001). Of the 62 participants 11 had scores above four on the SDS. Participants who reported a dependence upon caffeine (SDS = 5.500, standard error [SE] = 0.500) had a significantly (t(60) = 2.368, p = 0.02, η² = 0.09) higher estimated dose of caffeine (252 mg, SE = 49) than the estimated dose (135 mg, SE = 20) of those who were not dependent upon caffeine (SDS = 0.942, SE = 0.172).

Within a 2-month period, there were no significant differences in the number of times the caffeine-dependent (mean = 59.300, SE = 8.921) participants logged in compared to noncaffeine-dependent participants (mean = 66.577, SE = 3.527) (t(60) = 0.815, p > 0.05). However, participants reporting caffeine dependence spent less time (mean = 4.716 hours, SE = 0.929) on Blackboard than those not dependent upon caffeine (mean = 8.854 hours, SE = 0.919) (t(30) = 3.166, p < 0.01, equal variances not assumed).

As may be seen in Table 1, as the severity of dependence increased, students accessed fewer items on Blackboard (r = −0.313, N = 62, p < 0.05), spent less time on Blackboard (r = −0.254, N = 62, p < 0.05), and had lower marks on multiple choice quizzes (r = −0.272, N = 62, p < 0.05). The correlation between severity of dependence and the number of times a student logged in to Blackboard was not significant (r = −0.172, N = 62, p > 0.05). In other words, students who were dependent on caffeine logged in for about the same number of sessions, but exhibited a reduced breadth and depth of processing.

Discussion

With ongoing developments in affective computing, there is a very real potential for information to be delivered online that is contingent upon a consumer's mood, interests, or behavior.^1–3 This study related stimulant use to 2 months of tracked online study activity. Although caffeine-dependent and noncaffeine-dependent participants logged in a comparable number of times, people with greater caffeine dependency accessed fewer files, spent less time on a study-related Web site, and performed more poorly on online coursework.

In this study, caffeine dependency did not seem to influence the number of sessions spent upon Blackboard, however, it did seem to reduce the time students spent online. Nevertheless, we doubt caffeine made study more efficient. Caffeine-dependent individuals spent less time online, but seemingly did less study, accessing fewer items, and more importantly, completed fewer multiple-choice quizzes. This implies differences in the way information was processed when caffeine dependent.

Quality of decision-making can be related to information search. ³⁷ Individuals who engage in less information search are more likely to choose risky, money-losing options. ³⁷ Whereas, superior information search and integration have been associated with higher intelligence, ³⁸ and a more balanced canvassing of alternatives has been linked to better foreign policy outcomes. ³⁹ In particular, a poorer use of time can be associated with poorer grades in student cohorts. ⁴⁰

Laboratory studies have focused upon elementary processing operations, ²⁷ rather than complex performance. ²⁵ Streufert et al. ²⁸ examined the alleviation of caffeine withdrawal, and found caffeine increased the speed of decisions, but reduced options considered. Excess caffeine consumption facilitated speed of response, but decreased the utilization of opportunity. ²⁹ Streufert's studies indicate that caffeine use leads to faster, but less considered decisions and this is also indicated by these data. These findings are also supported by observations that caffeine-dependent individuals report more hypervigilant (panicked) decisional styles. ⁴¹

Caffeine abstinence may cause deficits ⁴² that require caffeine to alleviate withdrawal. ¹² After a missed morning dose of caffeine, withdrawal symptoms resemble a mild flu. ⁴³ However, our caffeine-dependent individuals did not exhibit higher levels of absenteeism that might occur if morning withdrawal were confused with illness.

Most research indicates that caffeine confers benefits within the workplace, ⁴⁴ although caffeine use can be associated with dependence.^12,42 While caffeine is likely to assist attentional processes in the short term, these data suggest that caffeine confers less benefits with long-term use, such that acute effects diminish for dependent users. ²⁵ Nevertheless, it is possible that heavier caffeine users observed in this article were self-selected.^45,46 Students under a degree of stress and struggling with heavier study commitments might require caffeine to maintain the necessary alertness.

These findings may also overlook other behaviors that are not being monitored. ⁴⁷ This study tracked online study behaviors. Links have been observed between heavier video game play and caffeine use ¹⁶ and work/study problems. ²⁰ Relationships have otherwise been reported between problem Internet use and caffeine.^15,17–19 Conceivably caffeine use was associated with lower levels of study, but higher levels of recreational activity.

Acute doses of caffeine can increase a person's motor activity levels, ⁴⁸ and caffeine dependence may be associated with a greater range and intensity of caffeine use, purchasing, and consuming caffeine in more locations. ³¹ This article reveals that caffeine can also affect cognitive processes, ²⁵ but caffeine seems to reduce the breadth or depth of online file access^28,29 during study.

Systems are already available to monitor and trigger interventions based upon students' online learning performance.^49,50 Others are being developed to promote mental health and behavioral change.^5,7 As the capability to monitor online activity grows,^51–53 it may be possible in the future to infer cognitive status^1,7 from behavioral indicators such as dwell times, ⁵⁴ keystroke dynamics, ⁵⁵ or mouse clicks. ⁵⁶ Such behavioral indicators are already being sought in other jurisdictions to detect specific behavioral problems.^57,58 This study suggests that a reduced breadth or depth of search might be a possible indicator of stimulant use, but such Bayesian behavioral profiling would require considerably more research. ⁵⁹

Conclusion

It may be possible to develop online behavioral indicators of stimulant use in humans. Heavier caffeine use may be associated with a reduced breadth or depth of online search during study, but further research is required with larger samples.