Hyperalgesia After a Drinking Episode in Young Adult Binge Drinkers

A Cross-Sectional Study

Dokyoung S. You; Hunter A. Hahn; Thomas H. Welsh Jr.; Mary W. Meagher


Alcohol Alcohol. 2020;55(6):608-615. 

In This Article



The Texas A&M University Institutional Review Board approved the study procedures, which were conducted in accordance with the Helsinki Declaration. All participants provided informed consent. Inclusion criteria were healthy adults between 18 and 30 years of age. Exclusion criteria were (a) any chronic physical and mental health issues, (b) prescription medication use (except contraceptives and vitamins), (c) needle phobia and (d) an injury or skin condition on the feet (pain testing site).

Potential participants completed an online prescreening survey evaluating health status and drinking history. They were asked to select one that describes their drinking pattern, with options never, former and current drinker. Those who endorsed 'Never' (57%) were classified as abstainers. 'Current' drinkers (37%) were asked to complete the Daily Drinking Questionnaires (DDQ, Collins et al., 1985). On the DDQ, participants reported the number of standard drinks and hours spent drinking for a typical week in the last 30 days. Moderate drinking was defined as consuming <4 (women)/5 (men) standard drinks on any occasion.

Of the total 2226 people who completed the prescreening (12% current binge drinkers), 177 people participated in this study. Before the laboratory visit, participants were instructed to avoid any pain and allergy medicine 3 days before the experiment and avoid caffeine and vigorous exercise 4 h before the experiment. After quantitative sensory testing (QST), participants reported their last drinking on the exit survey. Most participants (85%) received class credit for their participation, while the rest received the opportunity to win one of three $100 gift cards in a raffle.


The following self-report measures assessed binge drinkers' baseline psychological characteristics and emotional responses to pain testing. The Depression Anxiety Stress Scales (DASS) assessed the severity of depression, anxiety and life stress symptoms over the past week (Antony et al., 1998). Total scores range from 0 to 42 for each symptom domain, with normal ranges <10 for depressive, <8 for anxiety and <15 for stress symptoms (Lovibond and Lovibond, 1995). The Hangover Symptoms Scale (HSS) was administered to assess the frequency of 13 hangover symptoms over the past 12 months (Slutske et al., 2003), with scores ranging from 0 (0% of the time) to 4 (100% of the time). Items were dichotomized to identify symptom presence (1 = 1–4) or absence (0 = 0) and then summed. Therefore, total scores range from 0 to 13, with higher scores indicating more hangover symptoms. The Acute Hangover Scale (AHS) was used to assess nine acute hangover symptoms (i.e. hangover, thirsty, tired, headache, dizziness, loss of appetite, stomachache, nausea and heart racing) rated on a 0: None to 7: Incapacitating scale (Rohsenow et al., 2007). The Spielberger State-Trait Anxiety Inventory-6 (STAI) evaluated emotional responses to pain testing, with total scores ranging from 6 to 24 and higher scores indicating higher anxiety (Marteau and Bekker, 1992).

Quantitative Sensory Testing. Because alcoholic polyneuropathy often affects lower extremities (Chopra and Tiwari, 2012), pain testing was conducted on the dorsum of the nondominant foot (L5 dermatome). The QST protocol of the German Research Network was used to measure cutaneous mechanical, heat and muscle pressure pain thresholds (PPTs) (Rolke et al., 2006). Briefly, the cutaneous mechanical pain threshold (MPT) was measured by applying von Frey filaments (0.6, 1.4, 4, 6, 15, 26 and 60 gF, Stoelting Co. USA) in an ascending, and then a descending order. To prevent sensory fatigue and local sensitization, each threshold test occurred at a different location with an intertrial interval ≥ 10s. Individual MPT was calculated as the geometric mean of the three series of ascending and descending stimulus intensity ratings. If participants reported no painful sensation to the highest force von Frey (60 gF), the next highest force (100 gF) in the series was used as their pain threshold (Cameron et al., 2008). Heat pain threshold (HPT) was measured using a 9-cm2 Peltier thermode (ATS, Medoc Ltd., Israel). Temperature was ramped at 1 °C/s (baseline temperature 32°C) until participants indicated their first pain sensation. Tests were repeated three times at 30 s intervals and the values were averaged to calculate individual HPT. Muscle PPT was measured three times using a handheld algometer (FPX 50 model, Wagner Instrument, USA) at 30 s intervals and the values averaged to calculate individual PPT.

Blood Collection. Blood was collected in a chilled EDTA tube before and after QST. The blood samples were centrifuged for 15 min at 1000 × g in a cold room (4°C), aliquoted into 0.5-ml tubes, and stored in a −80°C freezer until assayed.

Enzyme-linked Immunosorbent Assay. Duplicate samples were analyzed according to the manufacturer's manual. Abnova (No KA1877) and IBL international (RE52061) kits were used for Epinephrine and Cortisol, respectively.

Physiological Measures. Continuous heart rate (HR), skin conductance level (SCL) and respiration rate (RR) were sampled at 1000 Hz using a MP150 (BIOPAC Systems, USA), interfaced with AcqKnowledge 4.2 for data acquisition. LabVIEW 8.0 was also used to generate a digital signal from the National Instrument 6008USB device connected to a BIOPAC Systems STP100C and UIM100C. A finite impulse response (FIR) bandpass filter was used to remove movement artifacts from HR (0.5 and 35 Hz) (Ruha et al., 1997) and RR (0.05 and 1.0 Hz). To remove artifacts from SCL, a 1 Hz FIR lowpass filter was used. After filtering, all data were visually inspected. Filtering corrected most noise in RR and SCL. However, HR data from three subjects containing uncorrectable noises (e.g. extreme movement artifacts) were not analyzed.


To recruit participants who voluntarily consumed alcohol, the experiment was conducted between Thursday and Sunday. For ethical reasons, participants did not receive any instructions about alcohol consumption. Participants who chose to consume alcohol within 2 days were classified in 'recent' alcohol drinking groups. Additionally, the experiment time was between 12 pm and 6 pm to minimize the effect of diurnal variation on pain and stress hormones. Participants were blinded to the study hypotheses and the experimenter conducting QSTs was blinded to participants' drinking patterns.

Figure 1 depicts the experiment timeline. Before and after the blood draw, BP, HR and body temperature were measured to make sure the levels were within normal limits. In the testing room, baseline physiological data were collected 5 min after the survey. The interval between different QSTs was 5 min.

Figure 1.

Experiment timeline [1 h]. V/S: Vital Sign including Blood Pressure, HR, and BT. MPT: Cutaneous Mechanical Pain Threshold, HPT: Heat Pain Threshold, PPT: Muscle Pressure Pain Threshold. Demo: Demographics, STAI: State-Trait Anxiety Inventory-6; DASS (Depression Anxiety Stress Scales); AHS: Acute Hangover Scale; HSS: Hangover Symptom Scale; HR: Heart Rate; BT: Body Temperature; SCL: Skin Conductance Level; RR: Respiration Rate.

Statistical Analysis

To examine the first hypothesis, a multivariate analysis of variance (MANOVA) with three pain thresholds was followed by a least significant difference (LSD) test. For the second hypothesis, a repeated-measures analysis of variance (ANOVA) compared the group difference in stress hormones and effective responses to QST. Then, an analysis of covariance (ANCOVA) examined how much of the group difference in pain sensitivity would be accounted for by stress hormones and negative effect. All statistical tests were two-tailed and P values < 0.05 were considered significant. SPSS 22 was used for all analyses.

Sample Size Calculation. This study was designed to recruit five naturally occurring groups: (a) abstainers, (b) moderate drinkers with a recent drinking episode, (c) moderate drinkers without a recent drinking episode, (d) binge drinkers with a recent drinking episode (main group of interest) and (e) binge drinkers without a recent drinking episode. Large effects of alcohol withdrawal on pain threshold were reported (Cohen's d > 1.65, Dina et al., 2006, Dina et al., 2008; Jochum et al., 2010). Sample size calculation indicated that at least 25 participants were needed in each group to achieve 95% power and an α of 0.05 for a between-subjects design. Our main interest was to evaluate group differences by drinking patterns and we sought to recruit an equal number of men and women per group. Thus, all the data were analyzed and presented by drinking group, but gender and gender by group interactions were examined and significant results were noted, if any.

Missing Data. There were few missing values (~1.7%), which were replaced with overall means. For pain data, three participants did not undergo HPT tests because of equipment malfunctioning. Therefore, these three participants' physiological data were replaced with the overall means. For psychological state variables, one binge drinker did not complete the STAI at baseline.


Table 1 depicts the demographics. There were no group differences in gender, race/ethnicity or cigarette use (Ps > 0.097). The abstainer group was about 1 year younger than the other groups, but all were young adults.

The age of first alcohol use was somewhat older in the moderate drinkers than the others, but no difference was found in years of regular alcohol use (P = 0.448). An ANOVA indicated no group difference in drinking frequency endorsed on the DDQ (P = 0.488); all types of drinkers endorsed drinking about 2–3 times a month. The respective means of typical drinks were 2.5 (SD = 1.6) and 3.1 drinks (SD = 2.4) for female and male moderate drinkers, and 6.6 (SD = 3.2) and 10.3 (SD = 10.2) for female and male binge drinkers. On the HSS, binge drinkers reported more hangover symptoms for the past 12 months. Notably, the HSS scores were even higher in the binge drinkers with recent drinking than those without it. Additionally, more hangover symptoms were reported on the AHS by binge drinkers with recent drinking, being consistent with the self-reported recent drinking episode. The respective means of recent drinks were 1.8 (SD = 1.0) and 2.0 (SD = 1.5) for female and male moderate drinkers, and 5.0 (SD = 1.6) and 6.0 (SD = 4.9) for female and male binge drinkers.

Baseline Psychological and Physiological Characteristics

Baseline psychological characteristics were similar between the groups (Kruskal–Wallis tests' Ps > 0.136) except for depressive symptoms (P = 0.022). Post-hoc comparisons with the Mann–Whitney test indicated that moderate and binge drinker groups reported more depressive symptoms than the abstainer group (Table 2). Depressive symptoms did not differ between abstainers and moderate drinkers with recent drinking (P = 0.075).

In comparing the baseline physiological states, a one-way ANOVA indicated difference only in SCLs, F(4, 283) = 3.12, P = 0.017. LSD post-hoc analysis indicated that higher SCLs were observed in the two moderate drinker groups and the binge drinker group with recent drinking, Ps < 0.05 (Table 2).