Effect of previous frequency of headache, duration of fasting and caffeine abstinence on perioperative headache

British Journal of Anaesthesia 1994; 72: 295-297

Effect of previous frequency of headache, duration of fasting and caffeine abstinence on perioperative headache L. NIKOLAJSEN, K. M. LARSEN AND O. KIERKEGAARD

SUMMARY

KEY WORDS Complications:headache. Pharmacology • caffeine Metabolism: fasting.

Minor postoperative morbidity, including headache, is common after general anaesthesia; the incidence of headache is reported to be as high as 54 % in patients prone to headache [1]. Headache is the primary symptom of caffeine withdrawal [2—6] and in a recent study, Fennelly, Galletly and Purdie concluded that perioperative headache was related to the daily consumption of caffeine [6]. In this study, we have examined the relationship between perioperative headache and various factors, including daily consumption of caffeine, in a multivariate fashion to eliminate interaction between single factors. PATIENTS AND METHODS

In an open study we examined 270 patients undergoing minor outpatient surgery under general anaesthesia. Four to six hours after operation each

LONE NIKOLAJSEN, M.D., KIM M. LARSEN, M.D., OLE KIER-

KEGAARD, M.D., Department of Anaesthesiology, Herning Hospital, Denmark. Accepted for Publication: September 30, 1993. Correspondence to L.N.

Downloaded from http://bja.oxfordjournals.org/ at Mount Royal University on July 15, 2015

We have examined the relationship between perioperative headache and various factors in 219 patients who fasted from midnight and underwent minor surgery under genera/ anaesthesia. Four to six hours after operation all patients completed a questionnaire on previous frequency of headache, daily consumption of caffeine and occurrence of perioperative headache. The duration of fasting, type of surgery, premedication and anaesthetic agents used were obtained from the anaesthetic record. After multivariate logistic regression analysis a significant risk of preoperative headache was found in patients who normally experienced headache more than twice a month (odds ratio (OR): 7.7; confidence interval (Cl): 2.9-20.1), had a daily caffeine consumption > 400 mg/24 h (OR: 5.0; Cl: 1.6-14.8) and who were anaesthetized after 12:00 (OR: 3.7; Cl: 1.4-9.8). The risk of postoperative headache was significantly greater in patients with preoperative headache (OR: 16.9; Cl: 6.5-43.8), daily caffeine consumption > 400 mgl 24 h (OR: 3.9; Cl: 1.5-9.6) and in those patients who received atracurium, which was similar to the risk of trachea! intubation (OR: 3.7; Cl: 1.7-7.9). (Br. J. Anaesth. 1994; 72: 295-297)

patient was asked to complete an anonymous questionnaire containing the following items: (1) gender, age; (2) previous frequency of headache: every day, once or twice a week, once or twice a month, never; (3) analgesic medications within the previous week. A list of nine analgesics prescribed commonly (caffeine- and non-caffeine-containing) was given to the respondents; (4) daily consumption of filtered, instant and decaffeinated coffee and tea (number of cups); (5) daily consumption of tobacco and weekly consumption of alcohol; (6) occurrence of headache before and after anaesthesia: (a) yes, a lot, (b) yes, some, (c) hardly any, (d) not at all, (e) don't know; (7) consumption of any food, beverages or tobacco from midnight and until the time of anaesthesia. The start of the period of fasting was assumed to be from midnight. For each patient, the time for start of anaesthesia, type of surgery, premedication and anaesthetic agents used were obtained from the anaesthetic record and listed on the questionnaire before completion by the patient. Caffeine consumption was calculated in milligrams per day using conversion factors derived from a previous source [7]: for example, filtered coffee contains caffeine 120 mg per cup, instant coffee 90 mg per cup and tea 50 mg per cup. Cola and chocolate consumption were not documented and the caffeine contribution from these sources was regarded as zero. Results were analysed using Spearman's rank correlation or the Mann-Whitney rank sum test. A forward stepwise logistic multiple regression model was used and the variable selection for entry into the final model was based on computed statistics for chisquare greater than 2.0. Variables were examined in dichotomous form with the following cut-off levels: previous headache more than "once or twice a month", actual headache more than "hardly any"; any drug, tobacco or alcohol consumption; fasting longer than until 12:00; and duration of anaesthesia more than 1 h. The cut-off level for caffeine consumption was chosen as > 400 mg/24 h based on a recent study of the calculated daily caffeine consumption in patients with and without perioperative headache [6]. For each variable the relative risk of headache, estimated as odds ratio (OR), was calcu-

BRITISH JOURNAL OF ANAESTHESIA

296

lated and in the multivariate model adjusted for interaction from other variables. P values less than 0.05 were considered to be statistically significant. RESULTS

TABLE I. Risk factors for preoperative headache. Odds ratio (OR) with 95 % confidence intervals (CI) and adjusted OR with 95 % CI after multivariate logistic analysis

Adjusted Headache more than twice a month Caffeine consumption > 400 mg/24 h Anaesthesia after 12:00 Females

OR

95 % CI

OR

Adjusted 95 % CI

5.6

2.4-12.6

7.7

2.9-20.1

3.6

1.4-9.7

5.0

1.6-14.8

4.7 3.4

1.8-10.6 1.3-9.2

3.7 2.5

1.4-9.8 0.8-7.8

TABLE II. Risk factors for postoperative headache. Odds ratio (OR) with 95 % confidence intervals (CI) and adjusted OR with 95 % CI after multivariate logistic analysis

OR Preoperative headache Caffeine consumption > 400 mg/24 h Use of atracurium Headache more than twice a month

95 % CI

15.4 6.5-36.1 2.6 1.2-5.5 1.6 3.5

0.9-2.8 1.6-7.6

Adjusted OR

Adjusted 95% CI

16.9

6.5-43.8 1.5-9.6

3.9

37 2.4

1.7-7.9 0.9-6.6

Bivariate analysis showed a correlation between postoperative headache and previous frequency of headache (P < 0.01; Spearman), preoperative headache (P < 0.000001; Spearman), and daily consumption of caffeine (P < 0.05; Spearman). There was no correlation between postoperative headache and the patient's age, gender, premedication, daily consumption of tobacco and alcohol, duration of fast or anaesthetic agents used. For multivariate logistic regression analysis the following variables were included: gender, previous frequency of headache, daily consumption of caffeine, tobacco and alcohol, preoperative headache, premedication, duration of fast and anaesthesia, anaesthetic agents used and consumption of food, beverages or tobacco from midnight and until the time of anaesthesia. There was a significant relationship between postoperative headache and preoperative headache, daily consumption of caffeine > 400 mg/24 h and the use of atracurium, which was similar to tracheal intubation (table II). The most important factor was preoperative headache. Fifteen patients (17 % of the smokers) had smoked on the morning of surgery; and despite being instructed to fast from midnight, 13 patients (6%) had eaten or drank. None had taken caffeinecontaining beverages. DISCUSSION

Perioperative headache is a distressing complication of anaesthesia, particularly in day-stay surgery. We found a highly significant correlation between previous frequency of headache and preoperative and postoperative headache. Others have found an increased incidence of perioperative headache (83 %) in those patients who were prone to headache and a decreased incidence (22%) in those who were not [1]. The association between postoperative headache and the use of atracurium, which was similar to tracheal intubation, may be explained by the use of

Downloaded from http://bja.oxfordjournals.org/ at Mount Royal University on July 15, 2015

Fully completed data sheets were obtained from 219 patients (81%). Forty patients did not return the questionnaire and 11 patients were excluded because of incomplete data sheets. Age, gender, type of operation, duration of fasting and anaesthesia, previous frequency of headache and daily consumption of caffeine did not differ in the excluded patients. Median age was 42.8 yr (range 18-80 yr); 7 1 % of patients were female. Operations included gynaecological (56%), urological (23%) and orthopaedic (21%) procedures. Eighty-one percent of the patients received premedication comprising apozepam 5—20 mg. Anaesthesia comprised i.v. induction with either thiopentone (55%) or propofol (45%), supplemented with nitrous oxide (100%), and fentanyl (68%) or alfentanil (47%), or both. Atracurium (33%) and suxamethonium (3%) were used for neuromuscular block. In 64% of the patients the trachea was not intubated. All patients underwent mechanical ventilation of the lungs. Monitoring was standard for low-risk patients: ECG monitoring was performed during anaesthesia, arterial pressure was determined by manual estimation using a sphygmomanometer and pulse oximetry was used as a measure of oxygen saturation. Fourteen percent of the patients had suffered from headache more than once a week before operation. Only two patients had taken caffeine-containing analgesics within the previous week. Caffeinated coffee or tea was consumed by 209 (95.4 %) patients. The calculated daily intake ranged from 40 to 2400 mg (median 682 mg). Eighteen percent of the patients suffered from headache before operation and 32 % suffered from headache after operation.

After bivariate analysis, we found a correlation between preoperative headache and previous frequency of headache (P < 0.0001; Spearman) and duration of fasting (P < 0.0001; Spearman). Women suffered from preoperative headache more than men (P<0.01; Mann-Whitney). There was no correlation between preoperative headache and the patient's age, premedication, daily consumption of tobacco, alcohol or caffeine, or consumption of food, beverages or tobacco from midnight and until the time of anaesthesia. A multivariate logistic regression procedure was used to analyse any relationship between preoperative headache and gender, previous frequency of headache, daily consumption of caffeine, tobacco or alcohol, premedication, duration of fasting and consumption of food, beverages or tobacco from midnight and until the time of anaesthesia. The risk of preoperative headache was significantly greater in patients with a previous frequency of headache more than twice a month, a daily consumption of caffeine > 400 mg/24 h and anaesthesia after 12:00 (table I). The most important factor was previous frequency of headache.

PERIOPERATIVE HEADACHE

perioperative headache that may be amenable to caffeine ingestion.

1. 2. 3.

4.

5. 6. 7.

REFERENCES McDowell SA, Dundee JW, Pandit SK. Paraanaesthetic headache in female patients. Anaesthesia 1970; 25: 334-340. Smith R. Caffeine withdrawal headache. Journal of Clinical Pharmacology and Therapeutics 1987; 12: 53-57. Griffiths RR, Bigelow GE, Liebson IA. Human coffee drinking: reinforcing and physical dependence producing effects of caffeine. Journal of Pharmacology and Experimental Therapeutics 1986; 239: 416-425. Van Dusseldorp M, Katan MB. Headache caused by caffeine withdrawal among moderate coffee drinkers switched from ordinary to decaffeinated coffee: a 12 week double blind trial. British Medical Journal 1990; 300: 1558-1559. Greden JF, Victor BS, Fontaine P, Lubetsky M. Caffeinewithdrawal headache: A clinical profile. Psychosomatics 1980; 21: 411-418. Fennelly M, Galletly DC, Purdie GI. Is caffeine withdrawal the mechanism of postoperative headache? Anesthesia and Analgesia 1991; 72: 449-453. Rail TW. Central nervous system stimulants. In: Gilman AG, Goodman L, Rail TW, Murad F, eds. The Pharmacological Basis of Therapeutics, 7th Edn. New York:

Macmillan Publishing, 1985; 589-603. 8. Miller M, Wishart HY, Nimmo WS. Gastric contents at induction of anaesthesia. Is a 4-hour fast necessary? British Journal of Anaesthesia 1983; 55: 1185-1187. 9. Splinter WM, Schaefer JD. Ingestion of clear fluids is safe for adolescents up to 3 h before anaesthesia. British Journal of Anaesthesia 1991; 66: 48-52. 10. Shevde K, Trivedi N. Effects of clear liquids on gastric volume and pH in healthy volunteers. Anesthesia and Analgesia 1991; 72: 528-531. 11. Goresky GV, Maltby JR. Editorial. Fasting guidelines for elective surgical patients. Canadian Journal of Anaesthesia 1990; 37: 493-495. 12. White BC, Lincoln CA, Pearce NW, Reeb R, Vaida C. Anxiety and muscle tension as a consequence of caffeine withdrawal. Science 1980; 209: 1547-1548. 13. Silverman K, Evans SM, Strain EC, Griffiths RR. Withdrawal syndrome after the double-blind cessation of caffeine consumption. New England Journal of Medicine 1992; 327: 1109-1114.

Downloaded from http://bja.oxfordjournals.org/ at Mount Royal University on July 15, 2015

a neuromuscular blocking agent, intubation or a combination of several factors. Intubation alone may cause headache by extension of the neck and compression of the arteries. For the preoperative period we found a strong correlation between the occurrence of headache and duration of fasting. A consistent finding in recent clinical studies has been that unlimited clear fluid ingestion up to 3 h before operation decreases perioperative morbidity and does not affect gastric contents [8-10]. Based on this finding, the Canadian Anaesthetists' Society now recommends that healthy patients undergoing elective surgery should be allowed unrestricted clear liquids until 3 h before surgery [11]. We found a positive relationship between patientreported headache and daily consumption of caffeine. This is in accordance with the findings of Fennelly, Galletly and Purdie [6]. Abrupt withdrawal of caffeine, in regular caffeine users, may precipitate withdrawal symptoms within 12-16 h, the most consistent being headache [2-6, 12, 13]. Caffeine causes cerebral vasoconstriction and withdrawal probably causes rebound vasodilatation which results in the development of headache [2]. Before anaesthesia, patients undergo periods of fasting similar to those capable of precipitating withdrawal. Our study has shown that perioperative headache was influenced by susceptibility to headache and duration of fasting. Also, our data suggest a relationship between daily consumption of caffeine and development of pre- and postoperative headache. The optimum duration of fasting has yet to be defined. It is worth considering if healthy patients, who are undergoing minor surgical procedures, should be permitted to ingest preoperative fluids. Patients with a large daily consumption of caffeine before operation have an increased incidence of

297