Erratum to “Sex differences in gamma-glutamyltransferase in people aged 40–42 years in two Norwegian counties”
Article Outline
Abstract
Gamma-glutamyltransferase (GGT) is widely used as a marker of alcohol intake, although it is documented that other factors are also associated with serum levels of GGT. The total population of men and women aged 40–42 years in two Norwegian counties was invited to participate in a health survey program. GGT was measured in 8116 men and 8689 women—67 % of the eligible population. In sex-specific multiple regression analyses, GGT showed a positive association with body mass index (BMI), cholesterol, ln
triglycerides, systolic blood pressure, and number of drinks per 2 weeks for both men and women. Glucose and ‘years of smoking’ were significant in women only. Cups of boiled coffee per day and physical activity in spare time were inversely associated with GGT level for both men and women. A significant positive interaction between alcohol intake and BMI was observed for men but not for women. The use of GGT as a marker of alcohol consumption in middle-aged persons should take into account sex, BMI and drinking of boiled coffee.
Keywords: Gamma-glutamyltransferase, Coffee, Body mass index, Alcohol use
1. Introduction
Despite its well-established clinical use as a screening test for alcoholism and alcohol abuse, gamma-glutamyltransferase (GGT) is, in the normal population, influenced by many other factors.
An association between body mass index (BMI), serum cholesterol, blood pressure, level of physical activity, cigarette smoking and serum GGT has been reported in different populations (Arnesen et al., 1986, Nakanishi et al., 2000, Poikolainen and Vartiainen, 1997). Furthermore an inverse association between serum GGT and coffee intake (Arnesen et al., 1986, Nakanishi et al., 2000, Poikolainen and Vartiainen, 1997), and dietary factors (Arnesen et al., 1986, Nakajima et al., 1994) has also been documented.
It has also been reported that the relationship between alcohol consumption and GGT has varied with level of BMI (Poikolainen and Vartiainen, 1997) and coffee intake (Kono et al., 1994).
A sex difference in GGT level has been documented (Nilssen et al., 1990, Pintus and Mascia, 1996) and it has been suggested that the lower level of GGT for women is likely to be of a physiological nature.
In this report based on a health examination of the total population of men and women aged 40–42 years in two Norwegian counties we studied factors that relate to GGT.
2. Material and methods
All men and women aged 40–42 in the two Norwegian counties of Rogaland and Nordland were invited to participate in cardiovascular screening during 1994–1995. Determination of GGT was done in 16 805 subjects, i.e. 67 % of the eligible population.
A questionnaire (Bjartveit, 1979, Skurtveit et al., 2001) was mailed to the participants, who returned the questionnaire while attending the medical examination. Alcohol consumption was recorded using the questions: (1) ‘Are you total teetotaller?’; (2) ‘How many times do you usually drink alcohol during 1 month?’ and (3) ‘How many glasses of beer do you usually drink during 2 weeks?’ (the same questions were given about wine and spirits). Information about total number of drinks per 2 weeks was calculated as the total number of glasses of beer, wine and spirits each participant usually drinks during 2 weeks.
The medical examination comprised measurements of weight, height, systolic- and diastolic blood pressure and collection of venous non-fasting blood samples for measurements of serum total cholesterol, glucose, triglycerides and GGT.
3. Results
Table 1 shows biologic variables and other lifestyle factors according to different GGT level for men and women. BMI, blood lipids and blood pressure steadily increase with GGT. The correlation coefficients are significantly different in men and women. The intake of boiled coffee decreases with increasing GGT, whereas the intake of other types of coffee increases except in the last GGT category (in this context, boiled coffee excludes beverages made by filter and espresso techniques). Alcohol intake increases with GGT; this applies to all forms of alcohol.
Table 1. Biological variables and lifestyle variables according to four groups of different GGT levels
| Men | Women | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| GGT (units/l) | r | GGT (units/l) | r | |||||||
| <10 (n=216) | 10–29 (n=4696) | 30–59 (n=2374) | 60+ (n=830) | <10 (n=1465) | 10–29 (n=6293) | 30–59 (n=689) | 60+ (n=242) | |||
| BMI (kg/m2) | 23.8 | 25.2 | 26.9 | 27.5 | 0.323 | 23.6 | 24.5 | 26.3 | 26.5 | 0.211 |
| Total cholesterol (mmol/l) | 5.3 | 5.7 | 6.1 | 6.4 | 0.261 | 5.2 | 5.5 | 5.7 | 5.8 | 0.166 |
| Triglycerides (mmol/l) | 1.7 | 1.9 | 2.4 | 2.9 | 0.285 | 1.2 | 1.3 | 1.7 | 1.9 | 0.233 |
| Glucose (mmol/l) | 5.1 | 5.2 | 5.4 | 5.4 | 0.068 | 5.0 | 5.1 | 5.3 | 5.2 | 0.100 |
| Systolic blood pressure (mmHg) | 129 | 133 | 138 | 139 | 0.193 | 123 | 127 | 130 | 131 | 0.170 |
| Diastolic blood pressure (mmHg) | 78 | 81 | 85 | 87 | 0.252 | 75 | 77 | 80 | 80 | 0.160 |
| Coffee boiled (cups/day) | 3.2 | 2.4 | 2.2 | 2.0 | −0.059 | 2.2 | 1.8 | 1.3 | 1.5 | −0.078 |
| Other coffee (cups/day) | 2.7 | 3.6 | 4.0 | 3.8 | −0.045 | 2.8 | 3.4 | 3.5 | 3.0 | 0.052 |
| Tea (cups/day) | 0.5 | 0.5 | 0.5 | 0.5 | −0.004 | 0.8 | 0.7 | 0.7 | 0.7 | −0.017 |
| Years of smoking | 10.9 | 12.0 | 12.5 | 13.0 | 0.016 | 10.3 | 11.6 | 12.9 | 12.0 | 0.076 |
| Daily number of cigarettes | 1.6 | 1.6 | 1.6 | 1.6 | 0.011 | 1.7 | 1.6 | 1.6 | 1.6 | −0.068 |
| Alcohol use (occasions/month) | 2.4 | 3.2 | 3.6 | 4.3 | 0.113 | 1.9 | 2.1 | 2.2 | 2.0 | 0.014 |
| Alcohol use (drinks/2 weeks) | 4.4 | 5.7 | 7.1 | 8.9 | 0.163 | 2.7 | 3.0 | 3.7 | 3.1 | 0.051 |
| Alcohol use (glasses beer/2 weeks) | 1.9 | 2.6 | 3.3 | 4.3 | 0.134 | 0.7 | 0.7 | 1.0 | 1.0 | 0.035 |
| Alcohol use (glasses spirits/2 weeks) | 1.0 | 1.4 | 1.8 | 2.1 | 0.102 | 0.4 | 0.5 | 0.6 | 0.4 | 0.027 |
| Alcohol use (glasses wine/2 weeks) | 1.0 | 1.3 | 1.3 | 1.8 | 0.054 | 1.2 | 1.4 | 1.5 | 1.2 | 0.022 |
| Physical activity (hard)a | 2.3 | 2.3 | 2.1 | 2.0 | −0.083 | 1.9 | 1.9 | 1.8 | 1.8 | −0.053 |
| Physical activity (light)a | 3.1 | 3.0 | 2.9 | 2.9 | −0.028 | 3.1 | 3.1 | 3.0 | 3.0 | −0.031 |
a Physical activity per week is graded as 1, none; 2, less than 1 h; 3, 1–2; 4, 3 or more hours per week. |
In a multiple regression analysis1 the independent variables explained 20% of variation of GGT (lntransformed) in men and 11% of the variation in women. In both sexes, BMI, cholesterol, lntriglycerides, systolic blood pressure, use of alcoholic drinks per 2 weeks were positively associated with GGT. In men an increase in BMI and total cholesterol of one unit corresponds roughly to increases in GGT of 5 and 10%, respectively. The corresponding figures in women are 2 and 5%. An increase in alcohol consumption of one drink per day (14 drinks in 2 weeks) is associated with an increase in GGT of 19% in men and 14% in women.
Cups of boiled coffee per day and physical activity in spare time were associated with a lower GGT level in both men and women. An additional two cups of boiled coffee per day corresponds to 3 and 5% decreases in GGT in men and women. Other types of coffee were not related to GGT.
Fig. 1 shows the relationship between the number of alcoholic drinks per 2 weeks and GGT (geometric mean) at three levels of BMI. In men the relationship is stronger in the category with high BMI. Men and women in the low BMI category had a lower level of GGT than peers with high BMI, irrespective of the alcohol intake. A positive interaction between alcohol intake and BMI was found for men (P=0.026) but not for women.
Fig. 1.
Serum GGT levels for men and women by alcohol drinking and BMI level. Since distribution of GGT was skewed to the right, geometric means were used as a measure of central location.
4. Discussion
This large study of men and women within a narrow age range provides further evidence that factors other than alcohol are also associated with serum GGT activity. GGT was related to BMI, total cholesterol, triglycerides, blood pressure, intake of boiled coffee (inverse) and physical activity (inverse). The associations were largely stronger in men. In addition, a positive interaction between alcohol and BMI occurred only in men.
As expected, the use of alcohol was an important predictor for elevated GGT in both sexes, although the relationship was weaker in women. However, a multitude of relationships between GGT and biological and life-style variables hamper the use of elevated GGT as a marker of high alcohol intake. In this study, people with low BMI and high alcohol consumption had lower GGT level on average than people with high BMI and no alcohol intake at all.
In this study there were few heavy drinkers; drinkers with ethanol consumption over 30 ml per day represented 1.5% of men and 0.1% of women. However, it is likely that there was an underreporting of alcohol intake in this study as an estimate of alcohol intake per year per capita gives only half of what was expected from the official sales statistics in Norway.
An effect modification of BMI on the GGT–alcohol relationship has been reported (Poikolainen and Vartiainen, 1997). We found a similar effect modification but this was present only in men. This might be a chance finding or it might be due to a much lower level in women. For instance the GGT level in women with body mass above 32 kg/m2 is comparable to the GGT level in men with BMI in the range 20–24 kg/m2. Another possibility is that the different distribution of excess fat in men and women may be implicated. The inverse relationship between intake of boiled coffee and GGT is in accordance with other studies (Arnesen et al., 1986, Nilssen et al., 1990, Nilssen and Førde, 1994).
The coffee deterpenes cafestol and kahweol (non-triglyceride lipids presents in coffee) which are high in unfiltered boiled Scandinavian coffee (Thelle et al., 1987), raised the total cholesterol level (Tverdal et al., 1990). It has also been shown that cafestol decreases serum GGT (Weusten-Van der Wouw et al., 1994). A possible favourable effect of boiled coffee on the liver must be weighed against an unfavourable effect on serum cholesterol.
In conclusion, our study confirms the importance of GGT as a marker for alcohol use. However, other variables ranging from life-style to biological characteristics are also important determinants of serum GGT in the middle-aged population.
References
- . The Tromsø heart study: distribution of, and determinants for, gamma-glutamyltransferase in a free-living population. Scand. J. Clin. Lab. Invest. 1986;46:63–70
- . Cardiovascular Disease Study in Norwegian Counties. Preliminary Report. Acta. Med. Scand. 1979;634:59–63
- . Coffee and serum gamma-glutamyltransferase: a study of self-defense officials in Japan. Am. J. Epidemiol. 1994;139:723–727
- . Carbohydrate-related regulation of the ethanol-induced increase in serum gamma-glutamyltranspeptidase activity in adult men. Am. J. Clin. Nutr. 1994;87–92
- . Coffee consumption and decreased serum gamma-glutamyltransferase: a study of middle-aged Japanese men. Eur. J. Epidemiol. 2000;16:419–423
- . Seven-year longitudinal population study of change in gamma-glutamyltransferase: the Tromsø study. Am. J. Epidemiol. 1994;139:787–792
- . The Tromsø study: distribution and population determinants of gamma-glutamyltransferase. Am. J. Epidemiol. 1990;132:318–326
- . Distribution and population determinants og gamma-glutamyltransferase in a random sample of Sardinian inhabitants. Eur. J. Epidemiol. 1996;12:71–76
- . Determinants of gamma-glutamyltransferase: positive interaction with alcohol and body mass index, negative association with coffee. Am. J. Epidemiol. 1997;146:1019–1024
- . Alkoholvaner blant 40-åringer i Norge. Endring over tid. Nordisk. Alkohol- and Narkotikatidskrift. 2001;18:296–302
- . Coffee and cholesterol in epidemiological and experimental studies. Artherosclerosis. 1987;67:97–103
- . Coffee consumption and death from coronary heart disease in middle aged Norwegian men and women. BMJ. 1990;300:566–569
- . Identity of the cholesterol-raising factor from boiled coffee and its effects on liver function enzymes. J. Lipid. Res. 1994;35:721–733
- 1 Readers interested in additional data from multiple regression analysis may refer to the journal website at http://www.elsevier.com/locate /drugalcdep under ‘Supplementary Materials’.
PII: S0376-8716(02)00319-8
© 2002 Elsevier Science Ireland Ltd. All rights reserved.
Refers to article:
- Sex differences in gamma-glutamyltransferase in people aged 40–42 years in two Norwegian counties
