SHOULD WE CONSIDER AN ACCEPTABLE DRINKING LEVEL SPECIFICALLY FOR POSTMENOPAUSAL WOMEN? PRELIMINARY FINDINGS FROM THE POSTMENOPAUSAL HEALTH DISPARITIES STUDY

JUDITH S. GAVALER

Department of Pharmaceutical Sciences, University of Pittsburgh School of Pharmacy, Pittsburgh, PA, USA

Author to whom correspondence should be addressed at: Department of Pharmaceutical Sciences, University of Pittsburgh School of Pharmacy, 1014 Salk Hall, Pittsburgh, PA 15261, USA. Tel.: +1 412 441 4229/+1 412 383 7312; Fax: +1 412 624 1850; E-mail: gavaler{at}pitt.edu/judygavaler{at}comcast.net

(Received 8 March 2005; first review notified 22 April 2005; accepted in revised form 2 May 2005)


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Aims: There are substantial neuroendocrine differences between postmenopausal (PMP) women and women with cyclic ovarian function; thus there are differences in hormonal responses to ‘acceptable’ levels of drinking, i.e. 7 total weekly drinks (TWD). The aim is to evaluate whether alcoholic beverage consumption for PMP women should be the same for all women, as is currently the case, or different from that for women with cyclic ovarian function. Methods: Carefully collected and cross-checked current drinking data in 318 PMP women from a spectrum of racial and ethnic groups have been analysed to determine if a plateau in estradiol (E2) levels at a particular alcohol consumption level other than 7 TWD can be identified as the drinking level above which no additional benefit in E2 levels occurs. Results: Levels of E2 are significantly different among abstainers and among drinkers within all racial and ethnic groups, with the disturbing exception of black PMP women. Although there are significant differences among racial and ethnic groups in the proportions of women who drink, there are no differences in TWD consumed during the previous month. Between ≤5 TWD and ≤6 TWD, a plateau in the correlation of E2 with increasing drinks/week occurs, and the range of the unstandardized multiple regression coefficient ± the standard error do not overlap between these same two drinking levels (P < 0.05). Conclusions: Consideration should be given to lowering the acceptable level of drinking for PMP women 5 TWD. Additional studies in which sample sizes for each minority racial and ethnic are larger need to be performed in the future.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Alcohol consumption has ubiquitous effects in postmenopausal (PMP) women from a spectrum of racial and ethnic groups on unperturbed PMP hormonal status (Gavaler et al., 2002Go), as well as on hormonal status perturbed by the use of estrogen replacement treatment (Gavaler, 2002Go, 2003Go; Gavaler et al., 2004Go). Alcohol is a significant negative predictor of estimates of PMP body fat mass (Gavaler and Rosenblum, 2003Go). In addition, among currently drinking PMP women prescription-treated blood pressure and diabetes, as well as obesity measures are significantly lower, while fitness and weekly exercise frequency as well as total calories and percentage of total calories as carbohydrates are significantly higher among drinkers (Gavaler et al., 2004Go).

The upper USDA recommendation for alcoholic beverage consumption for adult women of all ages has been 1 drink per day or 7 total weekly drinks (TWD), while in men the upper limit has been 2 drinks per day or 14 TWD (USDA, 1995Go). These levels have remained unchanged for years (Gomberg, 1944Go; Dufour, 1999Go; Blow and Barry, 2002Go; Register et al., 2002Go). One must wonder, however, what methods were used to determine these drinking limit recommendations in either gender; this is an important question because these recommended limits are viewed by the public as being safe.

In evaluating what may be an acceptable level of alcohol consumption for women, it is to be noted that women cannot be viewed as a homogeneous group. Specifically, PMP women and women with cyclic ovarian function differ substantially in the neuroendocrine control of hormone production. Thus, in terms of what may be an acceptable drinking level, PMP women need to be considered separately from women with cyclic ovarian function (Gavaler, 1990Go; McCoy, 2002Go).

In PMP women the hypothalamic–pituitary–gonadal axis is no longer intact because the ovaries can no longer respond (natural menopause), or are absent (bilateral ovariectomy performed at time of hysterectomy). It is important to note that women who have undergone a hysterectomy alone or with the removal of a single ovary still experience a natural menopause.

Further, PMP women are estrogen deficient by definition in the PMP state. Estrogen is produced by the conversion of androgens primarily of adrenal cortex origin by CYP19 (aromatase). This enzyme converts testosterone to estradiol (E2) and androstenedione to estrone. Further, it is particularly important to note that the activity of CYP19 is increased by alcohol consumption, increased age, in PMP women as compared to younger women, and by obesity (Hemsell et al., 1974Go; Bird et al., 1978Go; Crilly et al., 1979Go; Longcope et al., 1980Go; Meldrum et al., 1981Go; Ginsburg, et al., 1996Go; Djordjevic et al., 1998Go). In addition, alcohol-induced hormonal changes vary as an effect of race and ethnicity (Gavaler et al., 2002Go).

The decreased estrogenization of the PMP state originally led to the use of estrogen replacement therapy in PMP women. Thus, if acceptable drinking levels increase circulating E2 concentrations levels among PMP women can be viewed as being beneficial up to a point in alcohol consumption, then that point of alcohol consumption at which the maximum benefit occurs becomes an issue to be determined in PMP women from all racial and ethnic groups. The maximum benefit would be the level of alcohol consumption at which no further increase in E2 levels can be detected. E2 has been selected as the estrogen to be examined because E2 exerts a greater biologic effect than estrone due to the higher binding affinity of E2 to estrogen receptor binding sites (Schutt and Cox, 1972Go; Rosenblum et al., 1993Go). Further, it is important to emphasize that identifying the alcohol consumption level of maximum benefit in E2 concentrations is particularly important as the practice of prophylactic bilateral ovariectomy is increasing (Conklin et al., 1996Go: Kontoravdis et al., 1996Go; Gavaler, 2004Go), while the use of previously conventional estrogen replacement therapy has diminished substantially (Austin et al., 2003Go; Hersh et al., 2004Go).

The goal is to assess the possibility that an acceptable drinking level for PMP women may be lower than the USDA level of alcohol consumption for women as a group of 1 drink per day or a total of 7 drinks per week by examining statistically at what level of alcohol consumption no further benefit or increase in E2 levels can be identified.


    METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Study design
As the PHDStudy has been described previously (Gavaler et al., 1999Go, 2004Go), the description of methods will be brief. The study is a biomedical and psychosocial study, cross-sectional with historical prospective elements in design. Over sampling minority PMP women in Oklahoma was a major focus. The study in Oklahoma began in 1993 and recruitment ended in mid 1999, after which the study was moved to Pittsburgh. Data from Oklahoma PMP women are included in this report. Informed signed consent was obtained prior to enrollment into the study in compliance with the protocol approved by the Institutional Review Board. The study entry criterion was being PMP, confirmed by the ratio of follicle stimulating hormone (FSH) to luteinizing hormone (LH) having a value greater than ‘1’. The exclusion criteria included a history of current cancer or any disease severe enough to prevent a potential subject from coming or being brought to the clinic.

Data collection
To ensure the accuracy of the alcohol consumption data, the demographic questionnaire included quantity/frequency questions for the last month and for several other time points, a never-ever (in the past but not currently)-now drinking item, questions about preferred beverages, and drinking situations, all of which were spread throughout the demographic questionnaire. Alcohol data from the 3-day food and beverage record was also used. At the morning clinic visit, blood specimens were obtained for hormone measurements, and questions about when the most recent drink was consumed were also asked. Together, these data sources, and contacting the participant if necessary, resulted in the need to make corrections in <5% of participants. Using the algorithm of Turner, where 1 fluid ounce (fl oz) equals 48.41 ml and contains 22.3 g of alcohol, 12 fl oz of beer, 5 fl oz of wine and 1.5 fl oz of 80-proof liquor were set to one standard serving or drink (Turner, 1990Go).

Hormone assays
Hormone levels were measured using radioimmunoassay kits which have been used for years in our laboratory. For E2, the intra-assay coefficient of variation (CV) was 4.1%, and the inter-assay CV was 6.6%; the detection limit was 5 pg/ml. For LH, the standard was the first IRP 68/40; for FSH the standard was the second IRP 78/459. The intra-assay CV was 3.8% for LH, and 4.7% for FSH while the inter-assay CV was 5.8% for LH, and 6.1% for FSH.

Statistical analyses
Statistical tests used included one-way analysis of variance with the Tukey multiple comparison procedure, Pearson bivariate correlation coefficients, the independent group t-test, and multiple linear regression. Logarithmic transformations were used for all hormone levels. Data are shown as the mean ± the standard error of the untransformed data, as logarithms can be difficult to interpret. A two-tailed P-value < 0.05 was considered to be statistically significant.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Although the proportion of women who were current drinkers differed significantly among racial and ethnic groups, there are no detectable differences in the amount of alcohol consumed, likely due to the range of TWD within each group (Table 1). E2 levels in both drinkers and abstainers were significantly different among groups and within groups with the notable exception of black PMP women in whom this pattern was broken; there were no currently drinking Asian women.


View this table:
[in this window]
[in a new window]
 
Table 1. Characteristics of the study population using no hormone therapy

 
As there were no detectable differences in TWD among racial and ethnic groups, the entire population has been analysed in Table 2. Although the correlation coefficients of E2 with each increasing level of drinking, there appears to be a plateau after between 5 and 6 TWD. To test whether the apparent plateau was statistically significant, multiple linear regression with E2 as the dependent variable and TWD category as the predictor variable was used to generate the unstandardized regression coefficient and the standard error so that the range could be calculated. As may be seen, there is no overlap between the ranges for 5 and 6 TWD. Thus, it can be said that at 5 TWD there is a significant (P < 0.05) difference in the effect of drinking on E2 levels, which can be viewed as the maximum beneficial effect of drinking on E2 levels. As can also be seen, there was a significant difference in E2 concentrations between 5 and 6 TWD.


View this table:
[in this window]
[in a new window]
 
Table 2. Relationship between E2 and levels of alcohol consumption among PMP women

 
In Table 3, in addition to TWD, the relative roles of other factors known to influence PMP E2 levels [i.e. age, body mass index (BMI), bilateral ovariectomy (OVEX), and the estimated conversion of testosterone to E2] have been examined among PMP women within each racial and ethnic group drinking ≤5 TWD and drinking <7 TWD. In the PMP women consuming <5 TWD, with the exception of Asian American none of whom are drinkers and black PMP women who have already been demonstrated to have a paradoxical response to drinking, TWD is a significant predictor of E2 levels in all other racial and ethnic groups. In the women drinking <7 TWD, TWD is a significant E2 predictor only among Hispanic women, while the estimated conversion of testosterone to E2 enters the model in all racial and ethnic groups, even in Asian women none of whom drink. It is important to note that the role of age, OVEX and BMI do not show a consistent pattern among racial and ethnic group PMP women.


View this table:
[in this window]
[in a new window]
 
Table 3. Predictors of E2 levels in PMP women using no hormone therapy

 
Parenthetically, we have previously reported substantial variability in the E2 levels achieved among PMP women treated with estrogen replacement therapy (Gavaler, 2003Go; Gavaler et al., 2004Go), and have created categories based on the mean and standard deviation (SD) in 356 control, untreated PMP women. Treated women who achieved E2 levels less than or equal to the control mean +1 SD were categorized as being non-responders; treated women who achieved E2 levels greater than the control mean +6 SD were categorized as being over-responders; the remainder of treated women were characterized as therapeutic responders. As alcohol consumption has been demonstrated to be a predictor variable of the E2 levels achieved, for the sake of completeness, we briefly examined both the percentage of drinkers and TWD in each of the three categories in PMP women treated with either an oral or patch formulation. Among the 124 non-responders, 30.6% were current drinkers and the mean TWD was 3.3 ± 0.6. Among the 183 therapeutic-responders, 44.8% were current drinkers and the mean TWD was 4.7 ± 1.0. Among the 49 over-responders, 49.0% were current drinkers and the mean TWD was 5.7 + 1.4. Significantly there were differences among racial and ethnic groups only among current drinkers in the therapeutic-response group. Such a clear demonstration of the utility of setting acceptable drinking levels at ≤5 TWD in estrogen replacement treated was unanticipated.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
This study may be the first report in which consideration of the possibility of a different acceptable drinking level for PMP women has been systematically examined. The hypothesis was based on the difference in the control of E2 production on PMP women as opposed to women with cyclic ovarian women. In fact, this study may be the first hypothesis-based study to truly probe acceptable drinking levels in any group using a variety of statistical approaches that have substantiated each other. Now in PMP women, we can actually speak of ‘acceptable’ drinking; we no longer need to be bothered by how best to describe non-alcoholic drinking: social? moderate? low? light?

The plateau in E2 levels with increasing categories of alcohol has now been statistically confirmed. Further, the fact that tripling of the alcohol dose above the dose shown to be acceptable is required to breach the plateau in E2 levels in PMP women finally explains certain previous clinical observations (Gavaler, 1993Go; Gavaler et al., 1994Go). In addition, over 25 years ago when I was using a triplet-matched isocaloric liquid diet in ovariectomized rats, I could not figure out why E2 levels in the high dose animals did not differ from the middle dose animals, while the middle dose animals differed significantly from the control-fed rats. Now I know.

Together, these findings demonstrate that 5 TWD provide the maximum benefit in terms of the estrogenization of PMP women. These findings also indirectly demonstrate the need to use various methods of collecting alcoholic beverage consumption to ensure accuracy.

The vast majority of these new findings, however, have come from the data in the entire group of PMP women not treated with previously-conventional estrogen replacement, as well as in the entire group of PMP women treated with conventional oral or patch estrogen therapy (we now insist that our E2 levels are monitored regularly). The major question which must now be pursued is whether or not these findings can be replicated in the spectrum of racial and ethnic groups. We do know that among American Indian, Black, Hispanic, Jewish, White/American Indian Blend and White PMP women not treated with estrogen replacement in these analyses that significantly in higher proportions within each group who drink consume ≤5 TWD (P < 0.05 within each group). We know that the percentages of current drinkers differ significantly among racial and ethnic groups (P = 0.026). We also know that E2 concentrations among both drinkers (P = 0.026) and abstainers (P = 0.06) differ significantly among racial and ethnic groups. Further, among PMP women treated with conventional estrogen replacement therapy, we know that in the 124 therapeutic-responders there are significant differences among the racial and ethnic groups in E2 levels (P = 0.049 and in the proportion of drinkers (P = 0.046); we also know that among the 49 over-responders E2 levels (P = 0.051 and TWD (P = 0.047) differ significantly among racial and ethnic groups.

Clearly more research will be required to elucidate the paradoxical arithmetic decrease in E2 levels in black PMP women. A great deal more work will be required to expand these findings to apply within each racial and ethnic group analysed separately.


    ACKNOWLEDGEMENTS
 
This work has been supported in part by grants from the National Institute on Alcohol Abuse and Alcoholism (R01–11184 and R01–06672), the Office of Women's Health Research, and the Office of Minority Health Research.


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Austin, P. C., Mamdani, M. M., Tu, K. and Jaakkimmainen, L. (2003) Prescriptions for estrogen replacement therapy in Ontario before and after publication of the Women's Health Initiative trial. Journal of the American Medical Association 289, 3241–3242.[Free Full Text]

Bird, C. E., Finnis, W., Boromand, K., Murphy, J. and Clark, A. F. (1978) Kinetics of testosterone metabolism in normal postmenopausal women and women with breast cancer. Steroids 2, 323–335.[CrossRef]

Blow, F. C. and Barry, K. L. (2002) Use and misuse of alcohol among older women. Alcohol Research and Health 26, 308–315.[Medline]

Conklin, B. H., McGuire, P. A., Weiler, P. L. and Webb, D. I. (1996) A survey of the practice of prophylactic oophorectomy by gynecologists in the state of Alaska. Alaska Medicine 38, 71–74.[Medline]

Crilly, R. G., Marshall, D. H. and Nordin B. E. C. (1979) The effect of age on plasma androstenedione concentration in oophorectomized women. Clinical Endocrinology 10, 199–201.[ISI][Medline]

Djordjevic, D., Nikolic, J. and Stefanovic, V. (1998) Ethanol interactions with other cytochrome P450 substrates including drugs, xenobiotics and carcinogens. Patholgy and Biology 46, 760–770.

Dufour, M. C. (1999) What is moderate drinking? Defining ‘drinks’ and drinking levels. Alcohol Research and Health 23, 5–14.[ISI][Medline]

Gavaler, J. S. (1993) Effects of alcohol use and abuse on the endocrine status in expanded study samples of postmenopausal women. In Alcohol and the Endocrine System, Zakhari, S. ed., pp. 171–187. National Institute on Alcohol Abuse & Alcohol Research Monograph No. 23, NIH Pub. 93–3533. Washington, DC.

Gavaler, J. S. (1990) Alcohol and hormones in women: the reproductive years and the postmenopausal years. In Women: Alcohol and Other Drugs, Engs, R. C. ed., pp. 43–51. Kendall/Hunt Publishing, New York.

Gavaler, J. S., Deal, S. R., Rosenblum, E. R., Tivis, L. J. and Fagiuoli, S. (1994) Alcohol effects on postmenopausal hormone levels: normal women and women with alcohol-induced cirrhosis. Alcologia 6, 189–196.

Gavaler, J. S., Bonham-Leyba, M., Castro, C. A. and Harman, S. (1999) The Oklahoma postmenopausal women's health study: recruitment and characteristics of American Indian, Asian, Black, Hispanic and Caucasian women. Alcoholism: Clinical and Experimental Research 23, 220–223.[ISI][Medline]

Gavaler, J. S., Galvao-Teles, A., Deal, S. R. and Monteiro, E. (2002) Mechanisms for the increase in postmenopausal estradiol levels among moderate drinkers: findings in a multi-racial, multi-ethnic study population. Endocrinology, Metabolism and Nutrition [Lisbon] 11, 141–148.

Gavaler, J. S. (2002) Oral hormone replacement therapy: factors which influence the estradiol levels achieved in a multi-racial study population. Journal of Clinical Pharmacology 42, 137–144.[Abstract/Free Full Text]

Gavaler, J. S. (2003) Thoughts on individualizing estrogen replacement therapy based on the Postmenopausal Health Disparities Study. Journal of Women's Health 12, 757–768.[CrossRef][ISI]

Gavaler, J. S. and Rosenblum, E. R. (2003) Predictors of postmenopausal body mass index and waist hip ratio in the Oklahoma Postmenopausal Health Disparities Study. Journal of the American College of Nutrition 22, 269–276.[Abstract/Free Full Text]

Gavaler, J. S., Deal, S. R. and Rosenblum, E. R. (2004) Directions for unraveling the issue of health disparities: findings from the Postmenopausal Health Disparities Study. Alcohol and Alcoholism 32, 69–75.

Gavaler, J. S. (2004) Cohort trends in the incidence and age at natural menopause and bilateral oophorectomy among racial and ethnic groups: findings from the Postmenopausal Health Disparities Study. Endocrinology, Metabolism and Nutrition [Lisbon] 13, 1–9.

Ginsburg, E. S., Walsh B. W., Shea, B. F., Gao, X., Gleason, R. E., Barbieri, R. L. and Teoh, S. K. (1996) Effects of alcohol ingestion on estrogens in postmenopausal women. Journal of the American Medical Association 276, 1747–1751.[Abstract]

Gomberg, E. S. L. (1994) Risk factors for drinking over a woman's life span. Alcohol Health and Research 18, 220–227.

Hemsell, D. L., Grodin, J. M., Brenner, P. F., Siiteri, P. K. and MacDonald, P. C. (1974) Plasma precursors of estrogen. II. Correlation of the extent of conversion of plasma androstenedione to estrone with age. Journal of Clinical Endocrinology and Metabolism 38, 476–479.[ISI][Medline]

Hersh, A. L., Stefanick, M. L. and Stafford, R. S. (2004) National use of postmenopausal hormone therapy. Journal of the American Medical Association 291, 47–53.[Abstract/Free Full Text]

Kontoravdis, A., Kalogirou, D., Antoniou, G., Kontoravdis, N., Karakitsos, P. and Zourlas, P. A. (1996) Prophylactic oophorectomy in ovarian cancer prevention. International Journal of Gynecology and Obstetrics 54, 257–262.[CrossRef]

Longcope, C., Jaffee, W. and Griffing, G. (1980) Metabolic clearance rates of androgens and oestrogens in aging women. Maturitas 2, 283–290.[CrossRef][ISI][Medline]

McCoy, N. L. (2002) Longitudinal study of menopause and sexuality. Acta Obstetrics and Gynecology Scandinavia 81, 617–624.[CrossRef]

Meldrum, D. R., Davidson, B. J., Tataryn, L. V. and Judd, H. L. (1981) Changes in circulating steroids with aging in postmenopausal women. Obstetrics and Gynecology 57, 624–628.[Abstract]

Register, T. C., Cline, M. J. and Shively, C. A. (2002) Health issues in postmenopausal women who drink. Alcohol Research and Health 26, 299–307.[Medline]

Rosenblum, E. R., Stauber, R. E., Van Thiel, D. H. and Gavaler, J. S. (1993) Assessment of the estrogenic activity of phytoestrogens isolated from bourbon and beer. Alcohol: Clinical and Experimental Research 17, 1207–1209.[ISI][Medline]

Schutt, D. H. and Cox, R. I. (1972) Steroid and phytoestrogen binding to sheep uterine receptors in vitro. Journal of Endocrinology 52, 299–305.[ISI][Medline]

Turner, C. (1990) How much alcohol is in a standard drink? An analysis of 125 studies. British Journal of Addiction 85, 1171–1175.[ISI][Medline]

U.S. Department of Agriculture (USDA) (1995) Nutrition and Your Health: Guidelines for Americans, 4th ed. Home and Garden Bulletin No. 232, U.S. Department of Health and Human Services, Washington DC.





This Article
Abstract
Full Text (PDF)
All Versions of this Article:
40/5/469    most recent
agh173v1
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Request Permissions
Google Scholar
Articles by GAVALER, J. S.
PubMed
PubMed Citation
Articles by GAVALER, J. S.