The Impact of Education,Cultural Background,and Lifestyle on Symptoms of the Menopausal Transition: The Women's Health at Midlife Study

Abstract

Aims:

This study aimed to examine differences in symptom clusters among women in midlife from different cultural origins and to identify sociodemographic, lifestyle, and health characteristics that could account for the differences between the cultural groups in symptom reporting.

Methods:

Israeli women aged 45–64 were randomly selected according to age and population strata of three groups: long-term Jewish residents (LTR), Jewish immigrants from the former Soviet Union, and Arab women (mostly Israeli-born). Interviews were conducted with 540 LTR, 151 immigrants, and 123 Arab women. The survey instrument included the occurrence and rating of how bothersome to everyday function were 16 symptoms. Three outcome variables included hot flashes and two scales for mental and somatic symptoms extracted from exploratory factor analysis.

Results:

Multivariate logistic regressions showed that immigrants and Arab women (compared to LTR) had a significantly lower risk of reporting hot flashes and mental and somatic symptoms. Menopausal status was related only to hot flashes. Low education and depression were associated with the three symptom scales, whereas nonhealthy lifestyle was related only to somatic symptoms.

Conclusions:

Our main finding is that cultural group is an independent predictor of each of the three menopausal symptom scales. A possible explanation for the lower reporting of symptoms among Arab and immigrant groups is that they differ from the LTR in level of acculturation and attitudes toward menopause. These findings support the proposition of a cultural factor in menopausal symptomatology that needs to be addressed by clinicians caring for women at midlife.

Introduction

The final cessation of the menstrual cycle, occurring for most women between age 45 and 55, marks the end of a biological transition process of decreasing reproductive hormonal secretion. In parallel to this transition, social and physiological changes, such as added burden of family roles (coping with adolescent children and aging parents) and increasing age-related chronic morbidity, further affect women's quality of life. Several major studies on women's health at midlife demonstrated that various vasomotor, somatic, and psychological symptoms are associated with these changes and are bothersome to women in the midlife years, with varying degrees of severity both within and across population groups.^1

–4 These cross-cultural studies and others emphasize both the similarity and the diversity in the manifestations of the menopausal transition across ethnic/cultural groups.^5,6 The considerable variations in symptom frequencies across regions and cultural groups lend support to the proposition that social circumstances and perceptions of the menopause and aging play an important role in the experience of the menopausal transition. Although the importance of demographic and lifestyle factors in explaining part of the ethnic differences in symptom prevalence has been demonstrated particularly in Western countries, some cultural differences were observed in the associations of socioeconomic and lifestyle characteristics with symptom reports and quality of life at midlife.^7
–9 In light of the complexity of measuring the cultural effect in the context of immigration and globalization in the medical approach to menopause, the need for further investigations and more cross-cultural comparisons with similarly collected data is stressed in the literature.^5
–7

The diversity of the population in Israel in terms of national/religious composition (Arabs/Jews) and as a result of massive immigration waves allows us to address several issues raised in cross-cultural research of the menopausal experience. The current study focuses on three main population groups: (1) native-born/long-term resident (LTR) Jewish women, a heterogeneous group of women from either Western or mid-Eastern descent, who have lived in Israel since birth/childhood or early adulthood, (2) Jewish women from the former Soviet Union who immigrated to Israel after the fall of the Communist regime, in 1989 and later, and (3) Arab Israeli women, mostly Muslim and born in Israel, who lived there for their entire lives. These three groups represent >95% of the relevant age population strata.

The LTR, though heterogeneous in terms of educational attainment and ethnic origin, were exposed for most of their lives to the local healthcare system and norms of expression of health symptoms. The two minority groups are unique and represent socially opposed groups in terms of educational background, socioeconomic status (SES), religiosity, and lifestyle factors (smoking, diet, and exercise levels). These groups also differ in the extent and nature of acculturation into the majority group. Arab women born in Israel, although familiar with the Israeli health system, live both geographically and socially secluded in Arab localities or neighborhoods without much social mixing with the majority group. The immigrant group has undergone a major life change. With time, this group is more likely than the Arab Israeli group to go through the acculturation process and adopt the language and attitudes of the majority LTR group. At the time of our study, however, immigrants at midlife were still far from being fully acculturated.

Previous studies on health disparities between the minority groups and the majority Jewish LTR in Israel have shown that the immigrants from the former USSR reported poorer health compared with LTR¹⁰ partly because of the stress of immigration.¹¹ Women immigrants were exposed to higher stress, as many of them were single providers of care for older parents and children.^12
–14 Official data on the health of Arab Israelis emphasize their lower life expectancy and higher mortality rates from cardiovascular diseases and diabetes.¹⁵ On the other hand, epidemiological studies have shown that Arab Israelis reported better self-rated health¹⁶ but a higher level of demoralization¹⁷ than Jews.

In the current analysis, we first address the issue of what are the common core symptom groupings associated with the menopausal status across the cultural groups. Findings from the Study of Women's Health Across the Nation¹ (SWAN), a multiethnic U.S. study, showed that across all racial/ethnic groups, two factors, one consisting of hot flashes and night sweats and the other of psychological and psychosomatic symptoms (tense or nervous, blue or depressed, irritable, forgetful, headaches, stiffness, and heart pounding) were associated with menopausal status.² The findings from a multicenter study of women from the United States and four European countries³ showed similar results with regard to hot flashes and night sweats but differed as to a second symptom grouping, which did not vary across the countries and was related to markers of menopausal hormonal change: poor memory; sleeping problems; aches in the neck, head, or shoulders; and difficulty in sexual arousal. In contrast, findings from the multicenter Decisions At MEnopause Study (DAMES) (including U.S, Spanish, Lebanese, and Moroccan women) showed no consistent clusters of symptoms comparable across the participating countries,⁸ although hot flashes, vasomotor symptoms, and sleeping problems were significantly related to menopausal status in models adjusted for age, education, employment, smoking, physical activity, and country of residence.⁴ The following analysis contributes data on menopausal symptom groupings across unique cultural groups of Jewish, immigrant, and Arab women in comparison to the cited findings.

The second issue we relate to is if cultural background influences the reporting of menopausal symptoms (after taking into account the effect of SES, lifestyle factors, and physical and mental morbidity). Regression analyses of the SWAN and DAMES data showed significant differences in symptom reporting across the ethnic groups in the United States² and across countries.⁴ The current study addresses the last issue by assessing the effect of cultural group on symptom reporting while controlling for the role of education, physical morbidity, emotional well-being, and lifestyle factors. The analysis also assesses the effect of educational level and lifestyle characteristics as important factors in women's quality of life at midlife.

Thus, the main objectives of the current study were (1) to examine differences in symptom reports and symptom groupings across three major population groups: native born/LTR Jewish women, Jewish women from the former USSR, and Arab Israelis, (2) to investigate the relationship of sociodemographic characteristics, health status, menopausal status, depressive symptoms, and lifestyle factors to symptom reports, and (3) to investigate to what extent these variables account for the differences among the three cultural groups in their symptom reporting.

Materials and Methods

The Women's Health in Midlife National Study (WHiMNS) was designed to sample three main subpopulations of women in the relevant age group: Jewish residents born in Israel or who immigrated to Israel up to 1989 (LTR), Jewish immigrants from the former Soviet Union who arrived after 1989 (immigrants) and Arab Israelis. These three groups represent >95% of women aged 45–64 living in Israel at the time of sampling.¹⁸

The study sample consisted of women living in Israel as of January 2004, selected randomly from the National Population Registry, stratified by age (four 5–year age groups: 45–49, 50–54, 55–59, 60–64) and population group (LTR, immigrants, and Arabs). Participants were randomly selected in each of the 12 age and group strata. The study sample included 540 LTR, 151 immigrants, and 123 Arab women. The immigrants had been living in Israel 11.7 years on average (mean age of arrival was 45) compared with 44.3 years among LTR not born in Israel (mean age of arrival was 13). All Arab participants were born in Israel and lived there for their entire lives. Additional information about sample recruitment can be found elsewhere.¹⁹

Procedure

Data collection took place from June 2004 to March 2006 and was conducted at the participants' homes by trained interviewers from the same ethnic groups as the participants, using a structured questionnaire. The questionnaire was professionally translated to Russian and Arabic with backward translation to Hebrew.

Measures

An extensive structured questionnaire was created to collect information on women's physical and mental health at ages 45–64 in Israel and to examine a wide range of health behaviors, health beliefs, and attitudes to preventive behavior. In addition, a symptom questionnaire was included in order to estimate the prevalence and severity of various menopausal and general somatic symptoms among midlife women.

The symptoms questionnaire related to the occurrence of the following 16 medical symptoms in the past 6 months: back or neck pains, headaches, stomach aches, dizziness, chronic tiredness, sleeping problems, urinary problems, dry skin, heartburn, weight loss/gain, dry eyes, hot flashes, shortness of breath, irritability/moods, chest pressure, and memory loss. Each positive response was followed by a question with regard to how bothersome the symptom was on a scale ranging from 1 (not bothersome), 2 (a little bothersome), 3 (bothersome), to 4 (extremely bothersome). The 4-point bothersome scale was derived from a study by Greene,²⁰ and the list of symptoms was derived and modified from two questionnaires: the Revised Illness Perception Questionnaire (IPQ-R)²¹ and a menopause-specific quality of life questionnaire.²² For analyzing individual symptoms, each symptom was defined dichotomously as bothersome by a rating of ≥2 in the 4-point scale. Sociodemographic characteristics included age group (45–49, 50–54, 55–59, 60–64) and continuous age, years of education (0–8, 9–12, 13+), household monthly income (12,000+, 6,000–12,000, <6,000 Israeli shekels [ILS]), work status (employed currently, housewife/unemployed, retired), marital status (married/other), and study group (LTR, Russian immigrants, Arab Israelis).

Measures of health status and lifestyle behavior included the following characteristics. Menopausal status was based on menstrual characteristics and defined as the following categories: postmenopausal, indicating menses had stopped for at least 12 months; perimenopausal, indicating menses had occurred in the last 12 months but with irregularities; premenopausal, indicating menses had occurred in the past month with no changes in expected cycle during the past year.²³ Women (aged 45–54) who reported that their menses had stopped after gynecological surgery were included in the study as a separate category (21.0%).

Ever diagnosed chronic conditions were measured by a positive response (yes or suspected) to a direct question or currently taking medication for at least one of the following five conditions: hypertension, diabetes, heart disease, stroke, and cancer (as part of a checklist of diseases and medical conditions).

Healthy lifestyle was determined from the responses to questions about smoking, physical activity in the past year, and dietary habits. Conducting a healthy lifestyle was defined as a dichotomous variable (yes/no) as follows: (1) currently not smoking and (2) engaging in any physical activity in the past year and (3) adhering to at least three healthy dietary habits from a checklist of five items (including low % of fat in cheese, low frequency of eating fried food, low frequency of eating red meat, high frequency of consuming whole wheat bread1, low frequency of eating fast food).

Body mass index (BMI) was included as a 3-point categorical variable defined as follows: BMI <25.0 (normal), BMI in the range between 25.0 and 29.9 (overweight), and BMI ≥30.0 (obese). Use of hormone replacement therapy (HRT) was assessed by the question: Do you take now or did you take in the past hormonal replacement therapy for the menopause? (Yes, No).

Depressive symptoms were measured by a short form of the CES-D depression symptoms index.²⁴ The index includes 11 of the 20 original items, each referring to feelings reported in the past week on a scale of 0 (no), 1 (1–2 times a week), 2 (3–4 times), and 3 (5–7 times). The score was computed as the mean of responses to the 11 items. Levels of depressive symptoms were defined as low, middle, and high according to tertiles of the computed scale.

Data analysis

To account for the effects of the stratification design, all analyses were done by assigning strata weights to the specific age group and population subgroup of participants. Weights were calculated by dividing the number of women in the population in each strata cell by the number of women interviewed in each such cell. Statistical analyses were performed using SAS 9.1 statistical software (SAS Institute, Inc., Cary, NC) with the relevant procedures for taking into account the sampling design and sampling weights.

In the first stage of analyzing the symptom questionnaire, factor analysis was performed for 15 symptoms (excluding hot flashes) for the entire sample. Based on the rotated factor loadings of each symptom for the entire sample, symptoms were classified into two groups of (1) mainly mental/emotional symptoms and (2) general somatic symptoms. In order to assess the components of the menopausal symptomatology that are comparable across the cultural groups (objective 1), the factor analysis results within each cultural group were reported for the symptoms included in the two groups of symptoms. The scree plot for the entire sample and for each group clearly leveled off after two factors. This also coincided with the eigenvalue criterion >1.0 for the larger LTR group. In the two other groups, there were three additional factors with eigenvalues slightly >1.0. Procedure FACTOR of SAS software was used to perform principal component analysis of the symptoms data with a varimax rotation. Rotated factor loadings of each symptom on each factor were calculated. The resulting outcome scales were dichotomized by defining the upper tertile of each scale frequency distribution as bothersome symptoms.

The adjusted contribution of each covariate to the risk for reporting bothersome symptoms was estimated with logistic regression models using SAS procedure SURVEYLOGISTIC (objective 2). Each model included the variables that were significantly or borderline significantly (p < 0.20) related to at least one of the outcome scales in the univariate analyses. In order to present comparable models across each menopausal outcome scale, no selection criteria were employed in building the multivariate models. The variables included in the final models were population group, age (continuous), education group, family status, menopausal status, healthy lifestyle, diagnosis of chronic condition, BMI, level of depressive symptoms, and HRT use. Monthly income and work status were not entered into the final models because monthly income and work status were strongly correlated with education (school years); for example, 88% of women with low education (0–8 years) were not working, and 94% of them had low income.

The predictive discrimination ability of the models was estimated using a C-statistic. Results of the logistic models are presented as odds ratios (OR) with 95% confidence intervals (CI).

Results

The comparison of population characteristics among the three study groups revealed significant differences in every sociodemographic aspect (Table 1). The weighted age distribution of the study groups shows that immigrants are older in comparison to LTR and Arab women. Immigrant women had higher education but a lower income level compared with LTR. Arab women were the least educated (79% with <9 school years), reported being in the low-income bracket (95%), and were mostly housewives or unemployed (93%). The distribution of menopausal stages did not differ significantly across the three groups. The observed lack of variability in the background characteristics among Arab women (education, household income, and work status) prevented the inclusion of all sociodemographic indicators as separate variables in further analysis; therefore, educational level was entered as a proxy for SES.

Table 1.

Descriptive Characteristics of Population by Three Cultural Groups

	Long-term residents (n = 540) ^a		Russian immigrants (n = 151) ^a		Arab Israelis (n = 123) ^a
Characteristic	n	% ^b	n	% ^b	n	% ^b	p value ^c
Age, years							<0.001
45–49	130	30.8	36	27.9	27	33.9
50–54	134	30.7	39	28.4	28	26.7
55–59	142	21.6	32	20.4	32	22.8
60–64	133	16.9	44	23.3	34	16.6
Education, years							<0.001
0–8	86	14.5	0	0.0	100	79.3
9–12	208	39.1	36	24.7	13	12.6
13–30	244	46.4	115	75.3	8	8.1
Household monthly income, ILS							<0.001
>12,000	90	19.2	4	3.4	0	0.0
6,000–12,000	198	41.4	33	25.8	5	4.9
<6,000	205	39.4	99	70.8	111	95.1
Work status							<0.001
Employed currently	331	66.0	79	55.9	6	5.6
Housewife/unemployed	147	25.3	42	27.9	115	92.8
Retired	61	8.7	30	16.2	2	1.6
Family status							<0.001
Married	420	78.7	92	61.7	89	78.0
Other	119	21.3	59	38.3	34	22.0
Menopausal status							0.36
Premenopausal	52	12.3	16	12.3	15	16.7
Perimenopausal	91	21.3	21	15.8	17	19.6
Postmenopausal	317	53.2	98	61.5	77	53.5
Surgical menopause	79	13.2	16	10.4	12	10.2
Healthy lifestyle^d	216	39.5	24	15.5	10	9.7	<0.0001
Ever use of HRT	173	29.6	19	13.0	7	5.7	<0.0001
Chronic morbidity^e	261	44.7	95	60.1	63	47.0	<0.01
BMI							<0.0001
Normal, <25	241	46.6	31	21.2	19	16.1
Overweight, 25–<30	171	32.0	65	44.4	41	34.8
Obese, ≥30	117	21.4	53	34.4	57	49.1
Level of depressive symptoms							<0.0001
Low	203	38.1	17	11.2	12	9.2
Middle	212	39.4	57	37.3	39	33.7
High	124	22.5	77	51.5	70	57.1

Because of missing values, sum of numbers not always equal to 540, 151, 123.

Percents are population estimates projected from sample data.

Wald chi-square tests were used for comparison of categorical variables, taking into account the survey design.

Currently not smoking, and physically active in past year, and report healthy dietary habits.

Hypertension, diabetes, heart disease, stroke, cancer.

BMI, body mass index; HRT, hormone replacement therapy; ILS, Israeli shekels.

Regarding health and lifestyle indicators, the findings showed that immigrants reported the highest frequency of chronic disease (60%) compared with the other two groups (Arabs, 47%; LTR, 45%, chi-square p value <0.01). Almost 40% of the LTR engaged in a healthy lifestyle compared with only 10% of the Arab women and 16% of the immigrants (p < 0.0001). Immigrants and Arabs had higher levels of obesity (BMI > 30) and depressive symptoms compared with LTR (p < 0.0001).

The frequencies of individual symptoms reported as bothersome are presented in Table 2 for the entire sample (All group) and separately for each cultural group. The six symptoms most frequently reported as bothersome were aches in back/neck (63%), headaches (47.5%), nervousness (40%), hot flashes (34%), and sleeping problems (33%). Among the symptoms that differed significantly across the study groups, Arab women reported the lowest levels of nervousness, sleeping problems, fatigue, weight gain/loss, and dry skin, whereas Russian immigrants reported the lowest level of hot flashes compared with the other two groups. Immigrant women reported the highest levels of fatigue, weight gain/loss, dry skin, and shortness of breath.

Table 2.

Occurrence of Bothersome Symptoms by Cultural Groups

	All n = 814		LTR n = 540		Russian immigrants n = 151		Arabs n = 123
	n	% ^a	n	% ^a	n	% ^a	n	% ^a	p value ^b
Aches in back/neck	523	63.3	336	61.5	107	71.1	80	64.4	0.09
Headaches	385	47.5	236	44.9	79	52.8	70	57.9	0.02
Nervousness	303	39.8	222	42.7	48	33.0	33	28.9	0.006
Hot flashes	262	34.3	195	37.5	31	20.9	36	31.4	0.0002
Sleeping problems	263	32.9	194	35.7	43	28.2	26	20.4	0.002
Fatigue/weariness	254	31.7	176	32.7	61	40.9	17	12.3	<0.0001
Memory problems	234	29.2	163	30.5	42	27.8	29	21.6	0.13
Dizziness	228	26.9	141	25.1	48	31.4	39	32.9	0.13
Eye irritation	192	22.4	126	22.0	29	19.4	37	29.3	0.18
Heartburn	176	20.9	110	19.7	37	25.0	29	23.7	0.32
Weight gain/loss	154	19.4	97	18.9	41	26.6	16	13.2	0.02
Abdominal pain	147	18.2	101	18.9	27	17.9	19	13.5	0.31
Dry skin	132	16.4	98	17.6	26	17.7	8	6.0	<0.0001
Chest pain at effort	121	14.2	66	12.2	21	14.4	34	27.8	0.003
Shortness of breath	112	12.9	59	10.8	31	20.3	22	17.0	0.01
Urination problems	70	8.0	45	7.9	9	5.5	16	12.1	0.16

Percents are population estimates projected from sample data.

Wald chi-square tests were used, taking into account the survey design.

LTR, long-term Jewish residents.

Next, we conducted a factor analysis of the symptoms. On the basis of previous reviews of the literature, emphasizing that across different cultural groups, vasomotor symptoms cluster consistently as a distinct factor,^5,25 hot flashes was analyzed as a single symptom representing the vasomotor cluster and was excluded from the factor analysis. Table 3 presents the rotated factor loadings of each of the remaining 15 symptoms on the two factors in each study group. The results show that there are differences in the way symptoms cluster across the cultural groups. Of 7 mental/emotional symptoms (factor I) that clustered together in the All group (factor loadings ≥0.40), 6 had a loading ≥0.40 in the LTR group, 5 in the immigrant group, and 4 in the Arab group (though without a consistent pattern of the same core symptoms across groups). Among immigrants, there seemed to be a clear distinction between groups of mental/emotional symptoms (factor I) and general somatic symptoms (factor II), whereas among Arab women, factor II was composed of emotional (nervousness, fatigue, chest pain) and general somatic symptoms (such as dry skin, eye irritation). Among LTR, the emotional symptoms were grouped with mental symptoms and dizziness and aches in back or neck.

Table 3.

Rotated Factor Loadings of Each Symptom on Two Factors by Cultural Groups

	All Factors ^b		LTR Factors ^b		Russian immigrants Factors ^b		Arabs Factors ^b
Symptom list ^a	I	II	I	II	I	II	I	II
Group 1^c
Shortness of breath	0.68	0.07	0.62	−0.24	0.65	0.01	0.83	0.08
Sleeping problems	0.42	0.36	0.46	0.36	(0.31)	0.19	0.60	0.34
Memory problems	0.40	0.15	(0.36)	0.13	0.64	0.13	0.40	0.28
Nervousness	0.55	0.27	0.60	0.25	0.66	−0.06	0.20	0.48
Fatigue/weariness	0.45	0.38	0.60	0.21	0.49	0.32	0.05	0.73
Chest pain at effort	0.47	0.25	0.48	0.20	0.65	0.07	0.23	0.58
Weight gain/loss	0.61	0.16	0.52	−0.13	0.17	−0.26	0.77	−0.01
Group 2^d
Eye irritation	0.08	0.56	0.06	0.62	0.02	0.67	0.40	0.50
Headaches	0.22	0.53	0.39	(0.37)	0.01	0.68	0.18	0.64
Dry skin	0.00	0.50	0.02	0.58	−0.02	0.41	0.06	0.54
Dizziness	0.38	0.46	0.50	0.35	0.22	0.45	0.36	0.54
Aches in back/neck	0.37	0.43	0.42	0.36	0.38	0.48	0.43	0.46
Abdominal pain	0.00	0.60	0.11	0.58	0.34	0.55	0.31	0.03
Excluded symptoms^e
Urination problems	0.33	0.25	0.35	0.21	0.11	0.30	0.55	0.17
Heartburn	0.20	0.39	0.36	0.25	0.21	0.39	0.39	0.45

Hot flashes as a single vasomotor symptom was not included in the analysis.

Factors I and II, rotated factor loadings for the two groups of symptoms, for the entire sample, and each cultural group.

Group 1, mental/emotional symptoms with factor loadings ≥ 0.40 for the first factor based on the entire sample.

Group 2, general somatic symptoms with factor loadings ≥ 0.40 for the second factor based on the entire sample.

Somatic symptoms with loading of <0.40 for the first and second factor based on the entire sample.

LTR, long-term Jewish residents.

In an attempt to reduce multiple comparisons of individual symptoms across study groups and considering the power limitation related to the smaller sample size for the two minority groups, we allowed the formation of two scales, based on a comparable symptom cluster from each factor. The symptom clusters included all symptoms that showed a similar pattern of loadings in each group on the factor in question (and on the other factor) even if the loading was <0.40. Thus, the first 3 symptoms in the table, shortness of breath, sleeping problems, and memory problems, were combined to represent the mental symptom scale in the current analysis. Similarly, the factor loadings for the 8 general somatic symptoms in Table 2 showed that eye irritation, headaches, and dry skin have similar trends of clustering for the three study groups (except for similar size of loadings <0.40 for headaches in the two extracted factors among the LTR). Therefore, a combined cluster of the 3 symptoms was extracted to represent a general somatic scale. These two combined scales were scored each as a mean rating of experiencing the 3 symptoms (on a scale of 0–4). Because each scale combines prevalence and severity, its score represents overall burden of symptoms of this type. The definition of clinically meaningful bothersome outcome was the upper tertile in the distribution of each scale. For individually assessed symptoms (such as hot flashes), the presence of a bothersome symptom was defined as a rating of ≥2 on the response scale.

The crude associations of the potential correlates with the three outcome symptom scales are shown in Table 4. At this univariate level, only bothersome hot flashes differed among the cultural groups. In additon, bothersome hot flashes were significantly associated with age group, education, menopausal status, and ever use of HRT; bothersome mental symptoms were significantly associated with education, income, work status, menopausal status, and chronic conditions; somatic symptoms were significantly associated with education, income, work status, family status, and healthy lifestyle. The level of depressive symptoms was significantly correlated with the occurrence of the three symptom groups.

Table 4.

Level of Occurrence of Bothersome Symptoms by Background and Health Measures

		Hot flashes ^a		Mental scale ^b		Somatic scale ^b
Characteristic	n	% ^c	p value ^d	% ^c	p value ^d	% ^c	p value ^d
Total	814	34.3		30.3		34.6
Cultural group			0.0002		0.08		0.77
Long-term residents	540	37.5		32.3		34.6
Russian immigrants	151	20.9		24.3		32.6
Arab Israelis	123	31.4		25.1		36.9
Age, years			0.0002		0.45		0.57
45–49	196	31.8		28.6		41.0
50–54	201	45.1		30.9		33.7
55–59	206	31.6		28.1		35.0
60–64	211	24.1		34.9		32.4
Education, years			0.003		0.0001		0.0003
0–8	187	43.5		42.8		48.8
9–12	258	37.5		32.4		33.6
13–30	283	28.4		23.7		29.6
Household monthly income, ILS			0.32		0.0001		<0.0001
>12,000	94	31.2		23.8		19.0
6,000–12,000	236	37.9		34.4		26.8
<6000	415	39.7		45.3		46.9
Work status			0.67		<0.0001		<0.0001
Employed currently	416	37.4		30.2		26.7
Housewife/Unemployed	304	38.1		47.1		49.0
Retired	93	32.9		39.8		30.4
Family status			0.68		0.55		0.02
Married	601	34.8		29.8		32.3
Other	212	33.1		32.1		41.8
Menopausal status			<0.0001		0.006		0.57
Premenopausal	83	12.3		25.5		32.4
Perimenopausal	129	41.8		29.9		35.0
Postmenopausal	492	34.3		28.8		33.7
Surgical menopause	107	45.8		42.5		41.0
Healthy lifestyle			0.97		0.13		0.0002
Yes	250	34.4		26.7		25.5
No	564	34.3		32.0		38.0
Ever use of HRT			0.003		0.95		0.67
Yes	199	46.6		36.2		35.6
No	612	34.0		36.5		33.9
Chronic morbidity			0.85		0.001		0.22
No	395	34.7		24.9		32.5
Yes	419	34.0		36.3		36.8
BMI			0.20		0.15		0.18
Normal, <25.0	291	36.5		26.9		30.7
Overweight, 25.0–<30	277	30.0		29.1		35.7
Obese, ≥ 30.0	227	36.8		35.1		38.6
Level of depressive symptoms			<0.0001		<0.0001		<0.0001
Low	232	25.0		18.6		20.7
Middle	308	33.7		26.9		31.6
High	271	44.8		46.5		52.6

For Hot flashes, the occurrence of bothersome symptom was defined as a rating of ≥2 on the response scale.

For Mental and Somatic scales, the definition of bothersome outcome was the upper tertile (>1.3) in the distribution of each scale.

Percents are population estimates projected from sample data.

Wald chi-square tests were used, taking into account the survey design.

BMI, body mass index; HRT, hormone replacement therapy; ILS, Israeli shekels.

The results of the multivariate logistic regression models for bothersome occurrence of symptoms in the three scales are presented in Table 5. In these analyses, cultural differences were much more apparent. Both immigrant and Arab women had a significantly lower adjusted odds of reporting all three types of symptoms—hot flashes mental and somatic symptoms—compared with LTR.

Table 5.

Multivariable Logistic Regression Analysis of Occurrence of Bothersome Symptoms

	Hot flashes			Mental scale			Somatic scale
Characteristic	OR	(95% CI)	p value	OR	(95% CI)	p value	OR	(95% CI)	p value
Cultural group			0.0009			<0.0001			0.007
Long-term residents	1.00			1.00			1.00
Russian immigrants	0.47^*	(0.28-0.80)		0.53^*	(0.32-0.88)		0.60^*	(0.37-0.98)
Arab Israelis	0.43^*	(0.23-0.78)		0.24^*	(0.13-0.47)		0.45^*	(0.25-0.82)
Age in years, continuous			0.0002			0.99			0.06^**
Per 1 year increase	0.93^*	(0.89-0.96)		1.00	(0.96-1.04)		0.96^**	(0.92-1.00)
Education, years			0.007			0.005			0.05
0–8	2.43^*	(1.39-4.26)		2.54^*	(1.42-4.54)		1.90^*	(1.09-3.31)
9–12	1.40	(0.94-2.08)		1.56^*	(1.03-2.35)		1.09	(0.74-1.61)
13–30	1.00			1.0			1.0
Family status			0.37			0.48			0.18
Married	1.00			1.00			1.00
Other	0.83	(0.56-1.24)		0.87	(0.58-1.28)		1.28	(0.89-1.86)
Menopausal status			<0.0001			0.06		0.19
Premenopausal	1.0			1.00			1.00
Perimenopausal	5.60^*	(2.45-12.8)		1.10	(0.55-2.21)		1.16	(0.60-2.25)
Postmenopausal	6.66^*	(2.77-16.0)		1.03	(0.49-2.16)		1.46	(0.75-2.86)
Surgical menopause	10.4^*	(3.90-27.9)		2.02^**	(0.88-4.64)		2.18^*	(1.01-4.71)
Healthy lifestyle			0.38			0.67			0.04
Yes	1.00			1.00			1.00
No	0.84	0.56-1.24		1.09	(0.73-1.64)		1.51^*	(1.03-2.23)
Ever use of HRT			0.04			0.70			0.88
Yes	1.52	(1.01-2.30)		1.09	(0.71-1.66)		0.97	(0.65-1.44)
No	1.0			1.0			1.0
Chronic morbidity			0.72			0.02			0.66
Yes	0.94	(0.65-1.35)		1.59^*	(1.09-2.32)		1.08	(0.76-1.55)
No	1.00			1.00			1.00
BMI			0.48			0.92			0.59
Normal, <25.0	1.0			1.0			1.0
Overweight, 25.0–<30	0.78	(0.52-1.17)		1.06	(0.70-1.61)		1.24	(0.83-1.86)
Obese, ≥30	0.90	(0.57-1.43)		1.09	(0.69-1.73)		1.11	(0.71-1.73)
Level of depressive symptoms			<0.0001			<0.0001			<0.0001
Low	1.00			1.00			1.00
Middle	1.66^*	(1.07-2.57)		1.71^*	(1.06-2.76)		1.97^*	(1.25-3.09)
High	3.45^*	(2.10-5.66)		4.48^*	(2.66-7.53)		4.22^*	(2.59-6.87)

Each model included all variables from Table 4 excluding household monthly income and work status because of their very high correlation with education.

Sample size for each final model = 794; percent of women with missing data on few independent varibles = 2.5%.

*p < 0.05; **p < 0.10.

BMI, body mass index; CI, confidence interval; HRT, hormone replacement therapy; OR, odds ratio.

Menopausal status was significantly related to hot flashes, with increasing odds for perimenopausal status (OR 5.6) and postmenopausal status (OR 6.66) and the highest odds ratio for the postoperative menopausal status (OR 10.4) compared with premenopausal women. Women at the postoperative menopausal status also had a 2-fold risk for reporting bothersome mental and somatic symptoms. Less than 9 years of school was significantly related to higher risk of bothersome symptoms in all three scales Conducting a nonhealthy lifestyle was significantly associated with reports of bothersome somatic symptoms (OR 1.51, CI 1.03-2.23), and chronic conditions were associated with the report of mental symptoms. Ever use of HRT was significantly and positively associated with bothersome hot flashes (OR 1.52, 95% CI 1.01-2.30).

Moderate and high levels of depressive symptoms were significantly associated with hot flashes (OR 1.66 and OR 3.45, respectively), with mental symptoms (OR 1.71 and OR 4.48, respectively), and with somatic symptoms (OR 1.97 and OR 4.22, respectively). C-statistics for the models of hot flashes and mental and somatic symptoms were 0.72, 0.71, and 0.70, respectively, demonstrating a good predictive discrimination ability of the models.

We repeated the analyses for each individual symptom (data not shown). The results indicated a significantly lower adjusted odds for reporting emotional symptoms (nervousness, fatigue, weight gain/loss), abdominal pain, and aches in back/neck and a borderline trend for lower reporting of dizziness (p = 0.07) among Arab women compared with LTR. For immigrants, a significantly lower trend of reporting was observed only for nervousness and abdominal pain compared with LTR.

Discussion

This article sought to compare the reports of bothersome symptom clusters across three population groups of midlife women who differ either in religious/cultural origin (Arabs vs. Jews) or in life history (immigrants vs. LTR). In the three domains of symptoms under study, hot flashes and mental and somatic symptoms, we found significantly lower odds of reporting these symptoms among immigrants and Arab women compared with LTR, after taking into account differences in sociodemographic, health status, depressive symptoms, and lifestyle characteristics.

Our first objective was to identify clusters of symptoms among midlife women and to determine if they are comparable across the cultural groups. In general, the results for the full sample showed separate factors extracted for psychological/mental and somatic symptoms, as discussed by Greene in his review of seven factor analytical studies²⁶ (although the current list did not include several psychological symptoms related to anxiety or depression, such as panic attacks, feeling unhappy, crying spells, and others). The group-specific results showed differences between Arab women and Jewish women (both immigrant and nonimmigrant) in how emotional/psychological symptoms, such as nervousness and fatigue, are perceived across cultures. Interestingly, emotional and mental symptoms were extracted to the same factor in the LTR and immigrant groups as observed in Lebanon and Morocco (DAMES data) but into different factors in the Arab Israeli group, similar to women in the United States and Spain.⁸ In addition, among Arab Israelis and Jewish immigrants, headaches clustered together with general somatic symptoms, as observed in Morocco but not in the United States, Lebanon, or Spain.⁸ Thus, Arab Israelis do not seem to resemble Arab women from other countries in the way menopausal symptoms are correlated with each other. Comparison of the current results to similar cross-cultural studies is limited and should be approached with caution in light of the differences in the symptoms list and age range included in the various cited studies.

Cultural group and symptom reporting

Our second objective was to assess the effect of cultural/ethnic differences in symptom reporting using the pooled data. Therefore, analyses were performed based on the two extracted outcome scales (fulfilling minimal criteria for internal correlations across groups) and hot flashes rather than on the group-specific factors resulting from the factor analyses conducted separately for each group.

Significant differences in symptom reporting between minority women and the majority group were observed in the models predicting bothersome hot flashes and mental and somatic symptoms. Among Arab women, the trend of lower reporting in comparison to LTR is stronger than among immigrant women (in the adjusted models) and is also significant across additional individual symptoms not included in the scales, such as psychological symptoms (nervousness, fatigue, weight gain/loss) and psychosomatic symptoms (abdominal pain, aches in back/neck, dizziness). Similar findings with regard to psychological and psychosomatic symptoms were observed in a U.S. multiethnic longitudinal study (SWAN), where Caucasian women reported significantly more psychosomatic symptoms than minority women, that is, Hispanic, Chinese, Japanese American and African American women.²

Comparison of hot flashes across the cultural groups showed lower crude prevalence rates among minority women (especially immigrants) in comparison to LTR. In general, the prevalence of reporting hot flashes (recall time, past 6 months) in the age group 45–55 for LTR in the current sample (45%) was higher than the prevalence of hot flashes/night sweats in Caucasian Americans (36.6% in a recall period of 2 weeks, aged 42–52).²⁷ For Arab Israeli women, however, the rate was (35%), lower than in Lebanese (49%)²⁸ and Moroccan women (61%)²⁹ (in two studies using 1 month recall time). Considering all limitations of comparing across studies, there are indications that Arab Israeli women report less hot flashes than Arab women in countries where they form the majority group. It is possible that Arab Israeli women were less exposed to the Western notions about mental and physical expressions related to the menopausal transition and to the process of medicalization of the menopause among Israeli women. They live mainly in rural separate settlements that are not well integrated into the mainstream Israeli society. These findings are in line with another Israeli study showing better self-rated health among Arab Israelis compared with Jewish Israelis, although their objective health was worse.¹⁶ On the other hand, the findings with regard to vasomotor symptoms are not consistent across countries with a majority of Muslim women either. Comparing across countries in the Middle East region showed that women from Morocco had significantly higher odds of reporting hot flashes than Lebanese women.⁴

The low reporting of hot flashes among Russian immigrants is best compared with such reports from Eastern European countries, and that information was not available except for a study in Slovenia³⁰ showing that among women aged 40–65, 39% reported having hot flashes in the past 2 weeks, which is higher than the 23% estimated among Russian immigrants in the current study. This last comparison suggests that the social circumstances of being an immigrant, which are associated with problems of acculturation and added family burden for women, may be more important to their quality of life than menopausal change and aging.

As discussed in a comprehensive review of the literature,³⁰ variations in vasomotor symptoms across ethnic groups may reflect both differences in biological factors, such as endogenous estrogen, climate/seasonality, and health behaviors (smoking, alcohol consumption, physical activity), and in nonbiological factors, such as socioeconomic characteristics, stress, language, and level of acculturation. Although the data do not include measures for level of acculturation, the findings indirectly support the hypothesis of a cultural factor in the way hot flashes, mental symptoms, and somatic symptoms are perceived and reported, as the group differences remained even after controlling for various individual characteristics of the women. It is possible that such factors as lower willingness to discuss hot flashes or a tendency toward natural acceptance of these symptoms as part of menopause and aging by immigrants and Arab women played a role in the observed differentials.³¹ As suggested in the relevant literature, these findings strengthen the view that factors, such as level of acculturation, norms of expressing medical problems, and awareness of what constitutes an alarming medical symptom, may play a role in symptom reporting.^3,31 Women in the majority group may have been more exposed to Western attitudes toward health issues and more aware of expected menopausal changes and, therefore, more open to report them.

Education and symptom reporting

In the current study, low educational level (i.e., ≤ 8 school years) was associated with increased adjusted ORs of reporting all three outcomes, whereas high school education (9–12 years) was related only to increased odds of reporting mental symptoms in comparison to high education (13+ years). These findings are in line with other studies in Western countries, where negative associations between educational attainment and vasomotor and various other symptoms were observed.^1,32

–35 Among Caucasians and among major ethnic groups in the United States (except for Hispanic women) less than some college education was related to frequent vasomotor symptoms,³⁴ Whereas in Britain, lower education was related to vasomotor and general somatic symptoms,³³ and in Spain, to psychological and somatic symptoms.³²

The menopausal syndrome

Our data showed that hot flashes were clearly associated with the natural menopause. In line with findings from several population and cross-cultural studies, the results indicate that hot flashes were significantly higher in perimenopausal, postmenopausal, and postoperative menopausal women (compared with premenopausal) after adjustment for ethnic group, educational level, health status, lifestyle factors, and depressive symptoms.^3,33,34,36 The significantly higher odds for hot flashes, mental symptoms, and somatic symptoms among postoperative menopausal women are in line with findings from Britain relating early surgical menopause to several mental and somatic symptoms among women after hystercotomy³³ and from the DAMES data.⁴ The explanations in the literature suggest that the postoperative stage is linked to worse physical and emotional morbidity³³ and to a special burden of symptoms over and above those resulting from hormonal change.⁴ Mental and somatic symptoms were mostly unrelated to menopausal status in this study.

Lifestyle, use of HRT, BMI, depressive symptoms, and symptom reporting

The observed association of a healthy lifestyle with lower reporting of general somatic symptoms is in line with other findings supporting the role of physical activity and healthy nutritional habits in the reduction of chronic conditions^37,38 and the promotion of well-being among women at midlife.³⁹ However, our findings differ from findings from cited studies where separate indicators of lifestyle, such as physical activity and smoking, were associated with vasomotor and other symptoms.^1,35

We found that women who ever used HRT reported significantly more bothersome hot flashes. As no information was available about the time frame and length of use of HRT in relation to the occurrence of hot flashes, it could be assumed that most women who reported ever using HRT also reported (in a later part of the questionnaire) being bothered by hot flashes. Therefore, the use of HRT can be interpreted as a marker for bothersome hot flashes in this cross-sectional study design.

It is interesting to note that obesity (BMI > 30) was not related to any of the presented symptom scales (although in separate analyses for individual symptoms, it was significantly related to heartburn, chest pain at effort, and weight gain/loss). Findings from several other studies showed that vasomotor as well as other symptoms were positively associated with BMI.^1,27,36

The significant associations of depressive symptoms with reporting bothersome symptoms in the three scales under study correspond in part with longitudinal findings from the SWAN data, where baseline depressive symptoms were significantly related to frequent vasomotor symptoms among white and Hispanic women but were positively and nonsignificantly associated with these symptoms among African Americans, Chinese, and Japanese women.³⁴ On the other hand, findings from another Israeli study⁴⁰ show a significant correlation between depressive mood and nonspecific psychosomatic symptoms but not with a specific somatic factor that included vasomotor symptoms. Our findings are also partly in line with several cross-sectional studies focusing on depressive symptoms (as an outcome) and showing in one study that climacteric symptoms but not menopausal status were associated with higher rates of depressive symptoms⁴¹ and in another study that at least one physical symptom and vasomotor symptoms were significantly related to high depressive symptoms in the full sample.⁴²

Limitations

The study has several limitations with regard to the coverage of such topics as night sweats, current use of HRT, and language as a measure of acculturation. This is the result of using data from a larger study with many aims, focused mainly on issues of health behavior and, thus, limited in its coverage of other areas. Another limitation stems from the smaller sample size of the immigrant and Arab women groups, which precluded the estimation of models separately for each study group. Nevertheless, it is the first national survey of menopausal symptoms that compares the prevalence and severity of a wide list of symptoms across major population groups.

Conclusions

The main finding of this study is that symptom reporting among Israeli women at midlife is related to cultural background, educational level, and depression. Interestingly, the cultural differences remain after accounting for depression, education, and other sociodemographic and lifestyle characteristics. Although speculative, a possible explanation is that these groups differ in level of acculturation and in attitudes toward menopause and aging. These latter variables should be further explored using both qualitative and quantitative methods. These findings have relevance to public health professionals in emphasizing the need to strengthen clinicians' sensitivity to women from diverse cultural backgrounds who may not report specific symptoms because of cultural restraint. In addition, promotion of a healthy lifestyle and reduction of stress are important targets for intervention programs as part of individual care. The significant relationship of depression with symptom reporting underscores the importance of giving clinical attention to these symptoms among women at midlife.

Footnotes

Acknowledgments

This study was funded by the Israel National Institute for Health Policy and Health Services Research, grant number 63/02.

Disclosure Statement

The authors have no conflicts of interest to report.

References

Gold

, Sternfeld

, Kelsey

et al. Relation of demographic and lifestyle factors to symptoms in a multi-racial/ethnic population of women 40–55 years of age. Am J Epidemiol, 2000; 152:463–473.

Avis

, Stellato

, Crawford

et al. Is there a menopausal syndrome? Menopausal status and symptoms across racial/ethnic groups. Soc Sci Med, 2001; 42:345–356.

Dennerstein

, Lehert

, Koochaki

, Graziottin

, Leiblum

, Leventhal Alexander

. A symptomatic approach to understanding women's health experiences: A cross-cultural comparison of women aged 20 to 70 years. Menopause, 2007; 14:688–696.

Obermeyer

, Reher

, Saliba

. Symptoms, menopause status, and country differences: A comparative analysis from DAMES. Menopause, 2007; 14:788–797.

Melby

, Lock

, Kaufert

. Culture and symptom reporting at menopause. Hum Reprod Update, 2005; 11:495–512.

Obermeyer

. Menopause across cultures: A review of the evidence. Menopause, 2000; 7:184–192.

Obermeyer

, Sievert

. Cross-cultural comparisons: Midlife, aging, and menopause. Menopause, 2007; 14:663–667.

Sievert

, Obermyer

, Saliba

. Symptom groupings at midlife: Cross-cultural variation and association with job, home, and life change. Menopause, 2007; 14:798–807.

Crawford

. The roles of biologic and nonbiologic factors in cultural differences in vasomotor symptoms measured by surveys. Menopause, 2007; 14:725–733.

10.

Baron-Epel

, Kaplan

. Self-reported health status of immigrants from the former Soviet Union in Israel. Isr Med Assoc J, 2001; 3:940–946.

11.

Brodov

, Mandelzweig

, Boyko

, Behar

. Is immigration associated with an increase in risk factors and mortality among coronary artery disease patients? A cohort study of 13,742 patients. Isr Med Assoc J, 2002; 4:326–330.

12.

Ritsner

, Ponizovsky

, Nechamkin

, Modai

. Gender differences in psychosocial risk factors for psychological distress among immigrants. Compr Psychiatry, 2001; 42:151–160.

13.

Soskolne

. Single parenthood, occupational drift and psychological distress among immigrant women from the former Soviet Union in Israel. Women Health, 2001; 33:67–84.

14.

Remennick

. “I have no time for potential troubles”: Russian immigrant women and breast cancer screening in Israel. J Immigr Health, 2003; 5:153–163.

15.

Ministry of Health. Health in Israel 2005: Selected data. Jerusalem: Ministry of Health, Department of Health Information.

16.

Baron-Epel

, Kaplan

, Haviv-Messika

, Tarabeia

, Green

, Nitzan Kaluski

. Self-reported health as a cultural health determinant in Arab and Jewish Israelis: MABAT-National Health and Nutrition Survey, 1999–2001. Soc Sci Med, 2005; 61:1256–1266.

17.

Shemesh

, Kohn

, Blumstein

, Geraisy

, Novikov

, Levav

. A community study on emotional distress among Arab and Jewish Israelis over the age of 60. Int J Geriatr Psychiatry, 2006; 21:64–76.

18.

Central Bureau of Statistics. Statistical abstract of Israel 2002 (53) Jerusalem: Central Bureau of Statistics.

19.

Benyamini

, Blumstein

, Boyko

, Lerner Geva

. Cultural and educational disparities in the use of primary and preventive health care services among midlife women in Israel. Womens Health Issues, 2008; 18:257–266.

20.

Greene

. A factor analytic study of climacteric symptoms. J Psychosom Res, 1976; 20:425–430.

21.

Moss-Morris

, Weinman

, Petrie

, Horne

, Cameron

, Buick

. The revised illness perception questionnaire (IPQ-R) Psychol Health, 2002; 17:1–16.

22.

Hilditch

, Lewis

, Peter

et al. A menopause-specific quality of life questionnaire: Development and psychometric properties. Maturitas, 1996; 24:161–175.

23.

Brambilla

, McKinlay

, Johannes

. Defining the peri-menopause for application in epidemiologic investigations. Am J Epidemiol, 1994; 140:1091–1095.

24.

Kohout

, Berkman

, Evans

, Cornoni-Huntley

. Two shorter forms of the CES-D depression symptoms index. J Aging Health, 1993; 5:179–193.

25.

Avis

, Brockwell

, Colvin

. A universal menopausal syndrome? Am J Med, 2005; 118:37s–47s.

26.

Greene

. Constructing a standard climacteric scale. Maturitas, 1998; 28:25–31.

27.

Gold

, Block

, Crawford

et al. Lifestyle and demographic factors in relation to vasomotor symptoms: Baseline results from the Study of Women's Health Across the Nation. Am J Epidemiol, 2004; 159:1189–1199.

28.

Obermeyer

, Ghorayeb

, Reynolds

. Symptom reporting around the menopause in Beirut, Lebanon. Maturitas, 1999; 33:249–258.

29.

Obermeyer

, Schulein

, Hajji

, Azelmat

. Menopause in Morocco: Symptomatology and medical management. Maturitas, 2002; 41:87–95.

30.

Sievert

, Vidovic

, Horak

, Abel

. Age and symptom experience at menopause in the Selska Valley, Slovenia. Menopause, 2004; 11:223–227.

31.

Whiteman

, Staropoli

, Benedict

, Borgeestc , Flaws

. Risk factors for hot flashes in midlife women. J Womens Health, 2003; 12:459–472.

32.

Obermyer

, Reher

, Alcala

, Price

. The menopause in Spain: Results of the DAMES (Decisions at Menopause) Study. Maturitas, 2005; 52:190–198.

33.

Kuh

, Wadsworth

, Hardy

. Women's health in midlife: The influence of the menopause, social factors and health in earlier life. Br J Obstet Gynaecol, 1997; 104:923–933.

34.

Gold

, Colvin

, Avis

et al. Longitudinal analysis of the association between vasomotor symptoms and race/ethnicity across the menopausal transition: Study of Women's Health Across the Nation. Am J Public Health, 2006; 96:1226–1235.

35.

, Samsioe

, Borgfieldt

, Lidfeldt

, Agardh

, Nerband

. Menopause related symptoms: What are the background factors? A prospective population-based cohort study of Swedish women (The Women's Health in Lund Area study) Am J Obstet Gynecol, 2003; 189:1646–1653.

36.

Dennerstein

, Dudley

, Hopper

, Guthrie

, Burger

. A prospective population-based study of menopausal symptoms. Obstet Gynecol, 2000; 96:351–358.

37.

Lee

, Russell

. Effects of physical activity on emotional well-being among older Australian women. Cross-sectional and longitudinal analyses. J Psychosom Res, 2003; 54:155–160.

38.

Patterson

, Young

, Powers

, Brown

, Byles

. Relationships between nutrition screening checklists and the health and well-being of elder Australian women. Public Health Nutr, 2002; 5:65–71.

39.

Dennerstein

. Well-being, symptoms and the menopausal transition. Maturitas, 1996; 23:147–157.

40.

Becker

, Lomranz

, Pines

et al. Psychological distress around menopause. Psychosomatics, 2001; 42:252–257

41.

Bosworth

, Bastian

, Kuchibhatla

et al. Depressive symptoms, menopausal status, and climacteric symptoms in women at midlife. Psychosom Med, 2001; 63:603–608.

42.

Bromberger

, Harlow

, Avis

, Kravitz

, Cordal

. Racial/ethnic differences in the prevalence of depressive symptoms among middle-aged women: The Study of Women's Health Across the Nation (SWAN) Am J Public Health, 2004; 94:1378–1385.