Introduction

The developing world is experiencing an increase in the longevity of its population and an increase in the prevalence chronic, non-communicable illnesses (Smith and Mensah 2003). It has been estimated that over the last two decades 77% of the world’s new gain in elderly populations occurs in the developing countries. This increase in the age of the population has profound implications for the health care both in the present as well as in the near future.

The assessment of the health of a population is an accepted first step in establishing priorities for policy and action. In the global burden of disease studies conducted by World Health Organization, measures of morbidity and mortality by cause were derived from a combination of survey and outpatient morbidity data. Information gaps were filled with data derived from epidemiological models in the absence of full information (WHO 2002a, b).

Recent studies on the global burden of disease have shown that women bear a disproportionate share of the total burden of morbidity despite their generally longer life expectancy than men (WHO 2008; Murray and Lopez 1996). Women, who experience occupational risks in addition to the perils of reproduction in the fast-growing urban economies of low-income countries, are at exceptional risk of poor health (Murray and Lopez 1998; Avotri and Walters 1999). In Africa, most of the recent rapid population growth has been concentrated in towns and cities (UN Pop Div 2002). In many instances, health services have not kept pace with this population growth (UNICEF 1999). It is in the cities that lifestyle changes have become the most pronounced. These include changing diets, work that is more sedentary and new habits such as smoking and drinking, all of which have initiated a health transition, introducing new illness patterns to urban populations (Amoah et al. 2002).

The Women’s Health Study of Accra was conducted in 2003 in Accra, Ghana. Accra was selected as the site for this study because of the high level of migration from the rural areas to the urban environment. This was a community-based assessment of the prevalence of health conditions of adult women who currently resided in the urban environment. This study included two major components, a household survey of 3,200 women and a comprehensive clinical examination of 1,328 women. The health of women aged 50 and older in comparison to the younger women who participated in the clinical component of the study is reported here.

Methods

The study population

The area chosen for this study was the administrative unit known as the Accra Metropolitan Area, the core urban area within a larger region known as Greater Accra. The metropolitan area contains 1.66 million people and 373,540 households, according to the March 2000 census of Accra (McGranahan and Songsore 1996).

To ensure that this cohort represented all women aged 18 and older in the city, an extensive program of mapping and listing of eligible households was performed prior to the household survey (HHS) interviews. The 2000 Population and Housing Census provided a full enumeration of the population of Accra on census night, 26 March 2000. In the Accra metropolitan area, there were 1,731 occupied Enumeration Areas (EA) and each contained an average of 959 people. To increase the efficiency of the sample, the EAs were stratified by socio-economic status using the master sample (Megill 2002). Household facility indicators and education were used to stratify all EAs in Accra into four categories. Four socioeconomic status (SES) categories contained roughly equal numbers of people which were derived by dividing the SES index into quartiles. To keep the design effect small, the “take” of women was restricted to 18 women per EA. Once the 200 sample EAs had been selected with probability proportional to population size, eligible women who were usually resident in Accra aged 18 and older were listed by name and address. Over 60,000 names and addresses were collected. From this frame, study women were selected with probabilities fixed according to the SES status of the EA and the age group of the women. For the purposes of this study, the definition of the older woman was arbitrarily defined as age 50 and greater. Older women were progressively over-sampled as per previously described (Hill et al. 2007). From the original cohort of women completing the household survey, 1,321 women were selected in order of survey completion and then by over-sampling the women age 50 and older to participate in the comprehensive medical clinics (Duda et al. 2007). The additional seven women were neighbors that came to the clinic and were then included in the medical examination.

Women age 18 and older who were current residents of Accra, Ghana were eligible. Women were asked their age, including birth month and year. For women who were uncertain of their age, important events were presented to them as a reference and an estimate made for the most accurate age possible. The 39 pregnant women in this study were excluded from the obesity-linked analyses.

The clinical assessment—Comprehensive Medical and Laboratory Examination

The Comprehensive Medical and Laboratory Examination (CMLE) was performed by physicians at Korle Bu Teaching Hospital of the University of Ghana. Confidential HIV testing was provided with voluntary pre-test and post-test counseling by trained HIV health workers. Informed consent was obtained separately for the CMLE and HIV testing.

The history included select questions from the HHS for comparative purposes, as well as a review of the past medical history (PMH) and a review of systems (ROS) for specific and non-specific symptoms that occurred over the preceding four weeks. Questions regarding the family history, medication history and drug allergies were included.

A complete physical examination was performed, including HEENT, cardiovascular, pulmonary, gastrointestinal, breast, gynecologic, neurologic and musculoskeletal surveys. Anthropometric measurements were obtained on the women wearing lightweight street clothes without shoes. Weight was measured on a calibrated Salter scale to the nearest 0.1 kilogram (kg). Height was measured to the nearest 0.5 cm with the women standing upright with the head in the Frankfurt position (Norton and Olds 1996). Body mass index (BMI) was calculated as weight (kg) divided by height (meter2). Waist circumference was measured in centimeters (cm) at the mid-point between the lower ribs and the iliac crest. Hip measurements were taken at the maximal circumference of the buttocks. The waist to hip ratio (WHR) was calculated from the waist and hip measurements.

The laboratory evaluation included a complete blood count (CBC), fasting blood glucose (FBG), fasting lipid panel, hemoglobin genotyping, HIV testing and Papanicolaou (Pap) smear.

The CBC was performed using the impedance method on the Cell Dyne 1700 Series Analyzer (Abbott Laboratories, Abbott Park, IL.) The FBG was analyzed using the enzymatic assay kit ATAC PAK Glucose Reagents and ATAC 8000 Random Access Chemistry Analyzer (Elan Diagnostics, Dublin, Ireland). The hemoglobin electrophoresis testing was performed using cellulose acetate strips (Biosystems S.A., Barcelona, Spain) on the Fisons FEC 570 Power Pack Analyzer (Fisons, UK). These tests were performed at Korle Bu Teaching Hospital.

The lipid panel was performed using kits purchased from Randox Lab LTD (Antrim, Northern Ireland). HIV testing was performed using Determine HIV 1/2 Rapid Test (Abbott Laboratories, Abbott Park, Il.). Positive and indeterminate tests were repeated using the InstantScreen Rapid HIV 1/2 Assay (Morwell Diagnostics GmbH, Zurich, Switzerland). These tests were performed at the Noguchi Memorial Institute for Medical Research, University of Ghana, Accra, Ghana.

Statistical analysis

The data were coded and entered into the EPI6 database at the Institute for Statistics, Social and Economic Research (ISSER), University of Ghana. Statistical analyses were performed using SPSS 13.0 for Windows (SPSS, Inc., Chicago, IL.) including descriptive frequencies, univariate and multivariate logistic regression analysis, The multivariate logistic regression analysis used to assess the variables from the medical history included education, monthly income, perception of overall health, menopause status, years living at current residence, and marital status. Student t-Test and nonparametric analyses including Wilcoxan Rank Sum Test, Fisher’s Exact Test (2-sided) and Chi-square analysis. The Odds Ratio (OR) with a 95% confidence interval (CI) was used to describe the strength of the association. A p-value <0.05 is a statistically significant association.

Results

Characteristics of the women

A total of 1,328 women were examined in the medical clinic, 731 (55%) under age 50 and 597 (45%) age 50–100 years, the self-reported oldest age in the study. Table 1 shows comparison of age by highest level of education attained, monthly income level, residence, religion and ethnicity. There was a statistically significant difference in education level of the two groups, with the older women more likely to have had no formal schooling (p < 0.001) and to have a lower income level (OR = 0.90 [0.82, 0.98], p = 0.019). There was no statistical difference in area of residence, religion or ethnicity, although the trend for ethnicity revealed older women who were Ga and fewer who were Akan in comparison to the younger women.

Table 1 Characteristics of the women by age group
Full size table

Women age 50 and greater were significantly more likely to be married (OR = 3.30 [2.92, 3.72], p < 0.001) and to have lived in the same area or residence longer (OR = 1.33 [1.23, 1.43], p < 0.001). In a multivariate logistic regression analysis including education, monthly income, duration at residence and marital status, all variables remained significantly associated with age greater than 50 years as shown in Table 2.

Table 2 Multivariate regression analysis of demographic characteristics comparing women age 50 and greater with younger women
Full size table

Health status assessment

Women were asked questions from the standardized Short-Form 36 health assessment tool (Ware and Sherbourne 1992). Women age >50 years were statistically (p < 0.001) more likely to report that they expect their health to worsen over the next year (12.0% vs. 5.1%); report health worse than anybody they know (39.5% vs. 12.9%); disagree that their health is excellent (47.0% vs. 16.0%); agree that they seem to get sicker easier than others (28.0% vs. 11.2%) and feel that their health is worse this year than last (40.6% vs. 20.4%) in comparison to younger women.

Symptoms of menopause

The definition of menopause used for this study was no menses for 12 months and not pregnant or lactating. Women were defined as peri-menopausal if they had no menses for less than 12 months and were not pregnant or lactating and were age 50 or greater. Menopause status was known for 1,247 women, 573 (46.0%) pre-menopausal, 606(48.6%) post-menopausal, 68 (5.4%) peri-menopausal. Menopause status for women age 50 or greater was: 89.8% post-menopausal; 3.1% pre-menopausal, and 7.1% peri-menopausal. Menopause status for women under age 50 years was: 83.4% pre-menopausal; 9.8% post-menopausal and 6.8% peri-menopausal. Age 50 years and greater was significantly associated with post-menopausal and peri-menopausal status combined (p < 0.001).

Menopause status was not associated with known symptoms of menopause including: self identified weight gain as determined by change in clothes size, night sweats, hot flashes, mood swings, insomnia, irritability, sadness or depression. Only 86 post-menopausal women reported ever using hormone replacement therapy, 24(27.9%) of those were under age 50 and post-menopausal.

In a multivariate analysis that includes general characteristics and perceptions of health, age 50 years and greater remained statistically associated with post and peri-menopause status, education, monthly income, duration at current residence, perception of health, and marital status as shown in Table 3.

Table 3 Multivariate regression analysis of general characteristics and perceptions of health comparing women age 50 and greater with younger women
Full size table

Comparison of age with past medical history self-report

In the Past Medical History section of the HHS, women were asked a series of questions regarding illnesses that had occurred or were diagnosed within the previous 12 months. Age ≥50 years significantly associated with self-reported hypertension, diabetes, arthritis, stroke, glaucoma, cataracts, asthma, depression, blindness, and pancreatitis in the univariate analysis. Table 4 lists these conditions along with the odds ratio and confidence intervals. For those conditions which were uncommon in this cohort, a wide CI is noted.

Table 4 Comparison of women age ≥50 with younger women and self-reported medical history in a univariate analysis
Full size table

Controlling for variables that are significantly associated with age greater than 50 years, including education, perception of overall health, menopause status, duration of years living at current residence, marital status, and monthly income, women ≥50 years and older were more likely to report hypertension, diabetes, cataracts, arthritis, and depression than younger women as shown in Table 5. The variables stroke, glaucoma and pancreatitis had too few positive answers to be included in the multivariate analysis.

Table 5 Multivariate regression analysis of self reported medical health history comparing women age 50 and greater with younger women
Full size table

Overall, 646 (48.8%) of all women interviewed had experienced symptoms consistent with the local diagnosis of malaria. Malaria was the most frequently reported illness. Women > age 50 years were just as likely to have experienced an episode of malaria as women under age 50 years (49.6% vs. 48.1%, p=NS). Increasing age was also not associated with self-reported anemia, thyroid disease, hepatitis, myocardial infarction, peptic ulcer disease, urinary tract infections or cholelithiasis.

Comparison of age with review of systems reports

A detailed and extensive review of systems was taken that covered specific and non-specific symptoms that might have occurred during the preceding four weeks. Table 6 shows those symptoms which were significantly associated with age ≥50 years in a univariate analysis. Important findings include the high number of women with pain at any site, blurred vision, joint related problems and chronic back pain. Older women were more likely to wear glasses (p < 0.001) than the younger women. All other system variables were not associated with the older age group.

Table 6 Comparison of women age ≥50 years with younger women and the positive report of symptoms in a univariate analysis
Full size table

These variables which were found to be statistically significantly associated with age ≥50 years were analyzed with the significant variables education, menopause status, monthly income, perception of health, duration at current residence, and marital status. The review of system variables that remained significantly associated with age ≥50 years include: pain at any site, blurred vision, painful joints, stiff joints, chronic back pain, difficulty with ambulation, shortness of breath on exertion, loss of memory, chest pain on exertion, loss of teeth, loss of energy, urinary incontinence, constipation (p = 0.010), muscle cramps, confusion, and urinary urgency as shown in Table 7.

Table 7 Multivariate analysis and the comparison of women age ≥50 years with younger women and the positive report of symptoms in a univariate analysis
Full size table

Health care utilization and medication usage

Women > age 50 were statistically more likely to have ever been hospitalized than younger women (68.9% vs. 51.9%, OR 2.05 [1.63–2.57], p < 0.001). Older women were also more likely to have ever visited an out-patient clinic (94.3% vs. 88.5%, OR 1.48[1.07–2.05], p = 0.018). Older women were also more likely to have had a surgical procedure (31.5% vs. 17.6%, OR = 1.69[1.32–2.18], p < 0.001) and a previous blood transfusion (17.7% vs. 8.2%, OR = 1.63 [1.21–2.20], p = 0.001).

Overall, only 12.6% of women ever had a previous clinical breast examination, 1.3% ever had a mammogram, and only 1.9% ever had a previous Papanicolaou (Pap) smear or other cervical cancer screening procedure. There was no significant difference between the two age groups.

Women > age 50 were more likely to currently use medications prescribed by a physician (45.8% vs. 14.6%, p < 0.001), an herbalist (25.9% vs. 17.7%, p = 0.001) or a traditional healer (4.1% vs. 1.1%, p = 0.002) than younger women. Both groups were equally likely to use medications prescribed at a pharmacy (37.7% vs. 35.6%, p=NS).

Comparison of age with the physical examination results

The pertinent findings from the clinical examination include a significant association with age >50 years and elevated blood pressure, decreased visual acuity, obesity, and other clinical findings and are depicted in Table 8.

Table 8 Comparison of women age ≥50 years with younger women and the physical examination results
Full size table

Significant abnormal findings from the constitutional component included fatigue, cachexia, ill appearance and hyperkinesia. Dental abnormalities included caries and missing teeth. The abnormal speech patterns included slurred, stuttering, rapid and slow. Memory was tested by short term memory recall for five common local objects. The neck examination identified masses and abscesses. Abnormalities detected on the pulmonary examination included rales, rhonchi, inspiratory and expiratory wheezing. The cardiovascular examination identified decreased peripheral pulses suggestive of peripheral vascular disease, irregular rhythms, murmurs and carotid bruits. Lower extremity abnormalities included rubor with dependency, pitting and non-pitting edema, varicosities and thrombophlebitis.

Abnormalities of the rectal examination included masses, heme occult positive stool, prolapse, fissures, abscesses, hemorrhoids and warts. The abnormalities identified on the external pelvic examination included deformed labia, circumcision, ulcerations, vesicles, bladder prolapse, uterine prolapse, rectovaginal fistulas, rectovaginal masses, venereal warts and anal lacerations. Abnormalities of the gait assessment included difficulty getting out of a chair, wide based gait, abnormal tandem gait, and requires assistance for ambulation. While there was no significant differences between the two age groups for the examinations of the eyes, breast, abdomen or the internal pelvic examination abnormal clinical findings of those systems included cataracts, breast masses and bloody nipple discharge, and pelvic masses, bloody and infected cervical discharge and cervical lesions.

Comparison of age with laboratory results

Based on the laboratory results, women age 50 years and older were significantly more likely than younger women to have an elevated fasting blood sugar, elevated fasting serum cholesterol, and an elevated LDL as shown in Table 9. There was no association with age >50 years and results from the HDL, Pap smear, serum triglycerides, hemoglobin testing, or HIV status (Duda et al. 2005a, b).

Table 9 Comparison of women age ≥50 years with younger women and the laboratory tests
Full size table

Comparison of age and co-morbidities

The combination of three risk factors for cardiovascular disease—BMI >25 (includes overweight and obese women), hypertension and diabetes, were compared to each age group. Women age 50 and older were subdivided into two groups to determine the effects of advancing age and the prevalence of these risk factors.

Table 10 shows that 422/726 (58.1%) of women with a BMI >25 were hypertensive and 78/726 (10.7%) had an elevated FBG. There was a statistically significant increase in co-morbid illnesses with increasing age. 68.3% of women age 18–49 were overweight/obese without hypertension or diabetes compared to 12.3% of women age 50–64 and 8.6% of women 65 years and older (p < 0.001). There was a significant increase in hypertension in overweight/obese women when comparing younger women with women >50 years old (p < 0.001). When BMI >25, the risk of being hypertensive and/or diabetic is: age 18–49 OR 5.51 [4.3–7.1], p < 0.001; age 50–64 OR 3.75[2.9–4.8], p < 0.001; and >65 OR 2.49[1.9–3.3], p < 0.001.

Table 10 Comparison of co-morbid illnesses by age group
Full size table

Effect of education and income only on medical illness

There was a statistically significant difference in education level for the two age groups, as more women ≥50 years had no formal education compared to the younger women. Controlling for highest education level attained and monthly income, the most frequent health conditions reported including hypertension (OR = 12.46 [9.39–16.55], p < 0.001), BMI >30 (OR = 1.20 [1.10–1.30], p < 0.001), arthritis/joint pain (OR = 6.54 [2.78–10.10], p < 0.001), pain at any site (OR = 1.99 [1.57–2.52], p < 0.001), blurred vision (OR = 2.64 [2.04–3.41], p < 0.001), chronic back pain (OR = 2.02 [1.51–2.69], p < 0.001), numbness anywhere (OR = 1. 85 [1.39–2.45], p < 0.001), pins and needle sensations (OR = 1.65 [1.25–2.19], p < 0.001), palpitations (OR = 1,89 [1.40–2.54], p < 0.001), difficulty ambulating (OR = 7.81 [4.63–13.16], p < 0.001), shortness of breath with exertion (OR = 2.58 [1.80–3.69], p < 0.001), elevated (OR = 3.78 [2.31–6.18], p < 0.001),and hypercholesterolemia (OR = 2.23 [1.62–3.08], p < 0.001) continued to be significantly associated with age >50 years.

Relative risks for pertinent medical conditions using Cox regression analysis on measured variables

Cox regression analysis was performed on all dependent variables that were significantly associated with age ≥50 using the univariate analysis, and were analyzed by multivariate logistic regression analysis with the independent variables age as a categorical variable with 5 year increments, menopause status, highest education level achieved, monthly income, ethnicity and religion. The relative risk of hypertension, diabetes, legal blindness (Visual acuity >20/100), abnormal gait/mobility and memory are shown in Table 11. Other data are not included as no variable that was significant using the univariate analysis was insignificant using the multivariate logistic regression analysis.

Table 11 Relative risks for pertinent medical conditions using Cox regression analysis
Full size table

Discussion

The increase in longevity worldwide emphasizes the need to evaluate the status of the health of the aging population. Public health and medical care strategies are dependent on the prevalent type of conditions affecting the population.

Utilization of health services increases with age as illnesses become more prevalent (Leventhal et al. 1993). Cardiovascular disease is the leading cause of death in the developed countries and is steadily increasing in the developing world as well (AHA 2005; Roger et al. 2011; Bonow et al. 2002; WHO 2004). The prevalence of CHD morbidity has increased worldwide in part as a result of longer life expectancy and in part because of sedentary lifestyles, tobacco usage, obesity and diet. In Sub-Saharan Africa, it is estimated that prevalence of hypertension is 20 million and Women’s health in Accra a major risk factor for cerebral vascular accidents (WHO AFRO 2002). An estimated 150 million people are afflicted with type 2 diabetes globally and is expected to continue to increase (King et al. 1998; WHO 2002b). The Women’s Health and Aging Study identified chronic conditions that significantly increase with age, including visual impairment, functional disabilities, cardiopulmonary conditions, arthritis and musculoskeletal pain, and depression (Guralnik et al. 1995).

One caveat of performing such a study as the WHSA is that older individuals may underreport symptoms. The reporting of symptoms or conditions can be influenced by social, psychological, and cognitive factors (Kukull et al. 1994).

In our study, even with a difference in education level between the two groups, significant medical conditions were identified in the older women. Underreporting symptoms may also occur because of asymptomatic events, such as silent myocardial infarctions and atypical symptoms in older individuals and undiagnosed conditions because of lack of seeking health care.

The results from the WHSA identified the most common major medical problems in the older women which includes malaria, hypertension, obesity, diabetes and pain. Pain of any type, but particularly joint and back pain, is among the most common symptoms reported by the older women. In this study, the term “back” also included the neck for many respondents and the description of waist pain is consistent with back pain. This was a common report for most women, some of which may be attributable to the heavy loads women carried on their heads as young women. Malaria was as equally likely to affect the older women as the younger women. This finding emphasizes the double burden of illnesses that urban women in the developing countries must bear. Neither the highest level of education attained nor the level of monthly income affected the significant relationship between medical condition identified and age.

In this cohort of 1,328 women, only 46 women—8 of whom were 50 years or older—had no history of malaria and no history of joint pain in the last 12 months, and a normal blood pressure, BMI, normal, total cholesterol and LDL. As shown in Table 6, the risk of co-morbid diseases increases significantly with age. Risk factors for cardiovascular disease, which include hypertension, obesity, hypercholesterolemia, an elevated LDL and diabetes, are very common in these older women. Of concern is that although these conditions are significantly higher in the older women, they are highly prevalent in the women under age 50 years. Hence, targeted interventions for both younger and older women to address these health issues are critical to ensure a healthy aging population. The clinical signs of peripheral vascular disease as evidenced by diminished or absent pulses in the feet and accompanying suggestive skin changes is just under 10% in the older women, another indication of the prevalence of cardiovascular diseases in this population.

Gynecologic abnormalities are common in both older and younger women. Older women had significantly more abnormal findings related to external trauma such as female genital circumcision, prolapsed pelvic organs and fistulae, while lesions clinically suggestive of cervical cancer were identified in both age groups on the internal pelvic examination. This study did not identify any cervical cancers in the screening Pap smears, an issue that is currently being addressed. Only 1.9% of all of the women in this study reported ever having being screened for cervical cancer (Duda et al. 2005a, b; Chen et al. 2005).

Impaired visual acuity was identified in most women. For those who were illiterate, shapes were used instead of letters to facilitate an accurate measurement. While older women were more likely to wear eyeglasses than younger women, few women in the study owned a pair. Dental hygiene was also a significant problem for older women as dental caries and loss of teeth become more common with age.

To summarize, as populations age in the developing countries, health issues shift to include noncommunicable illnesses such as heart disease, hypertension, diabetes, cancers and frailty. Health care services need the support of the health ministry so that services can be expanded to provide not only reproductive health care to women but also a strong focus on general health care as the population ages. This places a major burden on the health ministry as funds should not be reallocated from existing programs but rather must be expanded to cover the changing medical and social needs of the population. In addition, other government sectors must be aware of the changing paradigm in the urban environment and address the needs for ready access to transportation, Women’s health in Accra food and housing. As migrations continue into the urban areas, the extended family network may not be as readily available as in the rural environments.

Conclusion

There are many priority health conditions for women who live in the developing countries. In addition to addressing issues related to poverty, sanitation, infectious diseases, malaria and other tropical diseases and reproductive health needs, as the population ages, chronic conditions that include hypertension, diabetes, pain, arthritis, poor visual acuity and poor dental hygiene play an important role in their overall health care management.