Aims Metabolic syndrome is a clustering of cardiovascular risk factors. Asian Indians, particularly women, have been reported to be at higher risk of developing metabolic syndrome. We aimed to estimate the prevalence of metabolic syndrome and its selected known determinants among adult Asian Indian women of lower socioeconomic status.
Methods We conducted a cross-sectional study over a period of 1 year from January 2008 to December 2008 in South Delhi, which included 300 women (>20 years) recruited through multistage systematic random sampling. Blood pressure and anthropometric measurements were taken. Biochemical tests were performed on blood samples collected after overnight fasting. Metabolic syndrome was defined using updated National Cholesterol Education Program/Adult Treatment Panel-III (NCEP/ATP-III) guidelines with modified waist circumference for Indians and International Diabetes Federation (IDF) criteria.
Results The overall prevalence of metabolic syndrome was 29.6% (95% CI 23.8 to 36.0) and 20.4% (95% CI 15.3 to 26.1) using NCEP/ATP-III and IDF criteria, respectively. The risk of metabolic syndrome increased with age and calorie intake. Most (203 (90%)) of the study participants were involved in physical activity with a low metabolic equivalent (MET) score but one-fifth (19.5%) had a calorie intake recommended for women involved in vigorous activity.
Conclusions The high prevalence of metabolic syndrome among women of lower socioeconomic status is a cause of concern, and calls for an effective public health response.
Statistics from Altmetric.com
Metabolic syndrome is an important risk factor for cardiovascular diseases and diabetes mellitus.1 The individual risk factors for metabolic syndrome are obesity, dyslipidaemia, hypertension and impaired blood glucose. Asian Indians, particularly women, have been reported to be at a higher risk of developing metabolic syndrome. Women have an increased predisposition to gain weight that occurs at various physiological transitions such as puberty, pregnancy and menopause, and this makes them more vulnerable for development of metabolic syndrome. Additionally, many women in the region occupy traditional household roles that may mean they are often involved in sedentary activity, eg, while doing household chores. In urban areas men involved in office based desk work could also be involved in low physical activity but in this population of lower sociol economic status they are likely to get involved in work which needs moderate to vigorous activity like manual labourers, paddy workers etc. Excess weight gain and physical inactivity are two important determinants of metabolic syndrome.2 Other potential determinants of metabolic syndrome are age, diet, alcohol, smoking, ethnicity, and so on.3 Traditionally, the presence of cardiovascular diseases has been observed more frequently in affluent populations. However, recent studies have reported the presence of its risk factors among people of lower socioeconomic status.4
Studies performed in India have reported the prevalence of metabolic syndrome among adults as to be from 11% to 56%, depending on the definition used (table 1).5–13 The prevalence has also been reported to be higher among women as compared to men. The majority of these studies were performed in urban settings involving people, mostly men, of high socioeconomic status. Studies from other countries have reported that low socioeconomic status is associated with metabolic syndrome, especially in women.14 Limited information exists regarding the burden of metabolic syndrome among populations of lower socioeconomic status in India, particularly women. Therefore, we aimed to estimate the prevalence of metabolic syndrome among adult women of lower socioeconomic status residing in the urban area of South Delhi, India.
The study was carried out in the urban community of Dr Ambedkar Nagar, a resettlement colony located in South Delhi. Dr Ambedkar Nagar is about 11 km from the All India Institute of Medical Sciences (AIIMS), New Delhi, India. This colony comprises mainly people who are migrants from nearby states and who largely are of lower socioeconomic status. This area was selected for the study as it is the urban field practice area of the Centre for Community Medicine (CCM), AIIMS; the CCM provides preventive, promotive and curative health services through a Mobile Health Clinic. A team of resident doctors of community medicine, undergraduate students and interns, a public health nurse, a medical social officer, a pharmacist, a laboratory technician and other support staff provide services through this mobile clinic. This area has 19 blocks, out of which families of 6 blocks receive domiciliary care from the Centre for Community Medicine. These 6 blocks have 2868 houses, with a total population of about 20 000 as of 31 December 2007. Four multipurpose health workers (two men and two women) visit these houses and provide basic health services.
Sample size and study participants
The sample size was estimated as 246 for 20% prevalence with 5% absolute precision, with an α error of 5%.5 In order to adjust for an expected non-response rate of 20%, the revised sample size was 300. All women aged more than 20 years who were residing in the area for at least 6 months were eligible to participate and those who gave consent were included in the study. The exclusion criteria for the study were: very ill and unable to participate, pregnancy, known HIV positive status, malignancy, any known endocrinological disorder and severe end-stage organ disease.
Study design and sampling framework
We conducted a community based cross-sectional study over a period of 1 year from January 2008 to December 2008 using a multistage sampling technique. In the first stage, three blocks were selected randomly from six available blocks in the urban field practice area of AIIMS, New Delhi. All the streets in the selected 3 blocks were listed blockwise, and from these 10 streets in each block were selected randomly. After that, from each selected street, 10 houses were selected randomly. Only one eligible woman was selected from each house. If there was more than one eligible woman in the same family, then only one was selected using a KISH table.15
The Institutional ethics committee of AIIMS, New Delhi, approved the study. We obtained written informed consent for the questionnaire-based interview, physical measurements and laboratory tests. Administration of interview schedules included data collection for sociodemographic profile, total calorie intake and level of physical activity. A revisit was made if the eligible woman was absent or the house was found empty. If on revisit the selected eligible woman was again absent, or the house was again found to be empty, then the next house was visited as replacement. If no eligible woman was found in the next family then the adjoining houses were visited until an eligible woman was found. While the interview and physical examination were performed at the domiciliary level, the collection of blood samples was performed mainly through a camp approach.
Data regarding food intake in terms of total calories were recorded based on the 24 h recall method. To document any variation in dietary intake for individuals the interviews were repeated on non-consecutive days. A global physical activity questionnaire was used for collecting data to assess the level of physical activity.16 All measurements were taken according to WHO standard guidelines. Waist circumference was measured to the nearest 0.5 cm using a non-elastic measuring tape and blood pressure was recorded to the nearest 2 mm Hg using a mercury sphygmomanometer. The instruments were standardised every morning before starting measurements. After administration of the interview schedule, measurements of waist circumference and blood pressure were taken.
Blood sample collection and laboratory analysis
The collection of blood samples was performed mainly using a camp approach. If the participant expressed an inability to come to the camp due to personal reasons, then a house visit was made to collect their blood sample. In all, 5 ml of venous blood was drawn, observing universal precautions, from the antecubital vein after overnight fasting. The collected blood samples were transported within 2 h to the departmental laboratory where biochemical analyses for fasting glucose (FBG), triglycerides (TG) and high-density lipoprotein-cholesterol (HDL-C) were performed using an autoanalyser (ECO automatic analyser, firmware V.3.13) and commercially available enzymatic kits (ERBA CHEM). FBG was measured by the oxidase/peroxidase method, TG and HDL-C were measured by the glycerol phosphate oxidase peroxidase aminophenazone and phosphotungstate/MgCl2 precipitation methods. None of the collected blood samples were spoiled. Participants were provided with the results of the biochemical tests performed. Appropriate treatment and referrals were provided for those in need. To ensure quality control, two samples with known results (one normal and another with an abnormal value) were run with each batch of blood samples to be tested. The samples with known results were provided by the manufacturing company of the enzymatic kits used. The intra-assay coefficient of variation was 0.02 and the interassay coefficient of variation was 1.3.
Metabolic syndrome was defined as per NCEP/ATP-III guidelines, with modified waist circumferences for Indians.3 International Diabetes Federation (IDF) guidelines were also used to estimate the prevalence of metabolic syndrome.17 Socioeconomic status was assessed based on Kuppuswamy's scale updated for 2007.18 The calorie content of food was calculated as per Indian Council of Medical Research (ICMR) guidelines.19 Since there was little variation between the two measurements of calorie intake, calorie intake on day 1 was used for calculation of dietary intake. Moderate physical activity refers to activities that result in an increase in heart rate but do not change breathing patterns. Vigorous physical activities refer to activities that cause an increase in heart rate as well as breathing pattern. The metabolic equivalent (MET) was defined as the ratio of the work metabolic rate to the resting metabolic rate; 1 MET is defined as 1 kcal/kg/h and is equivalent to the energy cost of sitting quietly.16
The prevalence of metabolic syndrome and its individual components were expressed in percentages. Continuous variables were expressed as mean±SD or median with quartiles. The 95% CIs were calculated wherever relevant. For calculation of 95% CIs for proportions, a simple asymptotic method without continuity correction was used: p±z √pq/n, where z is the 1−α/2 point of standard normal distribution and q=1−p. Independent t tests and χ2 tests were performed to compare means and to test for any differences in proportions. For calculations of physical activity, MET values were applied.16 Multiple logistic regression analyses were performed to examine risk factors for metabolic syndrome. All analysis was two-tailed and a p value <0.05 was considered statistically significant. The statistical package SPSS for windows (V.13.0) was used for all the analyses.
Description of study sample
Of the 300 participants that were interviewed, 226 (75%) provided blood samples. Thus, the analysis was performed for 226 participants, and the rest were excluded. The number of subjects below 30 years was 52 (23%), 30 to 39 years was 61 (27%), 40–49 years was 58 (25.7%) and 55 (24.3%) were 50 years or above. The mean age of the study participants was 40.0±11.4 years. The majority (214 (95%)) of the women were Hindu by religion, followed by Muslims (6 (2.7%)). In all, 99 (43.8%) of the women were illiterate, 104 (46%) had completed primary education, 14 (6.2%) were educated up to senior secondary level and 9 (4%) of the women were graduates or above. The majority (199 (88%)) of the women were homemakers. A total of 9 (4%) of the women were in government services. The proportion of participants who were of lower, upper-lower, lower-middle and upper-middle socioeconomic status was 51 (22.6%), 135 (59.7%), 33 (14.6%) and 7 (3.1%), respectively. None of participants were of upper socioeconomic status. Table 2 shows the clinical characteristics of the study participants.
We compared the characteristics of those who did not provide blood samples (74) with those who provided blood samples (226) with respect to age, waist circumference, blood pressure, calorie intake, educational level and socioeconomic status. Those who did not give blood samples had a lower mean age (36.5±10.9 vs 40.0±11.4 years, p=0.02), lower waist circumference (71.5±9.3 vs 77.2±10.1 cm, p<0.001), similar systolic blood pressure (120.2±13.3 vs 121.0±14.7 mm Hg, p=0.68), lower diastolic blood pressure (79.5±9.1 vs 82.3±10.6 mm Hg, p=0.04) and lower calorie intake (1539.0±457.8 vs 1763.9±631.4 kcal, p=.005) as compared to those whose blood samples were available. Proportions of different educational levels and socioeconomic strata revealed p values of 0.96 and 0.27.
Prevalence of metabolic syndrome and its determinants
The prevalence of metabolic syndrome was 29.6% (95% CI 23.8 to 36.0) and 20.4% (95% CI 15.3 to 26.1) using NCEP/ATP-III and IDF criteria, respectively. Among the individual components of metabolic syndrome, the prevalence was lowest for impaired fasting blood glucose and highest for low HDL-C (table 3). The components of metabolic syndrome varied in their rates of occurrence across age groups. Of the four age groups, women aged ≥50 years had the highest prevalence of raised blood pressure (40.3%), raised fasting blood sugar (36.4%), Low HDL-C (74.5%) and metabolic syndrome (41.8%) (table 4). The prevalence of diabetes (FBG ≥126 mg/dl) was 32/226 (14.2, 95% CI 9.6 to 18.7).
Most (203 (89.8%)) of the study participants were performing physical activity with a low MET score. Participants involved in moderate and low levels of physical activity totalled 4.4% (10) and 5.8% (13), respectively. The majority of participants 182 (80.5%) were taking in a total calorie intake recommended for sedentary and moderate workers; 44 (19.5%) participants had a calorie intake recommended for women involved in vigorous activity. The mean calorie intake of study participants was 1763.9±631.4 kcal/day.
Association of socioeconomic status with metabolic syndrome was not statistically significant (p=0.76). The prevalence of metabolic syndrome in the lower, upper-lower and lower-middle classes was 33.3% (17), 28.9% (39) and 30.3% (10), respectively. Out of seven participants of upper-middle class socioeconomic status, only one had metabolic syndrome. Physical activity level was not found to be associated with metabolic syndrome (p=0.06). The prevalence of metabolic syndrome among participants involved in low, moderate and vigorous activity was 32%, 10% and 7.7%, respectively. The association of metabolic syndrome with age and calorie intake was statistically significant. The prevalence of metabolic syndrome among those taking in <1875 kcal, 1876–2224 kcal and >2225 kcal was 18.7%, 39.6% and 52.3%, respectively. (p=<0.001). Multivariate logistic regression analysis revealed that the risk of metabolic syndrome increased with age and calorie intake (table 5). The age/calorie interaction was not statistically significant p=0.86 (table 6).
This study was performed with the objective of estimating the prevalence of metabolic syndrome and its selected known determinants among adult women of low socioeconomic status, because the current prevalence data is limited for this group. The prevalence of metabolic syndrome in our study was found to be 29.6%, as per NCEP/ATP-III criteria. The prevalence of metabolic syndrome was lowest in the 20–29 years age group (7.5%), while it progressively increased with age, with the age group ≥50 years (34.3%) having the highest prevalence. Most (203 (89.8%)) of the study participants were performing physical activity with a low MET score; 44 (19.5%) participants had a calorie intake recommended for women involved in vigorous activity.
Other studies have reported the prevalence of metabolic syndrome among Indians to range from 11% to 56%.5–13 Most of these studies were community based cross-sectional studies with participants aged 20 years and above belonging to different socioeconomic status groups. Our findings cannot be compared with other Indian studies, as those studies used either different cut-offs for various components of metabolic syndrome or used a different definition. In most of the studies metabolic syndrome was diagnosed using NCEP/ATP-III criteria using the higher cut-off of waist circumference than that recommended for Asian Indians.6 Studies which had used the same cut-off for waist circumference as in our study differed in cut-off for triglyceride level (>200 mg/dl) or impaired fasting glucose (>110).12 The European Group for the Study of Insulin Resistance (EGIR) definition was used by one study.8
It is possible that the differences in study setting could have contributed to the observed difference in prevalence between our study and other community based studies, as some of the studies were conducted in rural areas,9 and one was performed in a hospital setting.10 Misra et al20 performed a study among rural women during the same period and reported an overall prevalence of 12%, which is less than half of the level seen in the current study. This highlights the important contributory role of urbanisation for the development of metabolic syndrome. The possible reason for differences from the hospital-based study could be the significant difference in socioeconomic status among study participants in the two studies. People who report to health facilities are likely to be different from an apparently healthy person from the community.
We found that the prevalence of metabolic syndrome was 20.4% when it was defined by IDF criteria (with the cut-off for FBG ie, ≥100 mg/dl), whereas it was 29.6% by NCEP criteria. This finding is usual and expected among Indians because in IDF criteria inclusion of waist circumference is mandatory. Kanjilal et al11 had reported a similar pattern as in our study where the reported prevalence was 40.3% and 34.9% by NCEP and IDF guidelines, respectively.
In our study, the most common component of metabolic syndrome was low HDL-C (71.7%), which is higher than other reported studies in India.6 High prevalence of low HDL-C in our study participants is a major cause of concern because it points towards a higher risk of future metabolic complications in female populations of lower socioeconomic status. In light of the fact that, compared to Western people, Indians are predisposed to atherosclerotic complications at a lower level of dyslipidaemia, this observation suggests that there is an urgent need to address this issue.
The results of physical activity could not be compared because they are likely to vary due to different methods of data collection and a lack of a standard protocol. Additionally, 44% of our participants were illiterate, leading to low health awareness that could impact various risk factors such as physical inactivity for the development of non-communicable diseases. The mean calorie intake of the study participants was 1763.9±631.4 kcal/day. It was comparable to the calorie intake of young adult women living in urban areas as reported in a study where it was 1745±343 kcal/day.21 The findings from our study suggest that the majority of women (203 (89.8%)) were involved in a low level of physical activity, whereas 19.5% were taking in calorie levels recommended for women involved in vigorous activity. Among women with metabolic syndrome, a higher proportion of women (34.3%) were taking in calorie levels recommended for vigorous activity.
The study has some limitations. First, the response rate was 75% for blood collection, which may contribute to non-responder bias and influence the generalisability of results. Few characteristics of study participants who had given blood samples and those who failed to do so differed. As noted earlier, the participants who attended the blood collection camp were likely to be older, with higher diastolic pressure, were more obese and had a higher calorie intake. This raises the possibility of selection bias. Due to the presence of these factors the participants could have some perception regarding the risk of development of metabolic syndrome, which may have motivated them to have the blood examination done. These factors have been known to be positively associated with metabolic syndrome. Therefore, our study is likely to overestimate the prevalence of metabolic syndrome. However, the main reason quoted by participants who refused biochemical investigation was fear of needle prick.
The study documents the current prevalence of metabolic syndrome and its components among the low socioeconomic status group. The high prevalence of components of metabolic syndrome raises the possibility of a further rise in metabolic syndrome cases in future among this group, where almost one-third of the women already had metabolic syndrome. The existing significant gender gap in the diagnosis and treatment of this disease in women should be minimised by formulating an appropriate policy to prevent and control its spread among this population. Educating people with regard to adopting a healthy lifestyle could contribute to reduction in its prevalence and development of complications. Future research is required to find out the factors contributing to the high prevalence of metabolic syndrome and its components among this population.
Current research questions
Are the risk factors for metabolic syndrome in this population the same as in other groups?
Are women from lower socioeconomic groups at higher risk than upper socioeconomic populations for developing metabolic syndrome?
What is the most important predictor of occurrence of metabolic syndrome?
About one-third of adult women of lower socioeconomic status had metabolic syndrome.
The most common abnormality found among them was high prevalence of low HDL-C.
The majority of the women in this population were physically inactive.
Contributors SS: protocol development and writing of the report, including data collection and analysis, and manuscript preparation. PM, SK, AK, BN, NV: conceptualisation, guidance and review of literature, and revision and approval of thefinal draft.
Funding Not applicable; this study was carried out as a MD thesis.
Competing interests None.
Ethics approval Institute Ethical Committe, AIIMS, New Delhi.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.