Type of vegetarian diet, obesity and diabetes in adult Indian population

Studies from Western countries suggest that vegetarian diets may have a protective role against the development of obesity and diabetes [1‐5]. The European Prospective Investigation Study (EPIC-Oxford) found that mean BMI was highest in meat-eaters, lowest in vegans, and intermediate in fish-eaters and vegetarians [6]. In the Nurses’ Health Study, intake of red meat and processed meats were associated with increased risk of diabetes [7]. In Seventh-day Adventist cohort studies initiated in the 1960s–1970s, diabetes was less prevalent in vegetarian than in semi-vegetarian (those who ate fish and poultry, but <1 time/wk)_ or non-vegetarian church-goers and processed meat eaters [5, 7‐9]. These observational findings are also supported by experimental data which have shown that the selection of foods found in vegetarian diets may carry metabolic advantages for the prevention of type 2 diabetes [10].

India is experiencing an alarming increase in the prevalence of type 2 diabetes [10‐15]. The resulting morbidity, economic costs, reduced quality of life, and risk for complications make preventive strategies imperative. The contribution of the Indian diet to the increasing prevalence of diabetes in the country is not well understood. Within this, there is little information on whether the vegetarian diet confers a similar protective effect against obesity and diabetes that have been demonstrated in western studies. This is an important question given the ongoing preponderance of vegetarianism in certain social and religious groups in India coupled with an increase in meat eating associated with growth in western-style diets in some section of the Indian society. Moreover, vegetarianism in India is associated with unique characteristics. It is usually a lifelong pattern and adherence crosses multiple generations; it generally comprises high consumption of whole grains, legumes, nuts and seeds and dairy with spices and seasonings unique to the Indian diet. Hence, the combination/or the pattern of vegetarian diet may yield different findings than similar studies conducted in the West and it is thus possible to assess dietary associations with chronic diseases which have been difficult in the West due to low frequency. This study uses data from the third National Family Health Survey (NFHS-3, 2005–06), a survey of 109,041 Indian households which collected information on a wide range of dietary, societal, lifestyle, and environmental determinants of morbidity and chronic ailments including diabetes [16]. The NFHS-3 provides a unique opportunity to examine associations between types of vegetarian diet and diabetes and obesity in a large, nationally representative sample. We hypothesized that exclusively vegetarian diets, such as vegan, lacto- or lacto-ovo vegetarian, are associated with a lower prevalence of diabetes and obesity compared with a non-vegetarian diet.

This cross-sectional, population-based study adds to the limited data on associations between type of vegetarian diet intake and diabetes prevalence in developing countries. Our finding suggest that persons consuming a lacto vegetarian, lacto-ovo vegetarian or semi-vegetarian diet had a lower likelihood of diabetes compared with those consuming non-vegetarian diet after adjustment for a number of socioeconomic and lifestyle factors. These findings may be explained by adverse effects of meat and fish, protective effects of typical constituents of lacto-vegetarian and lacto-ovo vegetarian diet which have been demonstrated elsewhere [23‐26]. Our study indicates that the body mass index of Indian vegetarian diet consumers did not differ significantly from their non-vegetarian counterparts, but the male vegetarians appeared to be significantly (p < 0.0001) uniformly leaner than female vegetarians. This association between vegetarianism and non-leanness is in line with a health study among the Barbados Seventh-Day Adventists which found self-reported vegetarians of less than 5 years did not differ significantly from the non-vegetarians [27].

Our results are in consistent with those of previous studies among various seventh-Day Adventist churchgoers [1, 2, 5, 6, 28, 29], several other studies conducted in western countries [10, 30‐37] and an Indian study [38] which showed that increased conformity to vegetarian diets is protected against risk of type 2 diabetes and hypertension. Findings from an accumulating number of studies have also shown evidence that most vegetarian diets are not only nutritionally adequate but also associated with lower risk of certain chronic diseases, when compared with the effects of a more typical western diets [36]. Evidence from a number of other observational studies shows that certain dietary constituents are associated with protection against diabetes through the pathway of insulin sensitivity which is also confirmed through food trials [10]. A reduced risk of chronic disease has been reported in populations of vegetarians living in affluent countries [39‐41] and in case–control comparisons in developing countries [42]. Reduced consumption of animal fat and increased consumption of fruit, vegetables, foods that have a low glycemic index such as beans, legumes, nuts and cereals including whole grain and foods that reduce oxidative stress and chronic inflammation [2] may underlie such a protective effect. Whole grain is also a potential contributor to reduced diabetes risk in vegetarians and accumulating evidences from various prospective studies (both in men and women) and several meta-analyses shows that consumption of whole grains may reduce risk of chronic diseases including type 2 diabetes [43‐49]. A recent meta-analysis [49] shows that the summary relative risk per 3 servings per day was 0.68 (95%CI:0.58–0.81;I² = 82%;n = 10) for whole grains and 0.95 (95%CI:0.88–1.04;I² = 53%;n =6) for refined grains. In addition to the influence of fiber and glycemic load on postprandial glucose and insulin response [50], whole grains may also reduce the risk of type 2 diabetes through the action of nutrients such as vitamin E and magnesium [50‐52].

We did not find an expected association between vegans (who eat no animal products) and significant reduced diabetes prevalence in our study contrary to other western studies. The term ‘vegan’ which may not be correctly asked/interpreted/self-reported in Indian context might be an important reason for this unexpected association, as because vegans in western countries necessarily do not have any form of dairy, including butter or ghee or any animal products, including honey which might not be the case for Indian vegans; self-reported vegans in India probably have butter/ghee/honey in their diet. Second, it may be likely that vegans may be eating more of the refined-rice diets than non-vegetarians and unlike in the West, vegans in India would not be doing it for health-conscious/political reasons. Also, since this is a cross-sectional study, we may have some reverse casuality –diabetes patients changing to a vegan diet after the diabetes diagnosis. Another issue could be the statistical power. We found the odds ratio in vegan men is very similar to that in lacto-vegetarian men, while the result in women is null. Thus, the small sample size in the vegan group, which constituted only 1.6% of the sample, might have influenced the study results. However, though the more recent official statements from American Diabetes Association has clearly described vegetarian diets as healthful [53] and some studies [1, 2] shows that vegans have a least risk of type 2 diabetes, still the association between vegan diet and diabetes risk is open to question.

The notion that animal protein stimulates insulin secretion and possibly insulin resistance was proposed decades ago [54]. Red and processed meat consumption has been associated with increased risk of type 2 diabetes in a large number of cohort studies in the west [6, 55‐57]. Meat intake was associated with a higher risk of diagnosed diabetes in a study in Seventh-Day Adventists [8]. Several other studies around the globe have also reported an increased risk of diabetes or type 2 diabetes with a higher intake of processed meat [7, 9, 58‐62], red meat [7, 58, 59, 62‐64] and total meat [7, 9, 65], but in some studies the results have been inconsistent [55, 61].

The categories that we have used to distinguish different types of diet in our study is internationally recognized [2, 8, 17, 18] and have also proven to be categories that have markedly different risks of common diseases such as diabetes and hypertension in cohort [1, 2, 66] and observational studies in UK [37] but it is possible that the more-refined categories may provide better comparability.

The strengths of our study include the use of large nationally representative study sample which allows comparisons to be made between men and women and the ability to examine this association in adult Indian population. Also rigorous efforts were made in the NFHS-3 to obtain reliable self-reported data: the survey used local terminology and commonly understood terms to describe the disease, rigorously trained interviewers, supervisors and standard quality checks [16] (also see http://​www.​dhsprogram.​com).

The prevalence of diabetes in this large nationally representative survey was comparatively low (1.1%) than studies conducted in selected geographical areas or cities in India [11, 14, 67‐71]. The low diabetes prevalence in our study reflects the young age of this population, the use of self-reports rather than biochemical assessments and sampling from the general population that included a high proportion of respondents in rural areas [72]. Our study has added to this literature using a national population health survey with good coverage in rural areas across India. Estimates from a recent study of rural–urban migrants showed an age-adjusted prevalence of diabetes (diagnosed using both self-reports and fasting blood glucose in relatively affluent populations) of 10–15% in urban people and 5–6% in rural people of similar age to those recruited in NFHS-3 [73]. In most urban parts of India the health system is well enough developed for diagnosis of symptomatic diabetes, but at younger ages (<30 years) diabetes may not be symptomatic and NFHS-3 prevalence estimates are undoubtedly conservative, particularly for rural India where diagnosis may be much less likely to occur [74]. However, this ascertainment bias is unlikely to have been differential with respect to types of vegetarian diet consumption. In other words, although we clearly have sub-set of disease, it’s unlikely to be systematically different from entire group in terms of dietary patterns.

Previous research has shown a good agreement for self-reported diabetes when compared with medical records in a US population [75] and that self-reported health conditions demonstrate the expected relationship with socioeconomic status in India [76]. Studies in India also have shown that the difference between self reports and objective measurements according to education and awareness levels does not preclude the use of self-reports [72, 76]. On the contrary to the prevailing view that there is a positive (or null) association between measures of socio economic status and self-reported poor health/morbidities in less-developed countries, and that any potential “under-reporting” is not only smaller than the difference in prevalence of illness between the socially disadvantaged and the advantaged, the study by Subramanian et al. [76] showed that the same is even true within groups with the same objective diagnosis. In addition, our analyses considering respondents who reported ‘unknown’ for diabetes status were nearly identical to the main analyses (data not shown). Although our sample was relatively young (<50 years for women and men both), it is representative of the young population profile of India; 84% of the Indian adult population (18–69 years) and 47% of the total Indian population at all ages fall within the ages covered by this study [77]. Our study does exclude approximately 14% of the Indian population (men and women over age 50) due to the sample design of the NFHS. The prevalence of diabetes increases with age and whether a similar socioeconomic status–diabetes relationship exists among middle and older age groups in all parts India is not clear [72], although our findings are consistent with the previous studies that have included older ages.

The current national estimate for diabetes prevalence in India is about 7% of the adult population aged 20–79 years [72]; estimates being based on three relatively recent and large scale studies using a combination of oral glucose tolerance testing and self-reports of diabetes [11, 78]. There continues to be considerable uncertainty in estimates of diabetes for the whole of India due to the limited study locations (with a focus on urban areas), wide variation in survey sampling methodology, differences in diabetes diagnostic criteria and age groups studied [72]. These differences in study design have hindered direct comparison of the prevalence between studies, across regions and over time. The NFHS-3 provides an important benchmark because it is the first nationally representative survey of diabetes in India. Even if the prevalence estimates of diabetes have been underestimated in the NFHS-3, the observed diet–diabetes associations are reasonable and significant, and can be comparable to cohort and prospective studies on similar association in the west. Previous studies have largely overlooked the importance of modifiable dietary factors, which may be a key determinant of diabetes in Indians, given the varied nature of Indian diets. Further large-scale population-based surveys can be strengthened by using simple finger-prick blood glucose measurements in addition to self-reports.

In our analyses, the cross-sectional design precludes causal inferences and we were limited to the questions used to elicit lifestyle and dietary information. Given the high proportion of undiagnosed diabetes in developing countries including India (http://​www.​worlddiabetesfou​ndation.​org) where less than half of people with diabetes are diagnosed, there is a possibility that the exposure was associated with the likelihood of testing for diabetes, which may result in detection bias. Importantly the entire study may be with known diabetic subjects who would have altered diet and hence might have increased or decreased vegetarian diet consumption due to the dietary advice based on diabetes control and complications of diabetes like nephropathy. General dietary advice given to diabetic subjects is to include more whole grains, legumes, fruits and green leafy and other vegetables as this is evident in our data where more than 90% of the self-reported diabetics did report ‘daily’ or ‘weekly’ consumption of legumes, vegetables and fruits–all suggest that the dietary choices of self-reported diabetic subjects might have been modified to manage diabetes. However, despite these shortcomings rigorous precautions were taken in the NFHS to obtain reliable self-reported data such as the survey used the local terminology and commonly understood term of the disease, rigorously trained interviewers and supervisors and standard quality checks.

Nevertheless, our study has some other limitations. Misclassification of dietary information, although unavoidable, would most likely be non-differential and thus may attenuate the true association. There were relatively small numbers in some of the dietary categories, which should be considered when interpreting the findings in relation to these diets. There might be limitation of the dietary assessment method in NFHS-3 as well since there may be other foods that are associated with diabetes that are not asked to the respondents. We were also unable to distinguish between Type 1 and 2 diabetes diagnoses. Since the NFHS-3 questionnaire is interviewer administered, information on the inter rater compatibility, reproducibility and validity of questionnaire would be critical to evaluate the ability of such questionnaire to measure true dietary intake. But NFHS-3 being a part of Demographic and Health Surveys (available at http://​www.​dhsprogram.​com) which is conducted in more than 80 countries with similar questionnaire seems to be fairly valid to get an overall picture of frequency of dietary intake in a population [79]. However, under- and over-reporting could lead to a biased estimation of the association between dietary factors and diabetes. Although we adjusted for several confounding variables, we cannot exclude the possibility of residual confounding. However, if this was the case, similar effects would be expected for dietary components that are related to greater affluence, which was not observed.

Another limitation of our study is reliance on self-reports of diabetes which has resulted in a marked underestimation of prevalence, and its focus on people <60 years in whom diabetes is less common [74]. Self-reported data, especially in rural areas, can be flawed owing to several factors such as lack of awareness, low educational status, limited access to health services and hesitation to disclose diagnosed diseases. But in developing countries, self-reporting should not be a prohibitive limitation as medically diagnosed and/or biomarker-confirmed prevalence estimates are nearly impossible for nationwide prevalence estimates in low- resource and low-access settings such as India. Also since, the low and middle income countries has only very limited nutrition and health outcomes data, NFHS-3 is therefore the best available dataset to examine the relationship.

Valid data on physical activity were not available in NFHS-3 which is a limitation of this study since persons with healthier diets may be physically more active than other persons [80]. Therefore the lack of physical activity data may have confounded the results. Moreover, assessment of sedentary habits in this study was based on hours of TV watching. However, physical activity has in part been accounted for, indirectly, by adjusting for body mass index. In the present study, adjustment for socioeconomic and demographic factors, residential location, religion and caste/tribe status of the respondents did not markedly modify the adjusted result, suggesting that the associations are not completely explained by non-dietary lifestyle factors. Further studies are needed to determine whether the association between diet and diabetes is mediated by assumed nutrients or by lifestyle and socioeconomic and demographic factors related to frequency of food consumption.

The rising burden of diabetes in India requires a rapid response that integrates policies and programmes which enable effective prevention and control across diverse geographical and low-resource settings. Our findings on inverse association between types of vegetarian diet consumption and diabetes prevalence can be considered by policy-makers to promote healthy vegetarian diet consumption in Indian population and to discourage unhealthy non-vegetarian diets. There is, therefore, an opportunity to modify the direction and dimensions of this national epidemic through policy interventions (at the state level), which promote the availability, affordability and acceptability of vegetarian diets more specifically lacto vegetarian and lacto-ovo vegetarian diets and restrain the marketing and consumption of unhealthy non-vegetarian foods. This requires coordinated action at the level of governments, civil society and responsible sections of the food industry.

In conclusion, our findings are important for public health interventions in diabetes care in India which shows that, in a large sample of adult men and women in India, variants of vegetarian diets such as lacto-vegetarian and lacto-ovo vegetarian were associated with at least a 30% lower risk of diabetes. These results add to the limited evidence in developing countries that shows potential benefits of consuming vegetarian diets to reduce the development of diabetes. These findings need further validation by longitudinal and clinical studies but may well have public health significance for the Indian population. These findings, if replicated using objective and comprehensive methods of dietary intake and diabetes, may inform the development of interventions to address the growing burden of overweight and diabetes in India.

^aThe scale has a 200 kg capacity and weighs in 0.01 kg increments. The scale is powered by six AA batteries and has an “ON-OFF” switch located at the side of the scale. The SECA 874 digital floor scale is manufactured by Seca gmbh & co. kg. Hammer Steindamm 9 – 25. 22089 Hamburg. Germany. The scale can be procured directly from Seca. These instructions were adapted from instructions that accompany the scale and revised by Irwin J. Shorr, MPH, MPS.

^bThe state of a person’s health in terms of the nutrients in his or her diet. In Indian context, it also means inadequate and poor diet and repeated exposure to disease and illness, may or may not be based on any clinical test or measurement.