Why is diabetes on the rise

The critical literature review identified several factors that could potentially explain such a change in the prevalence of diabetes at a population level. The factors have been grouped in five categories Table 2 which are defined below. A description of these factors from to in New Brunswick is also presented to better understand how they could have contributed to the rise in prevalence of diabetes in this region.

In Table 2 , factors for which we found information and which we included in the analyses are presented in italics.

Concurrently, we noted a marked increase in the prevalence of prediabetes from to Table 1. Increases were also noted in the prevalence of obesity, hypertension, alcohol consumption, and immigration over the same period of time. During the study period, we also found the population was aging and that the prevalence of consumption of fruits and vegetables decreased.

However, the proportion of people reporting a physically active lifestyle, sedentary behaviours, tobacco smoking, having completed high school and university education, and being of low income evolved in directions that were opposite to the direction expected to be considered as a factor contributing to the increasing prevalence of diabetes. The literature review has shown that some environmental risk factors such as the presence of environmental pollutants such as nitrogen dioxide NO 2 , particulate matter PM , organic persistent pollutants and non-persistent pesticides , urbanization and rapid socioeconomic development could explain part of the increase in the prevalence of diabetes.

In this study, the only factor that was possible to measure was the urbanization based on the proportion of people living in a rural area. This proportion decreased during the study period and evolved in a direction supporting that it may have influenced the increase in prevalence of diabetes in New Brunswick Table 1.

Some factors characterizing the evolution of the disease, including survival time among individuals with diabetes, number of new cases of diabetes and conversion to diabetes from prediabetes population, could explain the increase in the prevalence of diabetes.

In this study, we found that only two of these factors evolved in the direction expected if these factors were to explain the increase in prevalence of diabetes. The incidence rate increased from to and a higher increase in the incidence rate was seen around and Table 1. The mortality rate of the population with diabetes decreased in the same period. In contrast, the conversion rate from prediabetes to diabetes could not explain the increase in prevalence of diabetes because it decreased during this period.

The literature review also suggested that a change in prevalence of disease could be attributed to changes in how the condition is identified. The detection effect could be related to an increase in the number of people being screened or diagnosed, an earlier detection of the disease and changes in diagnostic criteria. Accordingly we found that the number of people tested for HbA 1c increased from to and showed a higher increase around and Fig. Concurrently, the mean HbA 1c at detection decreased between and suggesting that people are being diagnosed earlier in the evolution of the disease Table 1.

It was also noted that the percentage of individuals with diabetes previously identified with prediabetes increased from to , also suggesting that people are being detected at an earlier stage of their disease. The average age at detection also supports the presence of a detection effect since data suggest that people are being diagnosed at a younger age.

Changes in factors could also be more pronounced in segments of the population, based on the year of birth birth cohort effect. In this study, the presence of a period effect was supported by an increase in the prevalence of type 2 diabetes from and in each age group, with the 40—49 year old age group displaying a greater increase Fig.

The presence of a birth cohort effect was also suggested since the increase in prevalence of type 2 diabetes was considerably greater among the youngest birth cohort compared to the oldest birth cohorts Fig. The prevalence of type 2 diabetes has more than doubled in the past 15 years in New Brunswick. Although this prevalence change is relatively greater than changes noted elsewhere, this is consistent with the increase in prevalence in diabetes observed in recent years in other provinces and countries around the world [ 69 , 72 ].

To help identify factors responsible for this increase, our literature review led to the identification of five categories of factors which together represent a comprehensive overview of factors that could explain a change in prevalence of diabetes. Our review included considerably more potential factors than what had been reported in other literature reviews [ 5 — 7 , 35 , 57 , 73 ]. Guided by this inventory, we assessed the changes in nearly all factors suggested to potentially influence the prevalence of diabetes in New Brunswick.

Through this work, we identified that changes in prevalence of diabetes in New Brunswick are likely attributed to a combination of numerous factors. Among individual-level risk factors identified in the critical review, our analysis suggests that the presence of other conditions, such as the aging population, obesity, hypertension and prediabetes could contribute to explain the increase in prevalence of type 2 diabetes in New Brunswick. Furthermore, although immigration increased in NB, it is difficult to conclude that ethnicity might have an effect on the prevalence of type 2 diabetes in NB since no information is available on ethnic origin of immigrants in the databases used.

An increase in prevalence of obesity or body mass index has been linked to increases in the prevalence of diabetes in many other studies [ 8 , 37 — 39 , 68 ] and authors suggested it is the most important contributor to increases in prevalence of diabetes [ 8 — 10 ].

Consistent with changes in body composition of the population, we also noted that diet quality decreased over the study period.

However, our data suggest that New Brunswickers became more physically active in the past 15 years, which does not concord with trends for obesity and diabetes. It has been suggested that this discrepancy may be attributable to measurement error associated with the use of self-reported measures [ 74 ].

Self-reported measures, such as physical activity levels, are suspected to have been influenced by an increase in social desirability possibly creating higher estimates over the years [ 74 ]. However, our results are consistent with the apparent increase in prevalence of a physically active lifestyle in the United States from and , which seems to have had minimal impact on reducing the burden of obesity [ 75 ]. Among the potential for environmental risk factors to have influenced diabetes, we only had access to data relating to urbanization and these suggested that the transition from a predominantly rural to an urban population accompanied the increase in prevalence of diabetes in New Brunswick.

This transition may mean that a greater proportion of the population may be exposed to obesogenic environments such as more sedentary work, higher use of car and public transport, television viewing and in short, lower levels of physical activity [ 6 ].

Although environmental pollution can be present in rural regions, it is also possible that urbanization has led to more exposure to pollutants. A meta-analysis has shown that the prolonged exposition to nitrogen dioxide NO 2 , particulate matter less than 2.

Changes in how diabetes evolves may also contribute to explaining a change in how the prevalence of disease increased in New Brunswick.

Most particularly, we noted a decline in mortality rates of people with diabetes from to , which is consistent with results from other studies [ 18 , 22 , 23 , 26 — 28 , 31 ].

A longer survival period in the population with diabetes, likely explained by better treatment and control of the disease [ 48 ], therefore contributed to the increase in prevalence. An increase in the conversion rate from prediabetes to diabetes could also have explained an increase in diabetes prevalence [ 65 ]. However, this study shows a decrease in the conversion rate from prediabetes to diabetes, therefore suggesting either better prevention efforts in this population at risk or a relatively higher proportion of people identified with prediabetes through more screening.

As our data indicated, it is highly probable that part of the increase in prevalence of diabetes can be attributed to a detection effect whereby the number of people tested with HbA 1c rose markedly during the study period.

Although the increase in testing was observed every year, sharper increases appeared to coincide with milestones such as the publication of the and guidelines encouraging type 2 diabetes screening [ 76 , 77 ] and the identification of HbA 1c as a recognised diagnostic tool by the American Diabetes Association in [ 78 ].

The possibility of earlier detection of the disease is also supported by a decrease in HbA 1c levels and age at detection during the study period. The increase in testing and an early detection may also partly be explained by the study period corresponding with the introduction of financial incentives for physicians to offer recommended care to their patients with diabetes , the implementation of an HbA1c tracking tool created to improve adherence to guidelines [ 79 ] and the Physician Practice profiles implemented in to aid physicians to identify at risk patients in New Brunswick.

The fact that an increase in the prevalence of type 2 diabetes was observed in all age groups over the 15 years of study supports the presence of a period effect. The percentage of change in the prevalence of diabetes was nevertheless higher among the 40—49 years old, as it is possible that this group benefited from a higher detection effect than other age groups. The period effect could be attributed to a combination of any of the other factors identified in this study, including urbanization, increases in immigration and an increase in the detection.

Other factors not measured, such as rapid socioeconomic development and increase in environmental pollution, could also be at play [ 19 ].

The data also presented evidence of a birth cohort effect as we observed higher increases in the prevalence of type 2 diabetes in the younger cohort groups. This could be explained by exposure early in life to some environmental factors related to an increase in obesity, as suggested by Soon et al. These results are consistent with others who demonstrated that steeper increases in the prevalence of diabetes in the youngest cohort parallel increases in the prevalence of obesity in the younger generations [ 19 , 20 ].

The development of more obesogenic environment may affect younger people to a greater extent than other age groups since it represents a great proportion of their relative environmental exposure [ 19 ]. Further, another study reported that weight gain between 25 and 40 years old was associated with a higher risk of diabetes than a weight gain after 40 years old [ 80 ]. The present study has some limitations that need to be acknowledged.

First, because of the descriptive design of the study, it was not possible to quantify and contrast the relative contribution of each factor. Second, the prevalence of diabetes and prediabetes could be underestimated due to the use of only one diagnostic method HbA 1c test in determining presence of disease and because HbA 1c was only endorsed as a diagnostic method after However, because the HbA 1c test had been in significant use prior to for the management of the disease [ 81 ], we hypothesise that most of the time a diagnosis of diabetes was followed by a HbA 1c test a short period after.

In the same way, the detection effect measured in part by the number of people tested with a HbA 1c could be overestimated by the fact that the HbA 1c values for prediabetes or diabetes screening had not been identified before and we did not consider other screening tests such as fasting glucose and oral glucose tolerance test which were often used before However, HbA1c test has different sensitivity, specificity and utility than other diagnostic tests such as OGTT and fasting glucose [ 35 ] and because of that, it is advisable to use the same diagnostic test over time.

Also, even though the HbA1c test is not universally accepted as a diagnostic tool [ 82 ] and may be affected by some individual conditions [ 83 ], it represents less individual variability than other tests and provides a better reflection of the glucose homeostasis in the long term [ 84 ]. Third, because it was not possible to distinguish the type of diabetes in the diabetes registry, it was not possible to exclude gestational diabetes cases and type 1 diabetes was excluded only on the basis of age.

Fourth, it was not possible to measure some risk factors because of lack of data available, including gestational diabetes, intra uterin environnment, nutritional transition status, family diabetes history, environmental pollution, rapid socioeconomic development and high triglycerides.

Fifth, our description of risk factors was performed concurrently to the change in prevalence of diabetes we were attempting to explain. Since diabetes has a long latency period, it may be appropriate to start describing those factors before This study also has strengths that need to be highlighted.

This is the first study that describes an overview of factors that could explain the increase in prevalence of type 2 diabetes and used a population based analysis to measure almost all of those factors. The majority of analysis was done on the entire New Brunswick population and almost all the data was available for the 15 year period. In conclusion, this study presents a comprehensive overview of potential factors that could explain the change in the prevalence of type 2 diabetes.

This study shows that for the past 15 years, the prevalence of diabetes in New Brunswick has increased considerably and this increase could be explained by many factors including some individual-level and environmental risk factors, the detection effect, the evolution of the disease and global changes. However, the increasing prevalence of diabetes observed may not be as impressive as it appears due to the significant influence of the detection effect and the dramatic decrease in mortality.

A better understanding of factors potentially responsible for the increase in type 2 diabetes can assist in making informed decisions about diabetes programs and policies. This study may be used as a template for other countries or provinces to identify factors that could explain the increase in the prevalence of diabetes in their respective jurisdictions. More research is needed to measure the relative contribution of each factor on the increase in prevalence of diabetes by measuring each factor directly at the individual level and to evaluate the change in risk factors earlier in the evolution of the disease.

International diabetes federation. Diabetes Atlas 6th edition. Accessed 15 May World Health Organisation. Public Health Agency Diabetes in Canada: facts and figures from a public health perspective. Canadian diabetes association. The cost of diabetes in New Brunswick. Diabetes prevalence and determinants in adults in China mainland from to a systematic review.

Diabetes Res Clin Pract. Article PubMed Google Scholar. Obesity and diabetes in Pacific Islanders: the current burden and the need for urgent action. Curr Diab Rep. Bhattarai MD. Three patterns of rising type 2 diabetes prevalence in the world: need to widen the concept of prevention in individuals into control in the community. J Nepal Med Assoc. CAS Google Scholar. Twelve-year trends in the prevalence and risk factors of diabetes and prediabetes in Turkish adults.

Eur J Epidemiol. Ethn Health. Smith JP. Nature and causes of trends in male diabetes prevalence, undiagnosed diabetes, and the socioeconomic status health gradient.

Ann Intern Med. Shifts in population dietary patterns and physical inactivity as determinants of global trends in the prevalence of diabetes: an ecological analysis. Nutr Metab Cardiovasc Dis. Forecasting Tunisian type 2 diabetes prevalence to validation of a simple model. BMC Public Health. Prevalence and determinants of overweight, obesity, and type 2 diabetes mellitus in adults in Malaysia.

Asia Pac J Public Heal. Google Scholar. Commentary: trends in prevalence of type 2 diabetes and prediabetes in South Asians—does it tell a story? Int J Epidemiol.

Prevalence, incidence and mortality of diagnosed diabetes: evidence from an Italian population-based study. Diabet Med. Trends in diabetes prevalence, incidence, and mortality in Ontario, Canada a population-based study. Increasing prevalence of type 2 diabetes in a Scottish population: effect of increasing incidence or decreasing mortality.

Recent Blog Articles. Health news headlines can be deceiving. Why is topical vitamin C important for skin health? Preventing preeclampsia may be as simple as taking an aspirin. Caring for an aging parent? Tips for enjoying holiday meals. A conversation about reducing the harms of social media.

Menopause and memory: Know the facts. How to get your child to put away toys. Is a common pain reliever safe during pregnancy? Harvard Health Blog Why are diabetes-related complications on the rise? Print This Page Click to Print. Art Savard. You might also be interested in…. Living Well with Diabetes Living Well with Diabetes helps you better understand and manage your diabetes.

Featured Content Recognizing the symptoms Monitoring blood sugar Weight-loss strategies for diabetes Alternative treatments for diabetes. Staying Healthy. Free Healthbeat Signup Get the latest in health news delivered to your inbox! Sign Up. Close Thanks for visiting. The Best Diets for Cognitive Fitness , is yours absolutely FREE when you sign up to receive Health Alerts from Harvard Medical School Sign up to get tips for living a healthy lifestyle, with ways to fight inflammation and improve cognitive health , plus the latest advances in preventative medicine, diet and exercise , pain relief, blood pressure and cholesterol management, and more.

I want to get healthier. Close Health Alerts from Harvard Medical School Get helpful tips and guidance for everything from fighting inflammation to finding the best diets for weight loss Close Stay on top of latest health news from Harvard Medical School.

Sign me up. This study aims to explore the latest trend in global and regional-specific diabetic burden by type, year, socioeconomic status and its associated risk factors, thus help to achieve the goal of prevention and control of NCDs in Data were collected from a set of possible sources, which include 21 possible Global Health Data Exchange data types ranging from scientific literature to survey data to epidemiological surveillance data The GBD study provides detailed epidemiologic estimates of more than diseases and injuries in countries and territories from to The overall GBD methodologies and specific diabetes methodology have been described 11 , 13 , Secondary analyses were performed by year, age, regions and socioeconomic status.

Ethics approval and informed consent were not required for this study because of public accessibility to the data. Gross national income GNI , a measure of the total domestic and foreign output, was calculated using the World Bank Atlas method.

Countries were divided into 4 categories according to GNI in The SDI ranges from 0 to 1, with a higher value implying a higher level of socioeconomic development. The calculations for the attributable burden of a given risk-outcome pair was explained in GBD 13 , attributable DALYs or mortality were estimated as total DALYs or mortality for the outcome multiplied by the population attributable fraction PAF. The PAF represents the proportion that the outcome would be reduced in a given population and time if there was exposure to the counterfactual level of the theoretical minimum risk exposure level As there were interactions between risk factors, all risk factors associated with DALYs may not be equal to the sum of each one.

Age-standardized rates of incidence, prevalence, death, DALYs were expressed as number per , population. Comparisons of national age-standardized rates of incidence, prevalence, death and DALYs among five SDI-based countries groups were assessed using the Kruskal—Wallis test, followed by Dunn's multiple comparisons. All statistical analyses were conducted using Prism software version 8; GraphPad. A P -value less than 0. The global disease burden of diabetes increased greatly from to Fig.

Globally, the incidence of diabetes increased from The age-standardized incidence rate increased from The global prevalence of diabetes increased from The age-standardized prevalence rate increased from 4, Global burden of diabetes mellitus from to DALYs: disability-adjusted life-years.

Global deaths due to diabetes increased from 0. The age-standardized death rate increased from Global DALYs increased form Age-standardized DALY rates increased from The global trend of type 2 diabetes, which accounted for majority of diabetes, was similar with that of total diabetes Fig. From to , age-standardized rates of type 2 diabetes increased from In terms of type 1 diabetes from to Fig.

The age-standardized rate decreased from 5. It can be observed that the diabetic burden increased gradually from to and are predicted with a rise from to in terms of incidence, prevalence, death and DALYs Fig. More specifically, there was a forecast increased to The geographic distribution of diabetic burden in varied by countries Fig.

The five highest prevalence were observed in China The top five countries of deaths were India , , China , , Indonesia 97, , United States 68, , and Mexico 64, Detail distribution information of age-standardized prevalence, mortality and DALYs rates of diabetes were displayed in Figure S1.

Global map of health burden of diabetes mellitus in Maps was based on EChart which is an open-source visualization library under an Apache 2. Based on the World Bank Income Level divisions, the incidence and prevalence of type 2 diabetes has increased greatly in all regions since Fig. With exception of the high-income regions, the remaining regions had an increase in the age-standardized mortality of type 2 diabetes especially in the lower-middle-income regions Fig.

The age-standardized DALYs rate of type 2 diabetes exhibited an increasing trend in overall income level regions, noticeably in the lower-middle-income regions Fig. With respect to type 1 diabetes, the incidence and prevalence slightly increased in recent years in high-income regions and remained stable in the remaining regions Fig. The age-standardized mortality and DALYs rates of type 1 diabetes trended downward in all regions Fig. For total diabetes and type 2 diabetes, the association between incidence, prevalence, mortality or DALYs rates with SDI index displayed an inverse U-shaped curve with the higher rates occurring in low-middle, middle, and high-middle SDI countries.

Globally, the burden of diabetes is prominently associated with metabolic risks i. In , the leading three risk factors were high BMI, dietary risks and ambient particulate matter pollution. High BMI was responsible for In low and low-middle SDI regions, diet low in fruits was a major risk factor.

Moreover, low, low-middle, and middle SDI regions were places where there was greater risk of deaths and DALYs from household air pollution from solid fuels. Interestingly, alcohol use appeared to be a protective effect in high-SDI regions.

SDI: Socio-demographic Index. This study presented a comprehensive picture of the numbers, rates, and increased trends of the burden of diabetes in countries and territories over the past 28 years. In , the global prevalence number and DALYs number of diabetes reached There was a projected increased to Similarly, the global trend of type 2 diabetes also increased.

However, the age-standardized rate of mortality and DALYs decreased steadily for type 1 diabetes. Interestingly, the low-middle, middle, and high-middle SDI regions were associated with higher burden, while high SDI regions had lower burden.

In addition, modifiable metabolic, environmental, and behavioral factors appeared to be most relevant risk factors for the diabetes burden. These estimates of GBD demonstrated the large and inexorably increasing burden of diabetes in the world. The estimated global prevalence number of diabetes had almost doubled since The estimated number was similar to data from IDF, which reported million people aged 18—99 years living with diabetes 4. The prevalence number for subjects aged 1—19 years was also included in GBD data, which increased from 5.

In addition, this study also showed that the increasing trend of global burden varied by diabetic type and regions. The incidence and prevalence number of type 2 diabetes increased worldwide, with higher incidence and prevalence in low-middle, middle, and high-middle SDI countries.

The increasing trend of type 1 diabetes has principally occurred in high income regions including Europe and the United States, in which there has been a reported 2. This indicates that people in low-middle, middle, and high-middle SDI countries could more be prone to type 2 diabetes because of social and economic transformation with increased food supply, a westernized diet and reduced physical activity.

Effective interventions should be conducted to change unhealthy lifestyles. On the other hand, for patients with type 2 diabetes, some methods including short duration intensive insulin therapy and reduction of body weight by intensive lifestyle interventions could be adopted to reverse diabetes 21 , This study demonstrated that the estimated death number due to diabetes has increased in recent decades and reached 1.

However, the mortality rate of type 2 diabetes in high-income areas and the global mortality rate of type 1 diabetes showed a downward trend. This estimated death number is similar to the data reported by the World Health Organization 1. The disparity in global death number between GBD and IDF most likely results from the data resources and methodology they used. GBD data come from death certificates listing diabetes as the condition most likely to be associated with direct mortality.

In contrast, the mortality reported by IDF is calculated from relative risks and total numbers of deaths from cohort studies comparing death rates in those with and without diabetes including both direct and indirect mortality 4. It is well known that diabetes is associated with increasing incidence and death due to cardiovascular and cerebrovascular disease, cancer, infectious disease, which results in increased indirect mortality associated with diabetes Our estimates showed that there was more than 67 million DALYs of diabetes in Age-standardized DALYs rates of type 2 diabetes exhibited an upward trend while type 1 diabetes showed a downward trend from to Several previous studies have reported that diabetic DALYs showed an increasing trend in developing countries.

In contrast, there has been a decrease in DALYs in some developed countries. Prevention and comprehensive control of diabetes including blood glucose, blood pressure and lipid profile should be emphasized, as these have all been shown to greatly decrease all-cause death and diabetic complications Favorable changes occurred in the age-standardized mortality and DALYs rates in type 1 diabetes over the past 28 years, which is consistent with other studies.

There is convincing evidence that individuals with type 1 diabetes have continuously improved since s, with respect to living with their condition 30 , This could be closely related to the progress in diabetes education, continuous monitoring of blood sugar, the widespread use of insulin and insulin analogues, and sensor-augmented pump therapy In contrast, there are still some improvements that could be made when caring for those with type 1 diabetes, especially for low income and low-middle income countries, which presented lower prevalence but higher rates of mortality and DALYs.

There is also a tremendous gap in life expectancy between patients with type 1 diabetes and that of the general population, even in high income countries. Recent studies demonstrated type 1 diabetes resulted in a loss of The situation is even worse for low income regions and low SDI regions where insulin is not readily available Strategies to balance distribution of medical and health resources in the world and inner a country should be formulated.

The higher increase in incidence and prevalence rates occurred in the lower-income regions as well as mortality and DALYs rate over the decades. This raises concerns for the less developed countries. These highlight that type 1 diabetes is still severe due to limited access to essential medicines particularly life-saving insulin and technologies Multiple factors are involved in type 2 diabetes including biological, environmental, behavioral and social factors which make it more complicated to measure the association between the DALYs and SDI As low-middle, middle, and high-middle SDI countries are experiencing rapid socioeconomic progress, the dietary pattern and lifestyles change greatly.

However, the basic infrastructures are insufficient to support healthy lifestyles and current healthcare services are unable to detect diabetic disease early and interfere timely. Intensive measures should be planned and implemented in the less developed countries to prevent further increases in diabetes DALYs and to reinforce healthcare services.

Our results showed that most of the diabetic burden is attributable to modifiable risk factors. As is well reported, high BMI accounts greatly for diabetes and has been continuously rising The incidence of diabetes could be reduced by weight loss 37 , Globally, diet low in whole grains, nuts and seeds, and fruits were the leading risks among dietary risks, which were much more pronounced in lower SDI countries.

Most developing countries are switching from traditional diets to higher intake of carbohydrates, fats and sugars Globalization and emerging supermarkets increase access to processed, high-fat, added-sugar and salt-laden foods. Relative low price and high accessibility of energy-dense but low-nutrient food decrease the consumption of whole grains, fruits and vegetables High BMI is greatly affected by dietary and physical activities which suggests policies could be targeted on improvements in healthy diet and adequate exercises.

In developing countries, government can provide more incentives for purchasing whole grain, nuts and seeds, fruits and vegetables and put restrictions or disincentives for less healthy products. The age-standardized DALYs rate attributable to ambient particulate matter pollution experienced a Though global household air pollution from solid fuels decreased, it was still a major risk factor in low and low-middle SDI countries.

Ambient and household air pollution may alter lung function, vascular homeostasis and insulin sensitivity, resulting in abnormalities in glucose homeostasis