The present study showed that almost 80% and 65% of Spanish men and women, respectively, weigh more than the recommended values for their height. Indeed, 28% of men and women presented general obesity, 36% of men and 55% of women presented abdominal obesity, and 89% of men and 77% of women have a WHtR ≥0.5. Our results also showed a significant association independent of age between the obesity measures and cardiovascular risk factors (e.g., diabetes, hypertension and hypercholesterolemia). As a result, 10-year coronary disease risk significantly increased with the categories of BMI, WC and WHtR, which could indicate an important disease burden in coming years.
Comparison with previous studies
The main Spanish study on cardiovascular risk factors (DORICA study) conducted between 1990 and 2000 in individuals aged 25–60 years reported a general obesity prevalence of 13% in men and 18% in women , lower than the DARIOS results in data collected since 2000. This supports the increasing trend in the prevalence of obesity evidenced in other studies [2–5]. On the other hand, the prevalence of general and abdominal obesity was lower in the nationwide ENRICA Study conducted between 2008 and 2010 than in DARIOS. The age ranges of the populations studied (>18 years in ENRICA and 35–74 years in DARIOS) may account for these differences . However, since 2004 two studies have reported age-specific prevalence of general obesity in the Spanish elderly population [28, 29] that is similar to DARIOS results. Since age is one of the main determinants of obesity, the prevalence is likely to increase as the population ages dramatically in coming years. The high prevalence of WHtR ≥0.5 observed could be related to the threshold chosen, which has been internationally recommended . However, a higher threshold (0.55) better discriminated cardiovascular risk in a population with high prevalence of obesity .
An international study by Doak et al. showed that Romanian men and Bulgarian women aged 25–64 years presented the lowest prevalence of obesity among European countries (10%). On the contrary, Scottish men and women presented the highest prevalence (28% and 26%, respectively) , similar to the DARIOS results. Finally, analysis of National Health and Nutrition Examination Survey (NHANES) data showed that the prevalence in the US is around 32% . The authors attributed the differences to the socioeconomic context of the countries studied .
Cardiovascular risk factors, obesity and sex
Obesity is key in the development of hypertension and diabetes [32, 33]. Indeed, both diseases were associated with general and abdominal obesity in the DARIOS data, independently of age. Hypercholesterolemia also showed a significant but weaker association, even though obese individuals in our sample presented the classical lipid disorder of hypertriglyceridemia and low HDL cholesterol .
In several population-based studies and a meta-analysis, the different measures of obesity were better discriminators and had a stronger association with cardiometabolic risk factors in women [10, 18, 24, 25, 35].
The sedentary life-style could be a possible cause, which is more prevalent in Spanish women than in men . In addition, sex-related differences in fat distribution  and in eating behaviours  may play a key role. Further cohort studies are needed to ascertain sex-related differences in the use of these variables as predictors of cardiovascular events.
Cardiovascular risk and obesity
A recent study has shown improved coronary risk prediction in men if a general obesity diagnosis is included , and higher mortality has been associated with overweight, general and abdominal obesity in men . In DARIOS results, the baseline coronary risk was higher in men, although 10-year coronary disease risk was strongly associated with overweight, general and abdominal obesity in women as well. Previous studies in Spain report that obesity did not increase the incidence of cardiovascular events; however, further cohort studies with longer follow-up are needed [41, 42]. In the Framingham Heart Study, for instance, obesity was associated with increased relative risk for development of cardiovascular disease in a population aged 35–75 and followed for 44 years .
Coronary risk and obesity types
Finally, there is some controversy about the obesity measurement (i.e., general or abdominal) that better correlates to cardiovascular risk [18, 24, 25, 35, 44, 45]. The abdominal obesity measures were significant predictors of cardiovascular events and death; BMI was not .
In our results, increased WC and WHtR implied higher coronary risk independently of BMI category. Surprisingly, men with WC ≥102 cm and BMI <25 kg/m2 presented the highest 10-year coronary disease risk. This finding could be explained by the sparse number of individuals included in this category. However, the subcutaneous fat storage in patients with high BMI seems to diminish cardiovascular risk compared to individuals with higher perivisceral fat storage . Another possible explanation may be the presence of sarcopenic obesity (i.e., age-related body composition changes characterized by decreased skeletal muscle mass and increased body fat mass)  that is more associated with cardiometabolic risk  and mortality in individuals with coronary heart disease . Both explanations may show the incapacity of the subcutaneous fat storage in these individuals due to genetics, ageing, sedentary lifestyle or unknown causes that result in ectopic fat storage with higher cardiometabolic risk [50, 51]. In women, on the other hand, a risk gradient was found between BMI and both WC and WHtR. Therefore, we believe that both types of obesity should now be measured in the clinical setting.
Strengths and limitations
The DARIOS Study includes 11 studies conducted in different regions of Spain in the first decade of the 21st century. All these studies used standardized methodology. The DARIOS data is drawn from 10 Autonomous Communities that comprise approximately 70% of the total Spanish population aged 35–74 years. In addition, the sample size (>28,000 individuals) and response rate (>70% in 8 out of 11 studies) ensure that our results accurately reflect the prevalence of obesity in Spain. The response rate was estimated according to the cooperation rate in the 2011 guidelines of The American Association for Public Opinion Research .
The cross-sectional design of the study limits the causal interpretation of the associations described. Therefore, cohort studies are needed to ascertain the role of obesity in the incidence of coronary events, particularly in our society, where the prevalence of this cardiovascular risk factor has dramatically increased in recent years . Notably, the cut-off point 0.55 for WHtR has shown higher predictive value for assessing the risk of diabetes and cardiovascular events [10, 53].