By Dr. Geoffrey Modest
A recent Swedish study found that socioeconomic status (SES) was independently associated with mortality, cardiovascular disease, and cancer in patients with type 2 diabetes (see doi:10.1001/jamainternmed.2016.2940).
Details:
- 217,364 people <70 yo, with type 2 diabetes in the Sweden National Diabetes Register (from 2001-2011), assessing all-cause, cardiovascular, diabetes-related and cancer mortality.
- Median age 58, 60% male
- 10% from non-Western countries: of these, 10% Latin America/Caribbean, 17% East or South Asia, 60% Middle East/North Africa, 14% sub-Saharan Africa
- Results (all adjusted for age, sex, duration of diabetes, marital status, income level, educational level, country of birth. further adjustment for smoking, HbA1c, eGFR, BMI, diabetes treatment, albuminuria, heart failure, MI, stroke, stage 5 CKD, and baseline cancer did not affect the associations much):
- 19,105 all-cause deaths: 60% cardiovascular, 37% diabetes-related, 34% cancer-related
- Marital status: comparing married vs single, overall 13.0 deaths /1000 vs 18.82 deaths/1000
- 27% decreased all-cause mortality [HR 0.73 (0.70-0.77)]
- 33% decreased cardiovasc mortality [HR 0.67 (0.63-0.71)]
- 38% decreased diabetes-related mortality [HR 0.62 (0.57-0.67)]
- No difference in cancer-related mortality, other than 33% decreased risk for prostate cancer [HR 0.67 (0.50-0.90)]
- Income: comparing lowest to highest income quintiles, overall 8.92 deaths /1000 vs 18.33 deaths/1000. This risk varied continuously as income level changes
- 71% increased all-cause mortality [HR1.71 (1.60-1.83)]
- 87% increased cardiovasc mortality [HR 1.87 (1.72-2.05)]
- 80% increased diabetes-related mortality [HR 1.80 (1.61-2.01)]
- 28% increased cancer-related mortality [HR 1.28 (1.14-1.44)]
- Income: comparing non-Western immigrants to native Swedes (with covariate adjustment)
- 45% decreased all-cause mortality [HR 0.55 (0.48-0.63)]
- 54% decreased cardiovasc mortality [HR 0.46 (0.38-0.56)]
- 62% decreased diabetes-related mortality [HR 0.38 (0.29-0.49)]
- 28% decreased cancer-related mortality [HR 0.72 (58-88)]
- Education: comparing college/university degree vs 9 yrs or less education. this risk varied continuously as educational level increased
- 15% decreased all-cause mortality [HR 0.85 (0.80-0.90)]
- 16% decreased cardiovasc mortality [HR 0.84 (0.78-0.91)]
- 16% decreased cancer-related mortality [HR 0.84 (0.76-0.93)]
Commentary:
- This study complements many prior studies from many countries over the past many decades showing that SES is a powerful predictor of morbidity and mortality, including both the development of and mortality from diabetes
- Sweden provides not just a large and rigorous database for analysis (clinical as well as individual-level data on risk factors and socioeconomic variables), but a country where SES has minimal effect on access to and use of health care services. For example, being hospitalized in Sweden costs approximately $10/day independent of the level of care or the number/type of interventions done. Immigrants in general receive evidence-based treatments earlier than native Swedes!!!
- Clearly there are limits to drawing definitive conclusions from an observational study. For example, some important covariates were not measured (e.g., alcohol intake, amount of smoking). Perhaps others as well.
- So, an important aspect of this study in Sweden was that there appeared to be relatively equal access to medical care in all groups. This again reinforces the concept that health care is much more than medical care, and health outcomes really depend a lot on broader social issues (and, this is in a country with a much larger social safety net and more extensive social programs/support networks than the United States, for example). In terms of the potential mechanism leading to increased mortality in those with lower SES, one might posit that their attendant stress is associated with many potentially adverse hormonal changes (especially as mediated by the known stress-related increase in cortisol), coupled with perhaps less healthy eating habits, lack of social supports/community cohesion: all possibly leading to increased morbidity and mortality.
Some other SES blogs:
https://stg-blogs.bmj.com/bmjebmspotlight/2016/04/28/primary-care-corner-with-geoffrey-modest-md-bmi-height-and-socioeconomic-status/ which found (through Mendelian randomization) that SES was related both to the individual’s height and BMI, and that part of the association was mediated through genetics, but mostly through social factors.
https://stg-blogs.bmj.com/bmjebmspotlight/2016/04/26/primary-care-corner-with-geoffrey-modest-md-life-expectancy-and-income/ found a striking relationship in the US between income and life expectancy.
And, for those readers who are microbiome-oriented, https://stg-blogs.bmj.com/bmjebmspotlight/2015/08/11/primary-care-corner-with-geoffrey-modest-md-early-life-stress-in-mice-changes-in-microbiome-and-later-anxiety/ which showed that early life stresses in mice leads to long-lasting adverse changes in microbiota