María del Cristo Rodríguez Pérez

Sex hormones and autoimmunity.

Autoimmune diseases occur more in women than in men, and this may be attributable to the role of estrogens. Androgens promote autoimmune diseases with a profile of type 1 cytokines, such as rheumatoid arthritis, whereas estrogens promote autoimmune diseases with a type 2 cytokine profile, like systemic lupus erythematosus. Both androgens and estrogens regulate the Th1/Th2 balance. Type 1 autoimmune diseases are improved when decrease type 1 cytokines (i.e. during fasting), or when there is a rise in type 2 cytokines (increased estrogens, as in pregnancy). Type 2 autoimmune diseases improve when type 2 cytokines are diminished (decreased estrogen, as in post-partum period) or when type 1 response is stimulated.

Inverse Association Between Obesity and Antinuclear Antibodies in Women

To the Editor: More than 70% of all autoimmune illnesses occur in women, a figure that has been attributed to stimulation of the Th2 response by estrogens. However, a little-explored relationship is that which may exist between obesity and autoimmune disorders in women. Since the discovery of leptin1 it has been known that the cytokine-producing capacity of adipose tissue is high. Serum concentration of leptin is 3- to 4-fold higher in women than in men. However, little is known about the reasons for this difference. To date no studies have investigated the relationship between leptin and autoimmune disease in the…

Efficiency of different strategies to detect autoantibodies to extractable nuclear antigens.

Autoantibodies to extractable nuclear antigens (anti-ENA) are identified mainly in samples positive for antinuclear antibodies (ANA). Although the method of choice for ANA screening is indirect immunofluorescence (IIF), several techniques are available to detect anti-ENA. The aim of this study was to compare the efficiency of five different strategies to determine anti-ENA. During a 2-year period we screened ANA in 30375 samples with IIF, and the 4475 samples ANA positive were tested for anti-ENA by double immune diffusion screening or fluoroenzymeimmunoassay (Screening FI); anti-ENA specificities were then determined by line immunoassay (LIA) or fluoroenzymeimmunoassay (FI). We compared five strategies that involved FI or LIA identification of anti-ENA with or without prior screening, or an algorithm that combined fluorescence pattern, number of anti-ENA specificities requested by the clinician and ANA dilution titer. One cost unit (CU) was defined as the cost of 1 test of ANA determination. We detected 553 anti-ENA positive samples. The most efficient strategy was the algorithm, at a cost of 3.3 CU per sample processed, the second most efficient strategy was screening plus FI identification (cost=3.8 CU), and the third most efficient strategy was screening plus LIA identification (cost=3.9 CU). The fourth most efficient strategy was FI identification without prior screening (13.3 CU per sample) and the least efficient was LIA identification without prior screening (13.6 CU per sample). In conclusion, an algorithm that combined techniques for detection, ANA titer, fluorescence pattern and number of specificities requested was the most efficient strategy for determining anti-ENA.

Validación de un cuestionario de frecuencia de consumo de alimentos adaptado para el estudio y seguimiento de la población adulta de las Islas Canarias

Background: The registering of eating habits requires a valid and reliable method. The purpose of this study is to validate the food intake frequency questionnaire, CDC-FFQ, which is an adaptation of another questionnaire, in order to assess the nutrition of the adult population of the Canary Islands. Methods: The CDC-FFQ questionnaire was given to 1,067 individuals taken from the general population (GP) and to 106 university students, aged 19 to 30. The second group was surveyed also in three 24-hour follow-ups. The nutrients were compared according to the CDC-FFQ in the GP and university students. The correlations were estimated between the CDC-FFQ and the follow-ups for nutrients and groups of foods and the concordance of the intakes of nutrients and groups of foods in the extreme quintiles, for the university students. Results: The mean values between nutrients of the CDC-FFQ between the GP, university students and the general population with university studies showed no significant differences except for vitamin B12 (p=0.004) and vitamin D (p=0.005). Correlations between the CDC-FFQ and the mean of the three follow-ups were obtained in the 0.202-0.601 range between nutrients adjusted by calories consumed in the case of the university students. By groups of foods, the correlations ranged between 0.243-0.542 for the CDC-FFQ and the follow-ups. The concordance of nutrients ranged between 39% and 100% and for groups of foods, between 41% and 100%. Conclusions: The CDC-FFQ questionnaire is valid for classifying the subjects in the relative ranges of their level of intake of foods and nutrients and, therefore, it could be useful in epidemiological studies with a diet assessment in the adult population of the Canary Islands.

Factores asociados al conocimiento y el control de la hipertensión arterial en Canarias

INTRODUCTION AND OBJECTIVES To analyze the factors associated with knowledge and control of hypertension in the adult population of the Canary Islands (18-75 years). METHODS We recruited a random sample of the general population aged ≥18 years. Hypertension was defined as systolic/diastolic blood pressure ≥140/90 mmHg or known hypertension (self-declared, or controlled hypertension <140/90 mmHg). The bivariate association of known and controlled hypertension with age, sex, anthropometry, serum lipids, medication, and lifestyle was corroborated by adjusting a multivariate logistic model. RESULTS We included 6675 participants. The prevalence of hypertension was higher in men (40% vs 31%, P<.001), who also had a lower frequency of treated and controlled hypertension. Female sex (P<.001), age ≥55 years (P<.001), obesity (P<.001), and diabetes (P<.001) were associated with known hypertension. The modifiable factors that, in spite of treatment, increased the risk of poor control of hypertension were alcohol consumption (>30 g/day, odds ratio [OR]=2.4, P<.001; >15-≤30 g/day, OR=2, P=.009; >5-≤15, g/day, OR=1.83, P=.004), obesity (body mass index ≥30, OR=2, P=.003; >24.9-<30, OR=1.7, P=.024), serum cholesterol >250 mg/dL (OR=1.6, P=.006) and elevated heart rate (>80 bpm, OR=1.45, P=.045; >70-≤80 bpm, OR=1.36, P=.038). CONCLUSIONS The awareness of hypertension increases with frequent use of the health system and with factors associated with known hypertension: female sex, age, underlying health problems. The modifiable factors associated with poor control of known hypertension are alcohol consumption, obesity, elevated heart rate, and hypercholesterolemia.

Factors associated with parietal cell autoantibodies in the general population

The presence in serum of parietal cell autoantibodies (PCA) is a characteristic of autoimmune gastritis. We determined the prevalence of PCA in the general population and investigate their association with type 2 diabetes, insulin resistance and lifestyle factors related with autoimmune gastritis. A cross-sectional study was performed, involving 429 individuals enrolled in a cohort study of the general population of the Canary Islands. All participants underwent physical examination, provided a blood sample and responded to a questionnaire regarding health and lifestyle factors. Serum concentrations of PCA, soluble CD40 ligand (sCD40L), C-peptide and glucose (to determine insulin resistance) were measured. The association of PCA with the other factors was determined with bivariate analysis, and logistic regression models were used to adjust the associations for age and sex. The prevalence of PCA was 7.8% (95% CI=10.3-5.3). The factors associated with PCA were female sex (p=0.032), insulin resistance (p=0.016), menopause (p=0.029) and sCD40L (p=0.019). Alcohol consumption (p=0.006) and smoking (p=0.005) were associated with low prevalences of PCA. After adjustment for age and sex, the association with PCA was confirmed for smoking (OR=0.1 [0.0-0.9]), alcohol consumption (OR=0.3 [0.1-0.9]), insulin resistance (OR=2.4 [1.1-4.9]), female sex (OR=2.4 [1.1-5.3]), sCD40L (OR=3.7 [1.2-11.4]) and menopause (OR=5.3 [1.2-23.3]). In conclusion, smoking and alcohol consumption acted as protective factors against the appearance of PCA in the general population, whereas female sex, menopause, insulin resistance and elevated serum sCD40L were risk markers for PCA. In patients who smoke or drink alcohol, clinicians should be cautious when using PCA to rule out autoimmune gastritis.

Autoantibody detection with indirect immunofluorescence on HEp-2 cells: Starting serum dilutions for systemic rheumatic diseases

Antinuclear antibodies (ANA) are determined, among other reasons, to identify samples which need a second test to detect the associated specificities. Our aim was to evaluate the clinical and economic impact generated by using an initial dilution for ANA of 1:160. We analyzed all samples for which ANA, anti-ENA and anti-dsDNA were requested over a 1-year period. ANA were detected by indirect immunofluorescence. Anti-ENA were analyzed with a combination of techniques. Anti-dsDNA were detected by radioimmunoassay. Cost analysis was performed by calculating the difference between two cut-offs (ANA 1:40 and 1:160). A total of 13,233 samples were processed for ANA, of which 59.9% were positive with the 1:40 cut-off and 39.2% with the 1:160 cut-off. At ANA titer 1:40, 0.2% of the samples were anti-ENA-positive and 2.2% were anti-dsDNA positive. Only ANA dilutions of 1:160 and higher showed significantly increased positive predictive value for anti-ENA (1.5 versus 0.2, p=0.029) and anti-dsDNA (8.3 versus 2.2, p<0.001) compared to the 1:40 titer. With the 1:160 cut-off, 16.6% fewer ANA tests, 41.8% fewer anti-ENA determinations and 36.4% fewer anti-dsDNA tests would have been needed. The average saving was 0.87 cost-units per sample (1 unit=2.06euro). We conclude that setting the starting dilution for ANA at 1:160 avoids unnecessary studies, increases the positive predictive values of ANA for anti-ENA and anti-dsDNA, and generates clinical and economic benefits.

Validación del modelo REI para medir la clase social en población adulta

Fundamento: La clase social generalmente se mide de manera categorica y basada en la ocupacion laboral, lo cual tiene multiples limitaciones. El objetivo de este trabajo es elaborar un indicador cuantitativo de clase social, facilmente estandarizable, validarlo en poblacion adulta y comprobar su aptitud para medir el impacto de la clase social como determinante de salud. Metodos: estudio transversal de 6.729 individuos para medir la clase social con las variables: Renta familiar per capita, Indice de hacinamiento, Estudios realizados, Ocupacion laboral y Situacion laboral. Se crearon dos modelos y mediante curvas COR se selecciono el mejor para validarlo analizando su capacidad de estimar los riesgos relativos de: residir en barrio pobre o rico, mantener un patron dietetico tipico de clases sociales pobres y presentar problemas de salud actualmente asociados a la pobreza. Resultados: el modelo solo incluyo las variables Renta, Estudios e Indice de hacinamiento (REI), produjo un indicador con rango de valores entre 4 y 21 y mostro correlacion inversa con la edad (r= -0,28; p <0,001), con el consumo de papas (r= -0,17; p<0,001) y con el consumo de legumbres (r= -0,03; p=0,01), ademas de correlacion directa con el consumo de ensalada (r = 0,10; p<0,001); su sensibilidad para detectar la residencia en barrio pobre alcanzo el 97% para valores menores a 10. REI estimo que las clases sociales pobres presentan riesgos significativos de situacion laboral de desempleo (OR=5,4), ocupacion laboral de baja cualificacion (OR=40,9), habitar en barrios pobres (OR =30,2), bajo consumo de ensaladas (OR = 2,2), gran consumo de papas (OR = 4,5) y alto consumo de legumbres (OR = 1,6). En ambos sexos las clases pobres presentaron mayor riesgo de problemas de salud, con mayor fuerza en las mujeres: sedentarismo (OR = 1,8), obesidad (OR = 4,4), obesidad abdominal (OR = 5,4), sindrome metabolico (OR = 3,4) y diabetes mellitus (OR = 2,0). Conclusiones: REI es un indicador valido, no basado en la ocupacion ni en la situacion laboral, facilmente estandarizable, apto para medir cuantitativamente la clase social en estudios que precisen analizar el impacto de la misma como determinante de salud.

On the problem of type 2 diabetes-related mortality in the Canary Islands, Spain. The DARIOS Study

AIMS To compare diabetes-related mortality rates and factors associated with this disease in the Canary Islands compared with other 10 Spanish regions. METHODS In a cross-sectional study of 28,887 participants aged 35-74 years in Spain, data were obtained for diabetes, hypertension, dyslipidemia, obesity, insulin resistance (IR), and metabolic syndrome. Healthcare was measured as awareness, treatment and control of diabetes, dyslipidemia, and hypertension. Standardized mortality rate ratios (SRR) were calculated for the years 1981 to 2011 in the same regions. RESULTS Diabetes, obesity, and hypertension were more prevalent in people under the age of 64 in the Canary Islands than in Spain. For all ages, metabolic syndrome and insulin resistance (IR) were also more prevalent in those from the Canary Islands. Healthcare parameters were similar in those from the Canary Islands and the rest of Spain. Diabetes-related mortality in the Canary Islands was the highest in Spain since 1981; the maximum SRR was reached in 2011 in men (6.3 versus the region of Madrid; p<0.001) and women (9.5 versus Madrid; p<0.001). Excess mortality was prevalent from the age of 45 years and above. CONCLUSIONS Diabetes-related mortality is higher in the Canary Islands population than in any other Spanish region. The high mortality and prevalence of IR warrants investigation of the genetic background associated with a higher incidence and poor prognosis for diabetes in this population. The rise in SRR calls for a rapid public health policy response.

C-peptide as a risk factor of coronary artery disease in the general population:

OBJECTIVE To analyse the association between serum C-peptide and coronary artery disease in the general population. METHODS Follow-up study of 6630 adults from the general population. They were stratified into group 1 (no insulin resistance: C-peptide < third tercile and glycaemia < 100 mg/dL), group 2 (initial insulin resistance: C-peptide ⩾ third tercile and glycaemia < 100 mg/dL) and group 3 (advanced insulin resistance: glycaemia ⩾ 100 mg/dL). RESULTS After 3.5 years of follow-up, group 2 had a higher incidence of myocardial infarction (relative risk (RR) = 4.2, 95% confidence interval (CI) = 1.7-10.6) and coronary artery disease (RR = 3.5, 95% CI = 1.9-6.6) than group 1. Group 3 also had increased incidences of both diseases. In multivariable analysis of the entire population, groups 2 and 3 showed significant risks of myocardial infarction and coronary artery disease (RR > 3 and RR > 2, respectively). However, when people with diabetes were excluded, the increased risks were corroborated only in group 2 for myocardial infarction (RR = 2.8, 95% CI = 1.1-6.9; p = 0.025) and coronary artery disease (RR = 2.4, 95% CI = 1.3-4.6; p = 0.007). CONCLUSION Elevated C-peptide is associated with the incidence of myocardial infarction and coronary artery disease in the general population. It can be an earlier predictor of coronary events than impaired fasting glucose.Objective: To analyse the association between serum C-peptide and coronary artery disease in the general population. Methods: Follow-up study of 6630 adults from the general population. They were stratified into group 1 (no insulin resistance: C-peptide 3 and RR > 2, respectively). However, when people with diabetes were excluded, the increased risks were corroborated only in group 2 for myocardial infarction (RR = 2.8, 95% CI = 1.1–6.9; p = 0.025) and coronary artery disease (RR = 2.4, 95% CI = 1.3–4.6; p = 0.007). Conclusion: Elevated C-peptide is associated with the incidence of myocardial infarction and coronary artery disease in the general population. It can be an earlier predictor of coronary events than impaired fasting glucose.