Fibromyalgia and Obesity

The Association Between Body Mass Index and Disability, Depression, History of Abuse, Medications, and Comorbidities

Carmen E. Gota, MD; Sahar Kaouk; William S. Wilke, MD

Disclosures

J Clin Rheumatol. 2015;21(6):289-295. 

In This Article

Results

The diagnosis of FM was determined by the judgment of physicians (W.S.W. and C.E.G.); 96.7% of FM patients reported widespread pain for more than 3 months' duration, 98.3% reported fatigue, and 84.9% reported unrefreshing sleep. The ACR 1990 criteria were met by 79.7% of patients, 80% of patients reported at least 11 of 18 tender points, and 79.5% met the survey criteria by having an SIS score of 5.75 or greater. The results reflect previously reported concordance rates between clinical diagnosis of FM, ACR 1990, and survey criteria.[19] There was no difference between normal, overweight, and obese FM patients in the percentage of FM patients who met the ACR 1990 criteria or the survey criteria.

Of 305 FM patients, 224 had data to calculate BMI (73.2%). The distribution of FM cases by weight category is presented in Table 1. For analysis, because only 1 patient was underweight, we merged the data with the normal-weight group. We compared all demographic and clinical data between normal-weight, overweight, and obese groups. Obese patients were further classified into class I, 21.4%; class II, 10.3%; and class III, 12.1%. No further differences were found by including obesity subclasses (data not shown).

Table 2 compares variables among normal-weight, overweight and obese FM groups. We found no differences between normal-weight, overweight, and obese FM groups with regard to gender, race, insurance type, employment, education, marital status, or alcohol, tobacco, or drug use. Significant differences, however, were found for many variables, including medication categories, but other than mood measures (depression, anxiety, and bipolar disorder), variables that are intrinsic to the diagnosis and/or process of FM were not significantly different. Specifically, no differences were noted within BMI groups for the duration of FM, as well as any of the FM symptoms, including percentage of patients having at least 11 of 18 tender points, meeting the ACR 1990 criteria, survey criteria, RPS, SIS score, FIQ subsets, and FIQ total score.

Although ESS scores were no different among BMI categories, mean BMI differed between patients with sleep apnea versus those without. For 41 patients with sleep apnea, the BMI mean was 36.5 (SD, 8.8) kg/m2: 7.4% were of normal weight, 14.8% were overweight, and 53% were obese. We did not include sleep apnea data in Table 2 because of the limited number of patients with any data.

We performed bivariate linear correlations among variables and BMI (Table 3), which was then analyzed by creating a multivariable model of BMI built using these variables. Other than depression, none of these included core process FM variables.

Variables demonstrating little importance for predicting BMI were excluded using a step-down procedure and a penalized estimation. We found a low R 2 value for this model, at 0.15, which signifies that this particular model predicts only 15% of the BMI change. The only variables that remained significantly predictive of BMI change were the history of sexual abuse and the number of medications taken for FM, whereas disability measured by HAQ-DI and depression measured by PHQ-9 became insignificant. This analysis demonstrates that increasing BMI is independently correlated with very few variables and importantly not with core process FM variables (Table 4, Table 5 and Table 6).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....