Materials and Methods

A comparative survey was conducted, during 2017 in a regional hospital, specialized in Traumatology and Orthopedics of the Mexican Social Security Institute in Northeastern Mexico. The coexistence of depression and cognitive impairment increases the prevalence of depression, so individuals older than 65 years with HF (radiological diagnosis corroborated by two hip surgeons), without cognitive impairment (Mini Mental ≥ 24 point) were included. Exclusions included: individuals under antidepressants treatment, use of antipsychotic medications, re-fracture and surgical wound infection.

Sample selection: from the register of daily admissions for hip service, using a random probability sampling with replacement and sample size for proportion in infinity population with similar characteristics and frequency of depression in hospitalized older adults of 29%⁹ and d: 0.05 for 310 participants.

To register, participants were questioned directly for: age, gender, schooling, comorbidities (pulmonary, cardiovascular, urological, renal, liver disease, cancer, endocrinopathies and brain disease), economic support (pension, family or work), caregiver's gender, housing (own, rented, borrowed or temporary stay), sensory deprivation (visual or auditory), total of current medications, alcohol intake and smoking. Fasting venous blood samples were taken to determine glucose, hemoglobin, hematocrit, urea, creatinine, urea nitrogen, leukocytes and lymphocytes, processed by the Ilab 650 IL (Instrumentation Laboratory SpA. Milano, Italy). The second part of the questioning explored the measurement of the clinimetric variables of the geriatric evaluation before the HF occurred and included the following 12 scales: 1) Yesavage¹² (15 question version), used for depression screening, hetero-administered, with ten affirmative items and five negative, score of 0-4 are normal, 5-8 mild depression, 9-11 moderate depression, 12-15 severe depression, consistent (Cronbach's Alpha 0.86) and reproductible (r = 0.81)¹³; 2) Inventory of social resources of an OA of Díaz-Veiga (ISR)¹⁴, values the social environment that supports the OA, through relationships: spouse, children, relatives and friends. When scored ≤ 17 a support plan is recommended; 3) Instrumental Activities of Daily Living Scale (IADL) (modified Scale of Lawton and Brody), assesses five items: ability to use the telephone, use of transportation, shopping, responsibility for medication and management of their economy, the rating ranges between 0 (maximum dependence) and 5 (total independence) consistent (Cronbach's Alpha 0.94)¹⁵; 4) Basic activities of daily life (BADL) (Barthel Scale modified by Shah et al.), with ten items: personal hygiene, bathing, eating, using the toilet, climbing stairs, dressing, moving (walking on a smooth surface or in a wheelchair), transfer chair/bed, stool control and urine control. The results vary between 0 (completely dependent) and 100 points (completely independent), with κ = 0.77 interobserver and κ = 0.85 intraobserver¹⁶; 5) The nutritional assessment was performed with the MNA (Mini Nutritional Assessment), hetero-administered, used as screening and consisting of six items, if a score is obtained ≤ 10, then the evaluation takes place, if score is between 17 and 23.5 patient is in risk of malnourishment and < 17 points, indicates malnutrition. (Cronbach's Alpha 0.83) and ICC: 0.89¹⁷; 6) Fragility was assessed with the Barber questionnaire, which contains nine items. Each affirmative answer is worth a point. One or more points suggest risk situation, sensitivity of 0.95 and specificity of 0.68, PPV 0.91¹⁸; 7) The pain was measured with the visual analog scale (VAS), on a numerical scale from zero to ten, in which zero is absence and ten the maximum intensity of pain. The patient tells the number that best indicates that intensity, reliable (r = 0.94), useful even for illiterates (r = 0.71)¹⁹. It was applied twice, and the median was taken as the definitive one; 8) Insomnia was evaluated with the Insomnia Severity Index Scale (ISI) and consists of seven items, the score ranges from 0 (absence of insomnia) to 28 (severe insomnia), Cronbach's Alpha of 0.92, sensitivity, specificity of 82.4%, 82.1%, respectively²⁰; 9) Abuse was analyzed with the screening questionnaire of the American Medical Association (AMA), which contains nine items, a positive response is considered suspicion of abuse²¹; 10) The risk of pressure ulcers was evaluated with the Barden and Bergstrom scale, consisting of six items: sensory perception, exposure of the skin to moisture, physical activity, mobility, nutrition, friction and danger of skin lesions; a score between 6 and12 qualifies as low, 13-18 medium and 19-24 high risk, respectively, specificity 70% and sensitivity 77%²²; 11) Anxiety was screened with the Goldberg Anxiety and Depression Scale, with subscales for depression and anxiety, each with nine items of yes/no answers. The initial four items for screening and the other five are formulated when positive responses are obtained from the screening; (Cronbach's Alpha 0.65), sensitive (83.1%) and specific (81.8%)²³. For the study, only the anxiety questionnaire was applied; 12) The risk of hospital falls was evaluated with the Scale of Risk Factors for Falls, that consists of five factors, fluctuates from zero to ten, qualifies the risk as low (0-1), medium (2-3) and high (4-10) with Cronbach's Alpha 0.99²⁴.

These measures were applied between one to four days before undergoing surgery of the HF. They were applied by personnel unrelated to the study (two trained doctors), blindly and independently at the patient's bedside. All the data were directly taken from the patient, as well as caregivers, the inter-observer concordance was estimated for all the instruments. The protocol was previously authorized by the Local Research and Ethics Committee, No. 1903.

Prior to the study, weekly training meetings were held for those in charge of questioning the patients (two doctors), and they were instructed on the data collection instruments. The geriatric assessment was carried out by personnel specialized in geriatrics and psychiatry. A pilot test was conducted before the start of the study. Participants who met the selection criteria and wished to participate, were informed of objectives of the study and asked for their signed consent.

Results

Four hundred and twenty-three participants were eligible to enter the study, and 310 participants were admitted. One hundred and thirteen did not meet the selection criteria, 50 because they were taking antidepressants and 63 had a cognitive impairment. The median age was 78 years, interquartile range (IQR), (72-83) and 224 were women (72.3%). Depression and anxiety were detected in 45.8% and 38.1%, respectively. The median score of the Yesavage scale, in the total population, was 4 points, (IQR: 2-6), in subjects with and without depression was 6.5 and 3 points, (IQR: 5-8), (IQR: 2-4), respectively; only 21 subjects (6.8%) presented ≥ 10 points. (Table 1). From the comparison of geriatric variables and physical health, between subjects with and without depressive symptoms, 79.6% (113 participants) were women and 4.8% of the subjects had chronic renal failure. Most caregivers were female, in 223 subjects. Clinical factors were compared between patients with and without depression (Table 2); the characteristics with statistically significant differences (p = 0.001) were age, number of medications, hearing and visual deficit, gender, diabetes and abuse; of depressed patients, 53.5% had diabetes and 65.5% had hypertension. In turn, we analyzed the results of the scale of evaluation of risk factors for falls between both groups; we found in 228 patients (73.5%) a high risk of falling (p = 0.03). When comparing the two groups and their laboratory results (Table 3), only glucose showed significant differences in those with depression. Also, between the two groups all the clinimetric variables showed statistically significant differences and in the scores of the ISR scales of Díaz-Veiga, of Lawton and Brody, MNA and Barthel, the result was p < 0.001, (Table 4). To compare both groups of the Barden risk scale to develop ulcers, 213 (68.7%) were classified as high risk, but without significant differences.

Discussion

As mentioned before, depression is a common condition among elderly patients with a high prevalence. The 45.8% prevalence found in this study is in keeping with results reported in other studies^2,3,25. Likewise, Szczerbińska K. et al.²⁶ reported twice-higher prevalence in hospitalized OA compared to OA in asylums. In our report, it was observed that subjects with depressive symptoms were ten years older, 81 vs. 71 year-old, which implies that advanced age and hospitalization are factors of strong association with depressive symptoms in OA. On the other hand, the prevalence of women with depression was 79.6%, higher compared to a study conducted in Chile in subjects hospitalized in an internal medicine service, with 28.1% of female depression²⁷. This difference could be to the fact that depression as well as HF is more frequent in women²⁸, since women have longer life expectancy.

Another aspect remarkable is the large proportion of patients with diabetes mellitus in the total population, and in the group cataloged with depression. Because 53.9% of the total were diabetic and had depression, higher than the estimate in the report of Subramaniam M. et al.²⁹, who refer a frequency of diabetics by 25.5% and a 6% of patients with depression and diabetes. This reflects that diabetes is accompanied by mental alterations, which must be treated all together.

On the other hand, it is interesting to note the possible association between depression and anxiety with insomnia in elderly patients; this is supported by the report of Press et al.³⁰, who pointed out that such alterations are more frequent in patients with anxiety than in individuals with depression. Similar results were observed in this report, since patients with anxiety were more frequently found to have these alterations and identified statistically significant differences between groups of patients categorized with and without anxiety (p < 0.001). Also, patients with depression presented a statistically significant difference (p = 0.004), but with lower frequency of patients with depression and insomnia. In contrast to our study, Perlis et al.³¹ observed a higher prevalence of depression in elderly patients with persistent insomnia. Similarly, another study observed that elderly patients with insomnia have a probability of more than 23% risk of presenting depression³². These discrepancies must be grounds for further studies.

Another highlight of the study is the statistically significant difference that was observed in the results of the clinimetric evaluation scales that were applied to patients, which may reflect that in these patients, depression could have occurred before the fracture, in subclinical stages that may still affect the functional results in the elderly patients that are going through HF surgery. This was concluded in a study that refers to the depressive symptoms that arise after a HF and are associated with a decreased functional state and that clinical and psychosocial variables predict the tendency for depressive status³³. Likewise, Shahab et al.³⁴ point out that a routine evaluation with instruments such as the Yesavage scale, can help screen patients at risk of unsatisfactory results in rehabilitation and predict functional results after a HF in OA. It is relevant to consider the results of the evaluation of the MNA, with 80 subjects (56.3%) with depression and at risk of malnutrition and 20 (14.1%) with depression plus malnutrition. These results may be associated with the altered state of mind of these patients, resulting in lower nutrient intake, placing them at a risk of malnutrition before being hospitalized, and higher risk of falls due to muscle weakness and causing a possible HF that when resolved by major surgery, they present susceptibility through delayed healing and infections of surgical wound; as observed in the review carried out by Abd Aziz et al.³⁵, who reported that infections are more frequent in these patients.