Association between Subclinical Hypothyroidism and Glycemic Control in Older Adults in a Medical Center in Peru

Objective To determine whether there is an association between subclinical hypothyroidism and glycemic control in older adults who received care at the “Centro Médico Naval” from 2010 to 2015. Methods This retrospective analytical study analyzed a secondary database of the care of elderly in the study hospital. The sample was comprised of 1,385 older adults. To detect an association between variables, the Poisson regression with robust variance was used at a significance level of 95%. The analyses were carried out with the STATA 16 program. Results Of the elderly 45.6% were between 71 and 80 years old; 58.4% were women and 43.8% had a normal body mass index. There was evidence of inadequate glycemic control in 8.1% and subclinical hypothyroidism in 15.2% of the elderly patients. Subclinical hypothyroidism was more frequent in the inadequate glycemic control vs. adequate glycemic control populations (41.1% vs. 13.0%). In the multivariance analysis, subclinical hypothyroidism (aPR = 2.22 95% CI [1.47–3.36]) was independent factor associated with inadequate glycemic control (p < 0.001). Conclusions A significant association was detected between subclinical hypothyroidism and inadequate glycemic control in older adults who presented at the “Centro Médico Naval” from 2010 to 2015.


Introduction
An older adult is more predisposed to acquire metabolic disorders due to multiple factors.For example, the elderly have greater prevalence of systemic chronic infammatory reaction, oxidative stress, damaged DNA, cellular senescence, tissular dysfunction, and decreased mitochondrial dysfunction [1].
Additionally, aging is associated with deterioration of insulin secretion by beta cells in response to endogenous incretin hormones, which in turn is associated with a reduction in insulin sensitivity and promotes beta cell death by inducing mitochondrial dysfunction.Furthermore, there is a decrease in preadipocyte replication with expansion of senescent cells in the adipose tissue that increases lipotoxicity and favors a proinfammatory state [2].Aging is also associated with changes in corporal composition and an increase in fatty mass, especially the visceral type, combined with a decrease of lean and skeletal mass [3], all of which causes a gradual intolerance to glucose and poor glycemic control.
Alterations in thyroid hormones have repercussions on the metabolism.For example, in hypothyroidism there is a decrease in insulin production by beta cells, and muscle cells develop less insulin sensitivity.Subclinical hypothyroidism shows an increase in insulin resistance due to a decrease in glucose transportation by the glucose 2 transporter (GLUT2) caused by a translocation of the GLUT 2 gene.Moreover, DM is a risk factor for thyroid disease because it alters thyroid-stimulating hormone (TSH) levels and the conversion of T4 to T3 in peripheral tissues.Hyperinsulinemia and insulin resistance can cause a proliferation of thyroid tissue, increasing the risk of developing lymphatic nodules and goiter [4].
In studies restricted to older persons, the reported prevalence of overt hypothyroidism has ranged between 0.2-5.7% and subclinical hypothyroidism between 1.5% and 12.5% of patients, making it the most prevalent endocrine disease among the aged, with a mayor prevalence in the female population [5].In more recent studies in Europe, the prevalence of hypothyroidism was 4.7%, 4.11% for subclinical and 0.65% for overt hypothyroidism, with women over 65 years of age being the most afected [6].Tis high prevalence can be explained by a decrease in organifcation and absorption of iodine and altered response to TSH in the elderly.In addition, changes in bioactivity, the sensitivity of thyrocytes to TSH hormone thyroid metabolism, and in the receptors and cofactors that modulate response to the infux of T3 cells have been described.In other words, these processes lead to a decrease in thyroid hormone production [7].
Regarding the variables related to thyroid function and glycemic control, it has been observed that among people with diabetes, females taking antihypertensive medications, who have a low educational level, are associated with thyroid dysfunction [8].Additionally, DM type 2 patients that also have subclinical hypothyroidism have a greater risk of developing complications like cardiovascular disease [8,9] and proliferative diabetic retinopathy [10].
Determining if there is an association between subclinical hypothyroidism and glycemic control in the elderly is relevant because it can lead to a better understanding of the pathogenesis and treatment of thyroid disease and diabetes.Complications may be prevented with better control.
Te present study, therefore, aimed to determine if there is an association between subclinical hypothyroidism and glycemic control in older adults who presented at the Medical Navy Center of Peru between the years of 2010 and 2015.

Study Design.
Tis study was a retrospective analytical study including a secondary analysis of the database of a study conducted in a Medical Navy Center (CEMENA) located in the Bellavista-Callao district in Peru, in which retired military personnel and their direct family members reside and receive care.Te data were obtained from a cohort of adults older than 60 years of age from 2010 to 2015.Te study was approved by the Ethics Committee of the Cientifca del Sur University (No. 113-CIEI-CIENTIFICA-2019, and registry code: 178-2019-PRE15).Participant anonymity was maintained and only necessary information was used to conduct the study.

Setting and Participants.
Te original population of the study was comprised of 1987 participants over the age of 60 years who were treated on an outpatient basis in the Geriatrics Service of the Naval Medical Center "Cirujano Mayor Santiago Távara" from 2010 to 2015.Te participants were divided into 2 groups: frst, older adults with inadequate glycemic control, and second, older adults with adequate glycemic control.Te geriatric patients with a preceding diagnosis of primary or secondary hypothyroidism, DM, and the ones that did not have enough information for the variable of interest (glycemic control) were excluded.Te fnal sample was comprised of 1385 older adults.

Data Collection Process.
In the present study, the dependent variable was glycemic control and was defned as glycemic monitoring based on the evaluation of glycosylated hemoglobin (HbA1c) values and fasting glucose at baseline.Good glycemic control was considered with fasting glucose values between 90 and 130 mg/dL (5.0-7.2 mmol/L) and glycosylated hemoglobin levels ≤7.5%.Inadequate glycemic control was considered if any of these values were altered.Te independent variable was subclinical hypothyroidism, which was defned as TSH concentrations above the upper normal limit together with free T4 concentrations within the reference range.Reference values were between 5 and 15 mU/L for TSH and between 0.7 and 1.8 ng/dL for free T4 [11,12].
Te analyzed variables were age; sex; education; marital status; body mass index (BMI), weight in kilograms divided by the square of height in meters; comorbidities, the coexistence of two or more diseases in the same person, for example, renal chronic disease with or without controls, arterial hypertension with treatment or recently diagnosed; polypharmacy, consumption of various drugs at the same time, i.e., more than 5 drugs were considered.

Statistical Analysis.
Te statistical analysis was performed using the statistical program STATA version 16.Descriptive statistics were used to determine percentages.Te chi-square test was used to analyze potential associations between patient characteristics (age, sex, education, marital status, BMI, comorbidities, and polypharmacy) and glycemic control in the bivariate analysis.Finally, for the multivariate analysis, the Poisson regression with robust variance was used to determine if there is an association between the characteristics of older adults and glycemic control.Te prevalence ratio (PR) was calculated with a confdence interval of 95% (95% CI).

Results
In this study, initially 1987 participants were selected; however, the fnal study sample comprised of data from 1385 older adults as those individuals that did not meet the inclusion criteria were excluded.Regarding the general characteristics, 45.6% of participants were between 71 and 80 years old; 58.4% were females; 65.3% were married; 43.8% 2 Current Gerontology and Geriatrics Research had a normal BMI; 60.8% had arterial hypertension; 5.3% had chronic kidney disease; 29.5% had polypharmacy; and 15.2% had subclinical hypothyroidism (Table 1).Among the elderly studied, 8.1% had inadequate glycemic control, and 91.9% had adequate glycemic control.Signifcant diferences were found with respect to sex, marital status, BMI, arterial hypertension, chronic kidney disease, polypharmacy, and subclinical hypothyroidism between the two study groups (adequate versus inadequate glycemic control) (Table 2).In the frst group, 72.3% were females, compared to only 57.4% in the second group.Te proportion of married older adults was greater in the frst group (83.3% vs. 71.2%,respectively), while the proportion of widowers was smaller in this group (10.2% vs. 21.5%).On the other hand, the proportion of the elderly with normal weight was greater in the frst group (57.4% vs. 50.0%),while the proportion of older adults who were overweight was lower (13.9% vs. 28.9%).Te proportion of older adults with chronic kidney disease (18.8% vs. 4.1%), polypharmacy (96.4% vs. 25.0%), and subclinical hypothyroidism (41.1% vs. 13.0%) was greater in the frst group compared with the second group.Finally, the proportion of arterial hypertension was lower in the frst group (52.7% vs. 62.1%, respectively).

Discussion
Te present study found an association between subclinical hypothyroidism and inadequate glycemic control.Tese results are similar to those reported in previous studies [13][14][15].Tis relationship may be explained by the mechanisms of thyroid dysfunction such as alteration of the expression of a group of genes that cause physiological anomalies and decrease transport of glucose to cells, decrease absorption of splanchnic glucose, and increase release of glucose by the liver.Moreover, subclinical hypothyroidism can cause insulin resistance due to a translocation of the gene that encodes GLUT2 [4].
A considerable proportion of the aged was found to be overweight and/or obese, or had arterial hypertension, chronic renal disease, polypharmacy, and/or subclinical hypothyroidism.Tese fndings may be explained by an increase in chronic comorbidities and polypharmacy present in older adults due to physiological, cultural, and socioeconomic changes [16].Older adults have a greater risk of acquiring subclinical hypothyroidism due to a reduction in thyroid hormones caused by physiological changes, such as an alteration of the absorption and organifcation of iodine, a decrease in the sensitivity of thyrocytes to TSH, and pathological changes including comorbidities such as diabetes mellitus [2,3,7].In the present study, the prevalence of subclinical hypothyroidism was 15.2%, which is greater than what has been reported in other studies [6,17].
Additionally, 8.1% of the older adults in the present study had inadequate glycemic control.Physiological changes that occur in older adults may explain the inadequate glycemic control, including a greater composition of fatty mass with an increase in adipocyte infammation, a reduction in lean mass, an increase in chronic systemic infammation with oxidative stress, and a decrease in beta cells and their response to incretins which in turn induces an increase in insulin resistance [2,3].
Participants who were overweight had a lower probability of having inadequate glycemic control, while those with low weight had a greater risk of presenting inadequate glycemic control.Tese results difer from those of other studies which found that an increase in BMI was associated with poor glycemic control due to hyperinsulinemia and insulin resistance which are related to obesity [18].Furthermore, blood glucose levels may be more difcult to control in obese people [19].Tere were other studies that had similar results to those of the present study, which showed that an increase in BMI is associated with a lower HbA1c.Tis fnding may be attributed to the fact that weight loss caused by accumulated chronic metabolic infammation   Current Gerontology and Geriatrics Research and certain behaviors, such as smoking, can alter pancreatic beta cells and increase HbA1c [20].
Polypharmacy can be associated with poor adherence to treatment, with a decrease in hypoglycemic medication consumption and poor glycemic control [18].Some studies have described a rate of therapeutic compliance less than 50% in patients with DM type 2 [21].
It is possible that polypharmacy could be a risk factor for hypothyroidism given that it was present in the majority of the participants (96.4%).Te medications most associated with altered thyroid function include amiodarone, which can be administered for arrhythmias, a common condition in the elderly [22].Other medications include lithium, antidepressants, antiepileptics, and rifampicin [23].
Similar to DM as the principal cause of inadequate glycemic control, there are other comorbidities that could alter glycemia such as high BMI, medications like corticosteroids, estrogens, and phenytoin, as well as conditions like acromegaly and Cushing's syndrome [24][25][26].
According to the present fndings that subclinical hypothyroidism is associated with inadequate glycemic control in the elderly are important because subclinical hypothyroidism is difcult to diagnose due to its subtle clinical manifestations.Moreover, in elderly with multiple comorbidities the manifestations may be masked or go undetected.Nevertheless, even though hypothyroidism can have imperceptible clinical manifestations, it can have severe health consequences.Tus, these results suggest the need for screening for one of these pathological conditions in the aged when the other condition is present, to achieve an early diagnosis and apply interventions for their control to minimize consequences.
One of the main limitations of this study was the limited studies performed on a national level, and consequently, a lack of data with which to compare the results found in this investigation.Moreover, the population studied was derived from a small sector (Naval Medical Center) that may not refect the prevalence of the country.Furthermore, it is not possible to determine any causal relationships with the fndings in this study; so, implications should be taken cautiously.However, this study contributes fndings that may have clinical relevance to the scientifc community and may motivate future investigations from diferent institutional and population samples.
In conclusion, a signifcant statistical association was found between the presence of subclinical hypothyroidism and inadequate glycemic control.We recommend performing similar investigations in populations from other regions and institutions to obtain a better picture at the national level.

Table 1 :
General characteristics of the study sample.

Table 2 :
General characteristics of the sample according to glycemic control.
PR: prevalence ratio, CI: confdence interval.1Modeladjusted for all covariables present in the table.Calculated by Poisson regression.Association between general characteristics and glycemic control in older adults from the naval medical center 2010-2015.