nueva página del texto (beta)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
INTRODUCTION
A global aging population has motivated researchers to focus on risk factors, preventive medicine, and chronic diseases in preparation for the coming health issues that this phenomenon will cause. The international economy and society itself will suffer from the worldwide changes in population age distribution. Between the years 2015 and 2050, the proportion of the worlds population older than 60 years will increase from 900 million to 2000 million, a rise from 12% to 22%1. In 2015, Mexico had one older adult (≥ 60 years of age) for every 10 young adults (15 years old), a figure that is expected to double by 20502. At present, Mexico City is the entity in Mexico with the highest index of aging, 59.1 in 2014, and a projection of 108.7 by 20303. Due to this global demographic transition, successful aging (prevention of disease/disability, high physical and cognitive function, and sustained engagement in social activities) has become a priority4.
The concept of aging suggests the emergence and accumulation of chronic conditions such as heart disease, diabetes mellitus (DM), cancer, and cerebrovascular diseases, which are also the main causes of death at this stage of life. Hence, the identification of adverse health outcomes in older adults has been addressed by the characterization of frailty as a biological syndrome. Frailty has been described as the loss of physiologic reserve and resistance to stressors that involves multiple physical, mental, and emotional deficits5. Frailty increases vulnerability to dependence and the risk of adverse events such as functional decline, falls, hospitalization, institutionalization, and mortality5,6. In Mexico City, a prevalence of frailty of 20.6% among older adults has been reported7, an intermediate level when compared with global reports from 4% to 59.1%8. Hence, the search for biomarkers that can help define frailty at a biological level could be used to improve the diagnosis of frailty, to identify pre-frail patients at an earlier stage, and to evaluate the biological outcome when frailty is modified by factors such as healthy eating or physical exercise. On the matter, telomere length, which has been associated with lifespan9, has had a controversial role in frailty description. Although no correlation was found for Chinese10 and German11 populations, we recently observed shorter telomeres associated with frailty in a Mexican population12.
Telomeres are protective and regulatory structures of the genome; their shortening is associated with chronic degenerative diseases such as DM13, obesity, and hypertension14,15 and has been associated with lifespan9. Lifespan depends on many internal and external factors such as aging, pollution, quality of life, and education, among others. Consequently, telomere shortening and frailty would seem to be intimately related. Most of the associations with telomere length mentioned above involve the production of oxidative stress that can affect DNA in a direct form. Oxidative stress has been associated with chronic damage in several conditions, including DM and hypertension9,16-18. At a molecular level, oxidative stress is capable of altering DNA and has been linked to telomere length reduction19. To determine the possible association between reactive oxygen species (ROS), telomere length, and frailty, we evaluated the effect of oxidative stress on telomere length as a first approach to explain telomere length shortening in frail patients. Then, we assessed both telomere length and oxidative stress, as biomarkers of frailty.
METHODS
Participants
This is a cross-sectional study based on 2015 data from the Cohort of Obesity, Sarcopenia, and Frailty of Older Mexican Adults (COSFOMA), a study involving 1252 adults ≥60 years of age affiliated to the Mexican Social Security Institute (IMSS), and who were residents of Mexico City7. The IMSS is part of the social protection system in health in Mexico, which provides services to salaried workers and their families, including medical care and economic benefits such as disability pensions or paid retirement. In Mexico City, IMSS has 48 primary care medical units attending 36.5% of the population and approximately 50.9% of older adults. A total of 202 individuals from the 1252 involved in COSFOMA, agreed to participate in the study. In the COSFOMA study, participants were chosen according to their place of residence through a simple random selection from the list of older adults affiliated to IMSS from the 48 first-level clinics of Mexico City7. Written letters delivered to their home addresses were used as invitations to inform them of the nature of the study and to provide them with an appointment at their clinic. When the older adults did not attend the appointment, a phone invitation was made and in some cases, a home visit. Blood samples were obtained by venipuncture of the median cubital vein using the vacutainer system. Samples collected from April to September 2015 were used in this study.
Ethics statement
This research protocol was conducted with the approval of the National Committee of Scientific Research as well as by the Ethics Committee for Health Research of the Instituto Mexicano del Seguro Social (R-2012-785-067). All participants gave written informed consent. The study was performed according to the Code of Ethics of the World Medical Association Declaration of Helsinki.
Evaluation of ROS
To evaluate by fluorometry the formation of ROS, 5 µl of serum and 85 µl of PBS 1X buffer were added to a 10 µl solution of dichlorofluorescein (DCFH) diacetate (CAS Number 4091-99-0, Sigma-Aldrich), incubated in darkness for 30 min at 37°C to allow for oxidation of DCFH to the fluorescent compound 2-7-DCFH by the presence of hydrogen peroxide. Samples were then read at 498 nm excitation and 522 nm emission (Cytation 5 Cell Imaging Multi-Mode Reader), previously calibrated with a standard curve. Data are shown as pmol/µL of serum.
Determination of lipid peroxidation
To evaluate lipid peroxidation, we measured malondialdehyde (MDA) as a reaction of thiobarbituric acid (CAS Number 504-17-6, Sigma-Aldrich); 50 µl of serum with 25 µl of PBS 1X were added to 50 µl of thiobarbituric acid. The sample was placed in a boiling bath at 94°C for 20 min, then centrifuged at 10,500 rpm for 15 min; the supernatant was read at 532 nm with a spectrophotometer (EPOCH), which was previously calibrated with a standard curve. Data are shown as µmol/µL of serum.
Sample processing and telomere length assessment
Genomic DNA was extracted from peripheral leukocytes by the salting-out procedure. Purified DNA samples were aliquoted in a concentration of 10 ng/µL and stored at −70°C until use. For telomere length assessment, we followed the quantitative polymerase chain reaction (qPCR) method published by O´Callaghan and Fenech20. The number of copies of telomere repeats was determined by the standard curve of Tel STD, while the standard curve of 36B4 STD was used as a housekeeping gene. After the qPCR reaction on a StepOnePlus real-time PCR System (Applied Biosystem), the Ct values of each sample were extrapolated in their corresponding curves by a linear regression test to determine telomere length. The Maxima SYBR Green/ROX qPCR Master Mix 2X (Thermo Scientific, California, USA) was used. The cycling conditions for both genes were: 10 min at 95°C, followed by 40 cycles of 95°C for 15 s, 60°C for 1 min, followed by a melting curve. Data are shown as kilobase-pairs (kb).
Frailty
The operationalization of the frailty phenotype was performed using the five criteria proposed by Fried et al.6 Frail adults are defined as those with three or more of the following criteria: self-report of weight loss, exhaustion, low physical activity, slowness, and weakness (low grip strength). The presence of one or two criteria indicates a pre-frail condition, whereas the absence of criteria indicates a robust or non-frail state7. Participants were classified as non-frail (score 0), pre-frail (score 1-2), and frail (score 3-5).
Statistical analysis
In general, the data analysis and visualizations were carried out with free (R)21 and commercial (SPSS and GraphPad) software. In particular, we used ggplot222 and ggsignif23 R packages. Quantitative variables are presented as the arithmetic mean and standard deviation (mean±SD). Qualitative variables are presented as absolute (n) and relative (%) frequencies. Body mass index (BMI) was reclassified into nutritional status according to the WHO criteria (underweight <18.5 kg/m2, normal weight 18.5-24.9 kg/m2, overweight 25-29.9 kg/m2, and obesity ≥30 kg/m2) to explore the possible correlation between telomere length and nutritional status. To explore the distribution of DCFH (pmol/µL of serum), MDA (µmol/µL of serum), and telomere length (kb) between frailty categories (non-frail, pre-frail, and frail), we generated boxplots and calculated their p-value by MannWhitney Wilcoxon test with Bonferroni correction.
The relationship between DCFH and MDA with telomere length by frailty categories was explored by scatter plots and linear regression with the Pearson correlation coefficient (r). The distribution of telomere length and oxidative stress among older adult groups was evaluated calculating p-values obtained by KruskalWallis non-parametric test. The association between three frailty categories with DCFH, MDA, and telomere length was analyzed by multinomial logistic regression analysis with p-values obtained by Students t-test. Previous analyses were carried out to determine a possible linear correlation between oxidative stress, lipid peroxidation, and telomere length in three conditions: non-frail, pre-frail, and frail groups of older adults. However, to establish the magnitude of the association against the frailty category, we performed a logistic regression considering the non-frail group as the reference category to compare the odds of being classified as pre-frail or frail considering oxidative stress (DCFH or MDA) or telomere length. The odds ratio (ORs) with 95% confidence intervals (95% CI) was obtained for precision. Pearson correlation coefficients, as well as multinomial logistic regression analysis, were adjusted for age, BMI, chronic illnesses (0, 1, ≥ 2), educational level (without studies, 1-6 years, and ≥ 7 years), drinking (binary), smoking (binary), and sex (binary), as possible confounders.
RESULTS
From the random selection of participants in the COSFOMA study, we invited 1252 members and recruited 202 older adults (≥ 60 years of age) who agreed to participate in this study. Table 1 shows the general characteristics as well as ROS measured by DCFH, lipid peroxidation measured by MDA, and telomere length measurements of the older adults, 133 females (65.8%) and 69 males (34.2%), ranging between 60 and 95 years of age (mean age: 68.9 ± 7.4). Regarding the frailty status, 14.4% of participants were classified as non-frail, 60.4% as pre-frail, and 25.2% as frail.
Table 1 General characteristics, oxidative stress (reactive oxygen species and lipid peroxidation), and telomere length in a sample of Mexican older adults 60-95 years of age (n = 202)
| Characteristics | Participants | n (%) |
|---|---|---|
| Age (mean ± SD years) | 69.89 ± 7.39 | |
| Sex | Female | 133 (65.8) |
| Male | 69 (34.2) | |
| BMI (mean ± SD kg/m2) | 27.71 ± 6.59 | |
| Nutritional status | Underweight | 8 (3.96) |
| Normal weight | 63 (31.19) | |
| Overweight | 78 (38.61) | |
| Obesity | 53 (26.24) | |
| Years of school | No schooling | 16 (7.9) |
| 1-6 years | 78 (38.6) | |
| ≥ 7 years | 108 (53.5) | |
| Chronic illnesses | None | 115 (56.9) |
| 1 | 57 (28.2) | |
| ≥ 2 | 30 (14.9) | |
| Smoking (currently) | No | 187 (92.6) |
| Yes | 15 (7.4) | |
| Drinking (currently) | No | 168 (83.2) |
| Yes | 34 (16.8) | |
| Frailty status | Non-frail | 29 (14.4) |
| Pre-frail | 122 (60.4) | |
| Frail | 51 (25.2) | |
| Reactive oxygen species (DCFH, mean ± SD pmol/μL) | 50.27 ± 24.58 | |
| Lipid peroxidation (MDA, mean ± SD μmol/μL) | 33.53 ± 13.18 | |
| Telomere length (mean ± kb) | 5.31 ± 1.72 |
DCFH: dichlorofluorescein; MDA: malondialdehyde; SD: standard deviation.
The aim of this study was to determine a possible association between oxidative stress measured by ROS (DCFH) and lipid peroxidation (MDA), telomere length, and frailty status. We hypothesized that older adults with frailty would have increased oxidative stress measurements associated with shorter telomeres when compared with both pre-frail and non-frail subjects. Contrary to our hypothesis, we observed no significant differences between ROS (DCFH, p = 0.76) and lipid peroxidation (MDA, p = 0.37) among the frailty categories, whether divided or not by sex. Between frailty categories, we did find a significant difference in telomere length (p < 0.001). This difference was explained by lower telomere length in the frail group compared with the non-frail (p < 0.001) and the pre-frail (p < 0.001) groups, while the differences between non-frail and pre-frail were not statistically significant (p = 0.065) (Fig. 1).
Figure 1 Oxidative stress and telomere length among older adults. The figure shows the distribution of dichlorofluorescein (p = 0.76), malondialdehyde (p = 0.37), and telomere length (p < 0.001), among frailty groups (green: non-frail; yellow: pre-frail, and red: frail). The p-value was obtained by KruskalWallis nonparametric test. Panels A, B, and C show with thick lines the differences within frail categories while panels D, E, and F show the differences between females and males by MannWhitney Wilcoxon test. Bonferroni correction < 0.016 was calculated to assign significant differences. ***p-values < 0.001. p-value was obtained by KruskalWallis non-parametric test.
Then, to assess the possible association between lipid peroxidation (MDA), ROS (DCFH), and telomere length with groups of older adults, we performed a multinomial logistic regression adjusted by age, sex, BMI, chronic illnesses, educational level, drinking, and smoking. Figure 2 shows the magnitude and precision of the association of oxidative stress and telomere length in pre-frail and frail older adults when considering the non-frail subjects as the reference category. We observed that the odds of being classified in the pre-frail group increased 0.989 per each pmol of DCFH (95% CI [0.97-1.007]; p = 0.232), and 1.011 per each µmol of MDA (95% CI [0.971-1.052]; p = 0.611), while every kb of telomere (a long telomere) protected 1.33 times from a pre-frail classification (OR = 0.749, 95% CI [0.543-1.032]; p = 0.012). On the other hand, the odds of being classified in the frail group increased 0.986 per each pmol of DCFH (95%CI [0.964-1.008]; p = 0.217), and 1.018 per each µmol of MDA (95%CI [0.971-1.066]; p = 0.466), while every kb of telomere (a longer telomere) protected 2.28 times from a frailty classification (OR = 0.438, 95%CI [0.299-0.641]; p < 0.001).
Figure 2 Association of oxidative stress and telomere length in Mexican older adults. The forest plot shows the multinomial logistic regression considering the non-frail group as the reference category to calculate the odds of being classified as pre-frail or frail considering oxidative stress (dichlorofluorescein or malondialdehyde) or telomere length. The odds ratio with 95%confidence interval was obtained for precision and adjusted by age, sex, body mass index, educational level, chronic illnesses, smoking, and drinking. Bold text indicates a statistically significant difference (p-value cutoff < 0.05).
Finally, we evaluated the relationship between DCFH and MDA on telomere length according to frailty status. First, we calculated the global effect of both DCFH (r = −0.031, p = 0.66) and MDA (r = 0.095, p = 0.18) on telomere length and observed no significant correlation between ROS or lipid peroxidation and telomere length. Then, we performed the analysis for each frailty category (Fig. 3). In addition, since the effect of obesity or adiposity on telomere length has been reported previously24,25, and although a negative association of BMI with telomere length is reduced among older adults25, we explored the possible correlation between telomere length and nutritional status (underweight, normal weight, overweight, and obesity), no significant association was found with any of the groups: underweight (r = 0.640, p = 0.088), normal weight (r = 0.089, p = 0.49), overweight (r = 0.110, p = 0.33), and obesity (r = 0.027, p = 0.85), nor when subjects were divided by frailty categories (data not shown).
Figure 3 Relationship between oxidative stress and telomere length among frailty categories. The figure shows the correlation analysis for A: dichlorofluorescein and B: malondialdehyde, with telomere length. The analysis was conducted on the overall study population. Linear regression for non-frail (green; square), pre-frail (yellow; triangle), and frail (red; inverted triangle) participants with Pearsons correlation coefficients (r) and p-values adjusted by age, sex, body mass index, educational level, chronic illnesses, smoking, and drinking is shown.
DISCUSSION
Frailty, telomere length, and oxidative stress were evaluated in a group of older adults to define their association and correlation. A first analysis was conducted to determine our hypothesized distinct distributions between oxidative stress, lipid peroxidation, and telomere length according to frailty groups; however, we observed a significantly different distribution only for telomere length. Telomere length was found to be associated with pre-frail and frail categories, as previously reported12. This specific variable has become of importance for older adults since, despite the philosophical debate about the definition of frailty, it has shown to have a predictive value on poor health outcomes. Here, we have found that a shorter telomere, which by itself has been previously related to a briefer lifespan, is associated with the frailty status. Even more, when considering the telomere length of non-frail older adults as the reference category, short telomeres were associated with pre-frail older adults while even shorter telomere length was associated with the frail status. As these results were adjusted for age, sex, BMI, chronic illnesses, educational level, drinking, and smoking, possible confounders included in our analysis, we consider that telomere length is a strong candidate as a biomarker of frailty status. To confirm this, the present study must be replicated in other populations for telomere length to be considered as a biomarker of frailty since telomere length is affected by environmental factors26.
Telomere length is associated with modifiable factors such as smoking and obesity, which are related to shorter telomeres, while exercise and a healthy diet preserve telomere length4,27. Furthermore, a recent study showed that frailty can be modified by exercise and nutrition, reversing the frailty state to non-frail28. Hence, telomere length determination as part of a frailty diagnosis could help in the prompt implementation of intervention measures, in either the pre-frail or frail status, on the above-mentioned factors to reverse this condition. If pre-frail status and frailty are assessed on time, more costly medical outcomes could be prevented, and quality of life improved. Research in this direction may help determine frailty in a more quantitative manner with a biological background. Although telomere length determination currently is not widely available, its future implementation could be considered as part of the frailty diagnosis, as a non-invasive tool. On that matter, the association of frailty status with telomere length could mean that besides its definition, frailty has a biological counterpart that changes according to the frailty status. Hence, we evaluated oxidative stress as a probable cause of telomere shortening that could be assessed as telomere length and found a mild effect of MDA on telomere length (r = 0.17; p = 0.06) for the pre-frail category. MDA is a marker for lipid peroxidation, i.e., the oxidative degradation of lipids that follows the excess of free radical formation. The increase of MDA in pre-frail subjects and the decrease in telomere length between pre-frail and frail subjects suggest that oxidative stress affects telomere length regarding a frailty status in older adults.
The main limitation of the present study was that even when the original cohort employed consisted of 1252 individuals, we could only include 202 participants, meaning our recruitment skills were insufficient. The hypothesis generated in this study could be explored in a longitudinal study that would provide further information on the effect of MDA on telomere length and frailty status. Furthermore, a more thorough analysis of oxidative stress, specifically the measurement of 8-hydroxy-2-deoxyguanosine, a specific biomarker of oxidative stress in DNA29, would enrich the knowledge regarding the effect of these two biological outcomes of frailty. It is worth mentioning that when drinking, smoking, and chronic illnesses were not considered as confounder variables, we observed a linear association between MDA and telomere length (p = 0.089). We are convinced that an ideal approach would have these as exclusion criteria to avoid masking this biological measurement.
We confirmed an association of telomere length with the frailty phenotype and found a low association between lipid peroxidation and telomere length. However, we were unable to prove or disprove whether the cause of telomere length changes in frail patients is due to oxidative stress. Moreover, we found no effect of drinking or smoking as confounder variables regarding telomere length or oxidative stress in frail older adults in this study.
