Abstract As in human populations, advances in nutrition and veterinary care have led to an increase in the lifespan of companion animals. Detrimental physiological changes occurring later in life must be understood before interventions can be made to slow or reduce them. One important aspect of human aging is upregulation of the inflammatory response and increase in oxidative damage resulting in pathologies linked to chronic inflammation. To determine whether similar processes occur in the aging dog, changes in markers of inflammation and oxidative stress were investigated in 80 Labrador retrievers from adulthood to the end of life. Serum levels of immunoglobulin M (p < .001) and 8-hydroxy-2-deoxyguanosine (p < .001) increased with age, whereas no effect of age was detected for immunoglobulin G or C-reactive protein unless the last year of life was included in the analysis (p = .002). Baseline levels of heat shock protein 70 decreased with age (p < .001) while those after exposure to heat stress were maintained (p = .018). However, when excluding final year of life data, a decline in the heat shock protein 70 response after heat stress was observed (p = .004). These findings indicate that aging dogs undergo changes similar to human inflammaging and offer the possibility of nutritional or pharmacological intervention to delay or reduce these effects. Canine, Longitudinal, Immunosenescence, Inflammation, Aging Just as in the human population, advances in nutrition and veterinary care have led to an increase in the lifespan of companion animals. A 2012 survey in the United States estimated that 33% of the pet dog population was between 6 and 10 years of age with 15% older than 11 years (1). The detrimental changes that occur later in life are characterized by a declining ability to respond to stress, increased homeostatic imbalance, and increased risk of diseases ultimately leading to death. These changes are referred to collectively as senescence and must be understood before interventions can be made to slow or reduce their impact on quality of life. One important aspect of senescence is the upregulation of the inflammatory response and increased oxidative damage that manifest themselves as a low-level chronic proinflammatory state (2). The consequences of such inflammation have been shown to be wide ranging in humans and include neurodegenerative disease and type 2 diabetes (3,4). The origins of this chronic inflammation are multifactorial but are postulated to be due in part to age-related changes in the immune system (2). In humans, age-related decline of the immune response has been termed immunosenescence and is well documented in dogs (5). During this process, extensive remodeling takes place affecting both the humoral and innate immune systems. Cell functionality may be compromised and lymphoproliferation in response to mitogens such as concanavalin A and phytohaemagglutinin is reduced in dogs over 8 years of age (6,7). Phenotypic changes in the relative numbers of B cells, T cells, and their subsets in peripheral blood have been reported in aging dogs (7,8) with alterations in the expression of cell surface markers of activation or memory (9,10). In a longitudinal study of diet restriction in Labrador retrievers, it was reported that earlier death was associated with lower lymphoproliferative responses to mitogens, reduced total lymphocytes, T cells, CD4-to-CD8 cell ratio, and higher B-cell percentages (8). In older dogs, changes in B-cell populations have been noted to influence responses to antigens and vaccination (11), whereas altered levels of immunoglobulin A, G, and M positive cells in peripheral blood and tissues have been observed (12,13). Collectively, these changes can result in chronic inflammatory cytokine production and slower resolution of inflammation after infection or assault (14). The release of proinflammatory cytokines such as interleukin 1, interleukin 6, and tumor necrosis factor-alpha has been shown to rise with age in both humans and dogs (5,15) and leads to a concomitant rise in circulating acute phase proteins (16). One such acute phase protein commonly measured in longitudinal studies of inflammation is C-reactive protein (CRP). This protein is produced in response to inflammatory cytokines, primarily IL-6, but also TNF-α. The measurement of CRP has some advantages over that of cytokines in that it remains stable over prolonged time periods and has a half-life of 19–20 hours; thus, its synthesis rate is determined by the intensity of the inflammatory process (17). Moreover, CRP levels are not affected by medications other than corticosteroids that directly affect the underlying inflammatory disorder, eating, circadian, or seasonal variation (18). A consequence of this pro-inflammatory state is tissue damage, which induces the production of reactive oxygen species (ROS) and in turn further oxidative damage (2). To protect cells from ROS, cells produce endogenous antioxidants to quench these damaging molecules. In a canine lifelong study of diet restriction, lower serum levels of a number of antioxidants were associated with aging (19). However, these authors urged caution in interpreting the data due to significant litter effects and lack of species specificity in the assays. A progressive decline in the ability to control oxidative damage may result in changes to cellular proteins, lipids, and nucleic acids (20). DNA and RNA damage due to ROS has been postulated to play a central role in age-related loss of physiological function, including immunosenescence (21). Cellular proteins are protected from oxidative stress by a range of physiological mechanisms, including a system of chaperones, termed “Heat Shock Proteins” (HSP) which promote the refolding or autophagy of damaged cellular components. If this system fails, an intracellular build-up of damaged protein occurs, triggering the production of further inflammatory cytokines (22). The effects of this cycle of inflammation, tissue damage, and oxidative stress have become the focus of much research in an area of human biology dubbed “inflammaging.” This process offers the possibility of multiple points of intervention where the process could be delayed or reduced. Before interventions can be identified and validated for the dog, further longitudinal studies to characterize the biological effects of advancing age are required. We hypothesized that similar age-related changes would occur in dogs as in humans, notably in markers of inflammation, immunosenescence, and oxidative stress. Here, we report findings from a prospective longitudinal study in the dog from adulthood to the end of life. Materials and Methods Experimental Design In 2003, a longitudinal study was initiated with the primary aim of investigating the physiological effects of avocado extract supplementation in canine diets. A secondary aim was to monitor a range of measures in order to characterize aging in the dog. The study started when the dogs were adults and continued until the end of life. Two cohorts of dogs began the study with a prefeeding baseline period during which they were habituated to the facilities, and individual food allowances to maintain ideal body weight and body condition were established. After this period, dogs were randomized into two treatment groups and offered a diet with or without avocado extract concentrate. Measures were taken biannually until death. Hematological and serum biochemical profiles for all dogs housed at the former IAMS Pet Health and Nutrition Centre (Lewisburg, Ohio) were previously as part of a larger data set to compile healthy reference ranges for older dogs (23), and data solely from cohort 1 dogs describing longevity and causes of mortality have also been published previously (24). Animals A first cohort of Labrador retrievers (Labs) entered the study in 2003 and a second cohort was added in 2005. Cohort 1 was made up of 39 (27 females and 12 males) neutered adults with a median age of 6.4 (range 5.3–8.1) years. Cohort 2 was made up of 41 (29 females and 12 males) neutered adults with a median age of 4.2 (1.9–6.0) years. Dogs were supplied by three different breeders. DNA samples from all dogs were analyzed via the WISDOM PANEL (MARS Veterinary, Portland, USA) of 321 DNA markers and referenced to data from 200 previously sampled pure breeds Labrador dogs in their database. This indicated that all study dogs were purebreds with the majority (68) coming from U.S. “Show” breeding lines and the rest (12) from U.S. “Field” breeding lines (data not shown). The last dog from cohort 1 died in 2015 aged 17 years, and in January 2016, ten dogs from cohort 2 were still alive aged between 12 and 15 years. Data from dogs in both cohorts including the surviving dogs were combined for this analysis. Animal Husbandry and Veterinary Care Throughout the study, all dogs resided at the Pet Health and Nutrition Centre, Lewisburg, OH, USA and procedures were approved by the Institutional Animal Care and Use Committee. Dogs were housed in indoor pens with free access to outdoor runs throughout the day. Additionally, all dogs received daily socialization with a qualified animal welfare specialist and supervised daily outside exercise. Indoor kennel temperature was maintained at ~22°C (range 18–24°C), a relative humidity of 50% (range 40%–70%), and air flow between 10 and 30 changes per hour. Indoor living areas had a light/dark cycle of 12 hours on/12 hours off, from ~6 am to 6 pm and regular access to natural light via pen windows. Exercise yards contained enrichment devices, shaded areas, and a variety of dog toys were provided in the indoor pens and outdoor runs. The general health and overall condition of each dog were monitored daily by the animal care staff. Standard veterinary care and preventive medicine were maintained throughout the study (see Supplementary Material). Medical conditions were treated by the supervising veterinarian using standard veterinary protocols; however, cancer and severe, or life threatening, conditions were managed individually by the veterinary care team based on the dog’s quality of life assessment. These assessments were performed using 10-point Likert scale ratings (25) to review food/water intake, pain/discomfort, mobility, hygiene, happiness, and number of “good days.” End-of-life decisions due to medical conditions and/or declining quality of life were based on the recommendation of the supervising veterinarian in consultation with a team comprised of health care, husbandry, research staff, and nonstudy veterinarians. Humane euthanasia was elected when quality of life was determined to be poor and/or medical or surgical therapy was no longer effective. Four dogs died unexpectedly of natural causes without an end of life decision, for the purposes of analysis, data from these dogs were not treated differently. Diets and Feeding Dogs completed a prefeeding baseline period of 7–12 months (cohort 1: 12 months, cohort 2: 7 months). The diet fed during the baseline period represented a standard quality dry dog food that was nutritionally complete and balanced. After the prefeeding period, dogs were randomized 1:1 into two treatment groups and offered a nutritionally balanced dry diet with or without avocado extract (<.10%; Kemin Industries, Des Moines, IA, USA). All diets met Association of American Feed Control Officials recommendations (26), for details of dietary composition see Supplementary Table 1. The daily food allowance for each dog was based on the number of kcal required to maintain each dog within an optimal body condition range. During the prefeed period, daily food allowances were adjusted as needed to ensure each dog was within their optimal body condition range. At the conclusion of the prefeed, daily food allowances were set for each dog and did not change unless warranted by a change in their body condition score (BCS) or for veterinary purposes. Assessment of BCS occurred on a quarterly basis by qualified graders using a previously described five-point scale (27), a score of 3 was considered ideal. Between-grader agreement was assessed quarterly by calculation of a kappa score. A dog’s daily food was changed if the quarterly BCS was ≤2 or ≥4. The maximum allowable change to an individual dog’s daily food allowance was ±50 grams and this food amount was maintained until the next quarterly BCS assessment. The daily allowance of food was split equally between two 30 minute meals, am and pm each day. Food intake (grams) was recorded daily. Fresh water was available at all times in each housing unit. Biochemical Measures Serum CRP, IgG, IgM and 8-hydroxy-2-deoxyguanosine (8OHdG), and peripheral blood mononuclear cell (PBMC) intracellular heat shock protein 70 (HSP70) were measured biannually in January and July each year. Blood was collected via the jugular vein into an SST serum separation tube (Fisher Scientific, Pittsburgh, PA, USA, Cat No. 02-683-148) early in the morning when dogs were in the fasted (>12 hours) state. After clotting, serum was separated by centrifuging at 3,000 × g for 8 minutes in a refrigerated centrifuge, harvested, stored at −80oC, and analyzed within 2 months. CRP quantification was by canine-specific ELISA according to the manufacturer’s instructions (ICL, Portland, OR, USA, Cat No. E-40CRP). IgG, the most abundant immunoglobulin in the body and a marker of the ability to sustain a humoral immune response, was measured by canine-specific ELISA according to the manufacturer’s instructions (Bethyl Laboratories, Inc., Montgomery, TX, USA, Cat No. E40-118). IgM, the first antibody to appear in response to initial antigenic exposure, was determined using a canine-specific ELISA as recommended by the manufacturer (Bethyl Laboratories, Inc., Montgomery, TX, USA, Cat No. E40-116). According to manufacturer’s instructions, 8-oxo-2′-deoxyguanosine (8OHdG), a marker of DNA oxidation, was measured using an EIA (Percipio Biosciences, Inc., Burlingame, CA, USA, Cat No. 21026). As a measure of the ability to mount a response to cellular stress, levels of intracellular HSP70 were determined in PBMCs at baseline and after exposure to heat stress. Blood was collected from the jugular vein into a vacutainer blood collection tube containing lithium heparin (Fisher Scientific, Pittsburgh, PA, USA, Cat No. 02-689-7). Peripheral blood mononuclear cells were isolated by density gradient centrifugation using Histopaque1077 (Sigma, St Louis, MO, USA, Cat No. H8889) according to manufacturer’s recommendations. Isolated PBMCs were added to two microtiter plates (for basal and stimulated measurements) at a cell density of 1 × 106 per well. For the stimulated measurement, the plate was exposed to heat stress (47oC) for 30 minutes before both plates were incubated at 37°C, 5% CO2 for 24 hours. Cell lysates were then prepared and HSP70 quantified using an ELISA as recommended by the manufacturer (Enzo Life Sciences, Inc., Farmingdale, NY, USA, Cat No. ADI-EKS-700B). For all measures, samples were assayed under similar experimental conditions and produced acceptable standard curves (R > 95% with <10% CV). All assays were duplicated and surrogate samples were used whenever available. All equipment was maintained with required preventative maintenance and calibration. All measurements presented in the current study were made within 1 month after samples were collected to avoid possible degradation of proteins. For all measures, raw data were censored only if a technical issue was identified and recorded at the time of analysis. Statistical methods: When cohort 1 completed the study, interim analysis of combined data from both cohorts indicated no effect of avocado extract on any of the measures taken (see Supplementary Table 2). Therefore, analysis of the effects of aging was carried out on the combined data. Median age of death with 95% confidence intervals (CI) was estimated using a Cox proportional hazards model taking into account age of death and current age of surviving dogs (see Supplementary Figure 1). Measures were analyzed using a joint model fitted using the joineR library in R v3.2.0 statistical software (28). This comprised a longitudinal model (incorporating mixed effects to allow for repeated measures on a dog through time) and a survival model (incorporating association with survival time, to ensure missing data due to death or censoring did not lead to bias).This model was unable to allow for a nested random effects structure (ie, dog within cohort); therefore, noise due to cohort was not included as a random effect. Distributional assumptions were checked to ensure robustness of the statistical models by performing residual checks (eg, for randomness and constant variance). Residuals were found to have increasing variability for a number of measures and these were log10 transformed prior to analyses. Diet was initially fitted as a baseline covariate in the survival model and age by diet interaction in the longitudinal model. Gender was included as a fixed effect in the longitudinal component of all models and identified as not statistically significant (p > 0.05) using ANOVA. Body weight and BCS were also included as fixed effects and identified as not statistically significant (p > 0.05). Further to this, body weight and BCS, when modeled individually were identified as not statistically significant in the survival component of the joint models (p = .217 and p = .313, respectively). When the effect of diet and its interaction with age had been established as nonsignificant, they were removed. To adjust for the inflated false positive rate due to the analysis of multiple measures, significance was tested against a Bonferroni adjusted level of 0.0071 (0.05/7). Resulting parameter estimates from the joint models were provided with 95% CI (obtained through bootstrapping with N = 1000 simulations). Significance was calculated assuming a t-distribution for the ratio of the parameter estimate and its standard error (as estimated from the bootstrap), with the degrees of freedom equal to total number of dogs. All data were included in the models, but figures present changes between 6 and 13 years as all dogs took part in the study between these ages. Analyses were carried out with and without data from the last year of life to minimize the influence of acute pathology and avoid reverse causality as suggested by Takata and coworkers in a study investigating inflammatory markers in healthy humans (29). Results Survival analysis determined the median age of death to be 13.5 years (95% CI 12.7, 14.0). Necroscopies were carried out on each dog; four dogs died naturally, two of which were found to have underlying neoplastic disease. Twenty-eight dogs were humanely euthanized because of severely limited mobility; of these, 12 had underlying neoplasms, 26 had neoplastic disease, 5 gastro-intestinal, 4 renal/urinary tract, 2 neurological, and 1 respiratory disease. When explored longitudinally, levels of CRP increased with age by 11.5% (95% CI −7.6%, 33%) between the ages of 6 and 13 years when the last year of life was excluded although this did not reach the levels of statistical significance (p = .275). However, when the last year of life was included, a significant (p = .002) association between CRP and age was detected (Table 1, Figure 1), with concentrations estimated to increase by ~30% (95% CI 11%, 57%) between the ages of 6 and 13 years (Table 1). Table 1. Effect of Age on Markers of Inflammation, Humoral Immunity, and Oxidative Stress Given as Fold Change Per Year of Age Estimated fold change per year of age (95% CI) p-value Including last year of life Excluding last year of life Including last year of life Excluding last year of life CRP 1.038 (1.015, 1.066) 1.016 (0.988, 1.042) .002* .275 IgG 0.992 (0.978, 1.003) 0.995 (0.978, 1.009) .261 .525 IgM 1.047 (1.031, 1.062) 1.057 (1.036, 1.077) <.001* <.001* 8OHdG 1.060 (1.043, 1.086) 1.058 (1.041, 1.086) <.001* <.001* HSP70 Baseline 0.755 (0.719, 0.792) 0.730 (0.688, 0.778) <.001* <.001* HSP70 Heat Stress 0.938 (0.887, 0.982) 0.915 (0.859, 0.968) .018 .004* HSP70 Heat Stress - Baseline 1.207 (1.182, 1.232) 1.225 (1.193, 1.263) <.001* <.001* Estimated fold change per year of age (95% CI) p-value Including last year of life Excluding last year of life Including last year of life Excluding last year of life CRP 1.038 (1.015, 1.066) 1.016 (0.988, 1.042) .002* .275 IgG 0.992 (0.978, 1.003) 0.995 (0.978, 1.009) .261 .525 IgM 1.047 (1.031, 1.062) 1.057 (1.036, 1.077) <.001* <.001* 8OHdG 1.060 (1.043, 1.086) 1.058 (1.041, 1.086) <.001* <.001* HSP70 Baseline 0.755 (0.719, 0.792) 0.730 (0.688, 0.778) <.001* <.001* HSP70 Heat Stress 0.938 (0.887, 0.982) 0.915 (0.859, 0.968) .018 .004* HSP70 Heat Stress - Baseline 1.207 (1.182, 1.232) 1.225 (1.193, 1.263) <.001* <.001* Note: CI = Confidence interval; CRP = C reactive protein; 8OHdG = 8-oxo-2’-deoxyguanosine; HSP70 = heat shock protein 70 from peripheral blood mononuclear cells cultured with (heat stress) or without (baseline) exposure to heat stress at 47oC; IgG, IgM = immunoglobulins G and M. *Statistically significant at the Bonferroni adjusted level of p < .007. View Large Table 1. Effect of Age on Markers of Inflammation, Humoral Immunity, and Oxidative Stress Given as Fold Change Per Year of Age Estimated fold change per year of age (95% CI) p-value Including last year of life Excluding last year of life Including last year of life Excluding last year of life CRP 1.038 (1.015, 1.066) 1.016 (0.988, 1.042) .002* .275 IgG 0.992 (0.978, 1.003) 0.995 (0.978, 1.009) .261 .525 IgM 1.047 (1.031, 1.062) 1.057 (1.036, 1.077) <.001* <.001* 8OHdG 1.060 (1.043, 1.086) 1.058 (1.041, 1.086) <.001* <.001* HSP70 Baseline 0.755 (0.719, 0.792) 0.730 (0.688, 0.778) <.001* <.001* HSP70 Heat Stress 0.938 (0.887, 0.982) 0.915 (0.859, 0.968) .018 .004* HSP70 Heat Stress - Baseline 1.207 (1.182, 1.232) 1.225 (1.193, 1.263) <.001* <.001* Estimated fold change per year of age (95% CI) p-value Including last year of life Excluding last year of life Including last year of life Excluding last year of life CRP 1.038 (1.015, 1.066) 1.016 (0.988, 1.042) .002* .275 IgG 0.992 (0.978, 1.003) 0.995 (0.978, 1.009) .261 .525 IgM 1.047 (1.031, 1.062) 1.057 (1.036, 1.077) <.001* <.001* 8OHdG 1.060 (1.043, 1.086) 1.058 (1.041, 1.086) <.001* <.001* HSP70 Baseline 0.755 (0.719, 0.792) 0.730 (0.688, 0.778) <.001* <.001* HSP70 Heat Stress 0.938 (0.887, 0.982) 0.915 (0.859, 0.968) .018 .004* HSP70 Heat Stress - Baseline 1.207 (1.182, 1.232) 1.225 (1.193, 1.263) <.001* <.001* Note: CI = Confidence interval; CRP = C reactive protein; 8OHdG = 8-oxo-2’-deoxyguanosine; HSP70 = heat shock protein 70 from peripheral blood mononuclear cells cultured with (heat stress) or without (baseline) exposure to heat stress at 47oC; IgG, IgM = immunoglobulins G and M. *Statistically significant at the Bonferroni adjusted level of p < .007. View Large Figure 1. View largeDownload slide Log10 CRP (mg/L) against age (years). Top: including the last year of life (p = .002). Bottom: excluding the last year of life (p = .275). Light grey lines are individual dog profiles with age, the black line the estimated average, and the dark grey shaded area the 95% confidence interval. Figure 1. View largeDownload slide Log10 CRP (mg/L) against age (years). Top: including the last year of life (p = .002). Bottom: excluding the last year of life (p = .275). Light grey lines are individual dog profiles with age, the black line the estimated average, and the dark grey shaded area the 95% confidence interval. Although serum concentrations of IgG were observed to fall slightly with age, this was not found to be significant (p = .261), even when the last year of life was excluded (p = .525; Table 1, Figure 2). However, the levels of IgM were estimated to increase by 38% (95% CI 24%, 53%) between the ages of 6 and 13 years. This increase remained significant when the last year of life was excluded (p < .001; Table 1, Figure 3). Figure 2. View largeDownload slide Log10 IgG (mg/mL) against age (years). Top: including the last year of life (p = .261). Bottom: excluding the last year of life (p = .525). Light grey lines are individual dog profiles with age, the black line the estimated average, and the dark grey shaded area the 95% confidence interval. Figure 2. View largeDownload slide Log10 IgG (mg/mL) against age (years). Top: including the last year of life (p = .261). Bottom: excluding the last year of life (p = .525). Light grey lines are individual dog profiles with age, the black line the estimated average, and the dark grey shaded area the 95% confidence interval. Figure 3. View largeDownload slide Log10 IgM (mg/mL) against age (years). Top: including the last year of life (p < .001). Bottom: excluding the last year of life (p < .001). Light grey lines are individual dog profiles with age, the black line the estimated average, and the dark grey shaded area the 95% confidence interval. Figure 3. View largeDownload slide Log10 IgM (mg/mL) against age (years). Top: including the last year of life (p < .001). Bottom: excluding the last year of life (p < .001). Light grey lines are individual dog profiles with age, the black line the estimated average, and the dark grey shaded area the 95% confidence interval. Irrespective of the inclusion of data from the final year of life, a significant (p < .001) positive association was observed between serum levels of 8OHdG and age. When data from the last year of life were included in the analysis, an increase of 51% (95% CI 34%, 78%) between the ages of 6 and 13 years was estimated (Table 1, Figure 4). Figure 4. View largeDownload slide Log10 8OHdG (ng/mL) against age (years). Top: including the last year of life (p < .001). Bottom: excluding the last year of life (p < .001). Light grey lines are individual dog profiles with age, the black line the estimated average, and the dark grey shaded area the 95% confidence interval. Figure 4. View largeDownload slide Log10 8OHdG (ng/mL) against age (years). Top: including the last year of life (p < .001). Bottom: excluding the last year of life (p < .001). Light grey lines are individual dog profiles with age, the black line the estimated average, and the dark grey shaded area the 95% confidence interval. A significant (p < .001) decrease in baseline HSP70 was associated with age when intracellular levels of PBMC HSP70 were investigated (Table 1). This association remained significant when data from the last year of life were excluded (Table 1) and was estimated to represent a reduction of 86% (95% CI 57%, 80%) between the ages of 6 and 13 years. To a lesser extent, a reduction in heat stress stimulated HSP70 was also observed, representing an estimated decrease of 36% (95% CI 57%, 12%) between the ages of 6 and 13 years (when the last year of life was included, Table 1). This reduction was not statistically significant at the corrected level; however, when the last year of life was excluded from the analysis, a significant reduction was detected (p = .004; Table 1). As decline in baseline HSP70 levels was greater than the decline in levels induced by heat stress, the difference between the two increased significantly with age (p < .001; Table 1, Figure 5). Figure 5. View largeDownload slide Log10 difference (ng/mL) between baseline HSP70 and HSP70 after exposure to heat stress against age (years). Top: including the last year of life (p < .001). Bottom: excluding the last year of life (p < .001). Light grey lines are individual dog profiles with age, the black line the estimated average, and the dark grey shaded area the 95% confidence interval. Figure 5. View largeDownload slide Log10 difference (ng/mL) between baseline HSP70 and HSP70 after exposure to heat stress against age (years). Top: including the last year of life (p < .001). Bottom: excluding the last year of life (p < .001). Light grey lines are individual dog profiles with age, the black line the estimated average, and the dark grey shaded area the 95% confidence interval. Discussion To characterize aging in the dog, changes in markers of inflammation and oxidative stress were investigated from adulthood to the end of life. Serum levels of IgM and 8OHdG increased with age, whereas no effect of age was detected for IgG or CRP unless the last year of life was included in the analysis. Baseline levels of PBMC HSP70 decreased with age while those after exposure to heat stress did not change significantly unless data from the final year of life were included in the analysis. In a number of mammalian and, indeed, invertebrate species, such changes have been reported to be associated with chronic low-grade inflammation in a process referred to as “inflammaging” (5). A study in Caenorhabditis elegans, an important model for studying the genetics of aging, indicated a gradual, age-dependent decline in the number of cells positive to anticytokine polyclonal antibodies, thus suggesting that cytokine-like molecules cross-reacting with antibodies raised against human IL-1a, PDGF-AB, and TNF-α, exist in lower invertebrates and likely play a role, in worm aging and lifespan (30). A number of mechanisms related to age have been proposed to underlie these changes including immunosenescence, the effects of increased visceral adipose tissue, decline in production of steroid sex hormones, and accumulative oxidative damage. CRP is produced by the liver in response to proinflammatory cytokines including IL-6 and TNF-α and is a sensitive, although non-specific, marker of inflammation in both human and the dog (31–33). In the current study, canine serum levels of CRP increased with age, although only reaching statistical significance when the final year of life was included. An association with age has been widely reported in the human population, where older people have levels within the normal range but higher than those of the younger population (32). Little information is available for the aging dog, but in agreement with our findings, one cross-sectional study in healthy beagles of various ages observed no age-related differences in CRP levels (34). It may be that dogs display a smaller increase in inflammation (as indicated by CRP) with age than is the case in humans while variability may be greater making small effects more difficult to detect. It is also possible that in dogs, other acute phase proteins, for example, serum amyloid A, are more informative of the inflammatory state than CRP as has been suggested recently (35). The loss of statistical significance when data from the final year of life were excluded may be unsurprising as elevated levels of CRP have been noted to be indicative of several acute diseases diagnosed in the study dogs including, malignant tumors (16) and increased CRP concentrations have previously been associated with poor clinical outcomes and death in dogs (36). It remains unclear, however, whether increased levels of CRP, or the particular cytokines which induce its production, are the direct causes of adverse effects or are the result of existing pathologies. Immunosenescence and accompanying immune dysregulation have been associated with physical impairment in the elderly human population (37). This loss of function is likely to be mirrored in the dog as similar age-related changes occur in both the immune systems of both species (5). For example, alterations in B- and T-cell repertoires have been reported in both aging humans and dogs (7,8,37). As in elderly humans, older dogs exhibit changes in B-cell populations which influence immune responses to antigens and vaccination (11,12,38). In studies of aging humans, there is evidence of a reduction in B-cell progression to memory cells, which leads to an overall reduction in the numbers of IgG producing cells and reduced circulating levels of IgG (37). This is in contrast to findings in aging mice where levels of IgG increased with age and isotype ratios changed with increased IgG1:IgG2a indicating changes in the humoral immune pathway (39). In agreement with a number of previous canine studies, here, little change in serum IgG was observed with age (9,12,40). Although in the current study levels of IgG were maintained, a shift with age toward increased levels of IgM did occur. In humans, expression of B-cell transcription factor E47 decreases with age (41). This factor regulates activation-induced cytidine deaminase expression lowering levels critical for class switching and antibody production (37). Impaired class switching leads to higher levels of IgM and reduced levels of IgG (37,41,42). This imbalance in humans has been shown to leave individuals vulnerable to infection and unable to mount an adequate response to vaccination (39,41,42). Similarly, studies in older dogs have suggested that the primary response to novel vaccines may be compromised (12,38). This could be indicative of a reduced ability to switch the primary IgM response to a longer lasting IgG response leading to the lower vaccine efficacy and increased vulnerability to infection well documented in elderly dogs (12,38). Current results are limited to changes in total IgG and IgM levels rather than individual isotypes, but to better understand the overall picture of humoral immune responses during canine aging, it would be essential to study changes of other isotypes and immunoglobulins, for example, IgA and IgE in future studies. Increased oxidative stress, resulting from an imbalance between the production and exposure to ROS and the host’s antioxidant defenses, is one proposed cause of senescence (43). If ROS are not cleared effectively, cellular damage to proteins, lipids, and nucleic acids occurs, ultimately resulting in cell death. When nuclear and mitochondrial nucleic acids are damaged by free radicals, oxidative derivatives are formed, the most abundant of these being 8OHdG, the oxidation product of deoxyguanosine. To protect the cell, 8OHdG is excised by cellular repair systems and can, therefore, be measured as a biomarker of oxidative damage (44). Serum levels of 8OHdG have been reported to be associated with a number of degenerative conditions including Alzheimer’s disease, arthritis, and type II diabetes (44,45) and to increase with age in both humans and mice (46). In the current study, we also observed a significant increase in the serum levels of 8OHdG with age. This increase could be a causative factor in a number of canine age-related diseases. For example, geriatric dogs exhibiting behavior related to cognitive dysfunction have been found to have 8OHdG levels which correlate with both age and dementia scores (47). Increased oxidative stress and serum 8OHDG levels have also been implicated in immunosenescence and chronic inflammation (48). Oxidative stress induces intracellular protein damage, which is encountered in a number of ways, including by HSP chaperones that promote protein refolding or autophagy (49). HSP70 is one of the key chaperone proteins that recognizes and transfers severely damaged proteins into the lysosome for degradation (49). This chaperone is constitutively produced intracellularly, but expression is upregulated after cellular stress, for example, heat or oxidative stress. In the current study, we observed an age-related reduction in baseline levels of intracellular HSP70 in a mixed population of PBMCs. Such a reduction has been reported previously in mice where tissue HSP70 basal levels show age-related changes (50). Here, levels of HSP70 after exposure to heat stress also declined with age, but to a lesser extent than baseline levels, only becoming statistically significant when data from the last year of life were excluded. This finding may suggest that HSP70 expression is increased when underlying pathology is present. If the baseline and heat stress responses are compared, with the caveat that they were not measured in the same assay plate, it would appear that with age the ability to produce baseline levels becomes impaired while the ability to mount a response to an acute stress is maintained. This profile may also be apparent in aging humans as while an age-dependent decline in primary cell HSP70 expression has been reported (51), the ability to mount a cellular heat stress response comparable with that of younger adults has also been noted (52). These observations may suggest that an adequate stress response is advantageous to survival and therefore preserved into old age even when constitutive HSP70 levels decline. The observation that the ability to maintain the HSP70 response to stress is particularly marked in human centenarians, and long-lived mice may also supported this (50,53). The consequences of reduced HSP70 levels have not been well characterized in the dog; however, they have been linked to cognitive decline and neurodegeneration (54). Intracellular HSP70 has also been noted to have a number of anti-inflammatory properties and can reduce responses to pro-inflammatory cytokines such as TNF-α and IL-1 by blocking activation of NFκB (54,55). Lower levels of intracellular HSP70 could therefore result in raised proinflammatory cytokines and hence the chronic inflammation associated with senescence and aging. The aim of this study was to characterize the effect of aging on markers of immunosenescence and oxidative stress in the dog. The majority of studies of canine aging have been cross-sectional in design and therefore infer changes over time by looking for differences between age groups. Here, we present data from the first longitudinal study of this size and power, allowing changes with time to be detected and followed under controlled conditions. Although all the individuals in the study were apparently free from gross pathology affecting their quality of life, minor conditions will have been present in individuals which could act as a confounding factor. However, the study was adequately powered and appropriately analyzed statistically to avoid undue bias. To reduce this risk further, data were analyzed including and excluding the final year of life as described by Takata and coworkers (29). As the study cohort was made up of dogs receiving nutritionally complete diets, with high standards of husbandry and veterinary care, the changes observed with age were small and therefore represented a slow move away from homeostasis. Such small changes could be influenced by minor differences in assay performance over time and this should be considered when drawing conclusions from the data. In the current study when assay performance issues were clearly identified at the time of analysis, the affected data points were censored. However, to avoid bias, no data were removed retrospectively; therefore, unidentified changes in assay performance could be a confounding factor. One advantage of a cross-sectional design is that it allows all samples to be analyzed at the same point in time; however, in longitudinal designs such as ours, samples have to be analyzed in batches to avoid issues of analytic stability, thus interassay variability can be encountered. It is therefore important that appropriate quality control standards are in place that can be applied to sample analyses over the duration of the study. Our findings are only representative of aging Labrador retrievers, although it is likely that they can be applied to the dog in general. However, breed differences in immune parameters have been identified previously (6), and therefore, variation in how these change with age could also occur. Further studies would be required to determine if our findings are applicable across breeds. Although the dogs were kept in conditions as close to those of pets as possible, it is unclear whether our observations are representative of pet dogs in general. Our data suggest that aging dogs undergo a change in immunoglobulin production and the ability to respond to intracellular oxidative stress. Furthermore, our findings demonstrate an age-related increase in oxidative nucleic acid damage and inflammation. To the best of our knowledge, it is the first time that such observations have been made in a longitudinal canine study of this size. Our findings indicate that as in humans, aging dogs exhibit a profile of changes indicative of “inflammaging,” a chronic stimulation of the immune system and reduced ability to respond to infections or stresses such as ROS. It is clear that this process contributes to the canine aging process. The pathophysiology of inflammaging is therefore of great therapeutic importance offering multiple targets for interventions to defend against and delay its effects. For example, there may be a need to optimize vaccination protocols for older dogs and to support their defenses against ROS by supplementation with antioxidants. Although many interventions have been suggested to address facets of inflammaging, a combined approach may be more effective in slowing the proinflammatory cycle. Although aging is unavoidable, the development of such strategies to support this process is essential to promote a long and healthy old age. Supplementary Material Supplementary data is available at The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences online. Funding This work was supported by The IAMS Company and the WALTHAM Centre for Pet Nutrition. They are a division of MARS Petcare. Conflict of Interest None. Acknowledgments We thank the research team at the Royal Canin Pet Health and Nutrition Centre (Lewisburg, OH) for their care of the study dogs and, in particular, Elizabeth Fuess for data collection and collation throughout the course of the study. We also thank Prof. D. Ingram and Dr. G. Roth for their contribution to the initial study design. Although the underlying data used to generate the research published herein were collected in collaboration with GeroScience, Inc., it did not involve in deriving the conclusions appearing in this publication. References 1. American Veterinary Medical Association . US Pet Ownership & Demographics Sourcebook . Schaumburg, IL : American Veterinary Medical Association ; 2012 . 2. Baylis D , Bartlett DB , Patel HP , Roberts HC . Understanding how we age: insights into inflammaging . Longev Healthspan . 2013 ; 2 : 8 . doi: 10.1186/2046-2395-2-8 . Google Scholar CrossRef Search ADS PubMed 3. Bettcher BM , Kramer JH . Inflammation and clinical presentation in neurodegenerative disease: a volatile relationship . Neurocase . 2013 ; 19 : 182 – 200 . doi: 10.1080/13554794.2011.654227 . Google Scholar CrossRef Search ADS PubMed 4. Paolisso G , Rizzo MR , Mazziotti G et al. Advancing age and insulin resistance: role of plasma tumor necrosis factor-alpha . Am J Physiol . 1998 ; 275 ( 2 Pt 1 ): E294 – E299 . Google Scholar PubMed 5. Day MJ . Ageing, immunosenescence and inflammageing in the dog and cat . J Comp Pathol . 2010 ; 142 Suppl 1 : S60 – S69 . doi: 10.1016/j.jcpa.2009.10.011 . Google Scholar CrossRef Search ADS PubMed 6. Faldyna M , Levá L , Knötigová P , Toman M . Lymphocyte subsets in peripheral blood of dogs—a flow cytometric study . Vet Immunol Immunopathol . 2001 ; 82 : 23 – 37 . Google Scholar CrossRef Search ADS PubMed 7. Greeley EH , Ballam JM , Harrison JM , Kealy RD , Lawler DF , Segre M . The influence of age and gender on the immune system: a longitudinal study in Labrador Retriever dogs . Vet Immunol Immunopathol . 2001 ; 82 : 57 – 71 . Google Scholar CrossRef Search ADS PubMed 8. Greeley EH , Kealy RD , Ballam JM , Lawler DF , Segre M . The influence of age on the canine immune system . Vet Immunol Immunopathol . 1996 ; 55 : 1 – 10 . Google Scholar CrossRef Search ADS PubMed 9. Reis AB , Carneiro CM , Carvalho Md et al. Establishment of a microplate assay for flow cytometric assessment and it is use for the evaluation of age-related phenotypic changes in canine whole blood leukocytes . Vet Immunol Immunopathol . 2005 ; 103 : 173 – 185 . doi: 10.1016/j.vetimm.2004.08.014 . Google Scholar CrossRef Search ADS PubMed 10. Lawler DF , Larson BT , Ballam JM et al. Diet restriction and ageing in the dog: major observations over two decades . Br J Nutr . 2008 ; 99 : 793 – 805 . doi: 10.1017/S0007114507871686 . Google Scholar CrossRef Search ADS PubMed 11. Mansfield KL , Burr PD , Snodgrass DR , Sayers R , Fooks AR . Factors affecting the serological response of dogs and cats to rabies vaccination . Vet Rec . 2004 ; 154 : 423 – 426 . Google Scholar CrossRef Search ADS PubMed 12. HogenEsch H , Thompson S , Dunham A , Ceddia M , Hayek M . Effect of age on immune parameters and the immune response of dogs to vaccines: a cross-sectional study . Vet Immunol Immunopathol . 2004 ; 97 : 77 – 85 . Google Scholar CrossRef Search ADS PubMed 13. Kleinschmidt S , Meneses F , Nolte I , Hewicker-Trautwein M . Distribution of mast cell subtypes and immune cell populations in canine intestines: evidence for age-related decline in T cells and macrophages and increase of IgA-positive plasma cells . Res Vet Sci . 2008 ; 84 : 41 – 48 . doi: 10.1016/j.rvsc.2007.03.009 . Google Scholar CrossRef Search ADS PubMed 14. Deitschel SJ , Kerl ME , Chang CH , DeClue AE . Age-associated changes to pathogen-associated molecular pattern-induced inflammatory mediator production in dogs . J Vet Emerg Crit Care (San Antonio) . 2010 ; 20 : 494 – 502 . doi: 10.1111/j.1476-4431.2010.00565.x . Google Scholar CrossRef Search ADS PubMed 15. Vasto S , Candore G , Balistreri CR et al. Inflammatory networks in ageing, age-related diseases and longevity . Mech Ageing Dev . 2007 ; 128 : 83 – 91 . doi: 10.1016/j.mad.2006.11.015 . Google Scholar CrossRef Search ADS PubMed 16. Nakamura M , Takahashi M , Ohno K et al. C-reactive protein concentration in dogs with various diseases . J Vet Med Sci . 2008 ; 70 : 127 – 131 . Google Scholar CrossRef Search ADS PubMed 17. Vigushin DM , Pepys MB , Hawkins PN . Metabolic and scintigraphic studies of radioiodinated human C-reactive protein in health and disease . J Clin Invest . 1993 ; 91 : 1351 – 1357 . doi: 10.1172/JCI116336 . Google Scholar CrossRef Search ADS PubMed 18. Fröhlich M , Sund M , Thorand B , Hutchinson WL , Pepys MB , Koenig W . Lack of seasonal variation in C-reactive protein . Clin Chem . 2002 ; 48 : 575 – 577 . Google Scholar PubMed 19. Stowe HD , Lawler DF , Kealy RD . Antioxidant status of pair-fed labrador retrievers is affected by diet restriction and aging . J Nutr . 2006 ; 136 : 1844 – 1848 . Google Scholar CrossRef Search ADS PubMed 20. Bansal M , Kaushal N. Oxidative stress mechanisms and their modulation . New York: Springer ; 2014 . Google Scholar CrossRef Search ADS 21. Blount DG , Heaton PR , Pritchard DI . Changes to levels of DNA damage and apoptotic resistance in peripheral blood mononuclear cells and plasma antioxidant potential with age in labrador retriever dogs . J Nutr . 2004 ; 134 ( 8 Suppl ): 2120S – 2123S . Google Scholar CrossRef Search ADS PubMed 22. Salminen A , Kaarniranta K , Kauppinen A . Inflammaging: disturbed interplay between autophagy and inflammasomes . Aging (Albany, NY) . 2012 ; 4 : 166 – 175 . doi: 10.18632/aging.100444 . Google Scholar CrossRef Search ADS 23. Bellows J , Colitz CM , Daristotle L et al. Defining healthy aging in older dogs and differentiating healthy aging from disease . J Am Vet Med Assoc . 2015 ; 246 : 77 – 89 . doi: 10.2460/javma.246.1.77 . Google Scholar CrossRef Search ADS PubMed 24. Adams VJ , Watson P , Carmichael S , Gerry S , Penell J , Morgan DM . Exceptional longevity and potential determinants of successful ageing in a cohort of 39 Labrador retrievers: results of a prospective longitudinal study . Acta Vet Scand . 2016 ; 58 : 29 . doi: 10.1186/s13028-016-0206-7 . Google Scholar CrossRef Search ADS PubMed 25. Likert R . A technique for the measurement of attitudes . Arch Psychol . 1932 ; 140 . 26. Association of American Feed Control Officials . Model regulations for pet food and specialty pet food under the model bill . IN: Oxford ; 2016 . 27. Burkholder WJ . Use of body condition scores in clinical assessment of the provision of optimal nutrition . J Am Vet Med Assoc . 2000 ; 217 : 650 – 654 . Google Scholar CrossRef Search ADS PubMed 28. Philipson P , Sousa I , Diggle PJ , Williamson P , Kolamunnage-Dona R , Henderson R . joineR: joint modelling of repeated measurements and time-to-event data, R package version 1.0-3 . 2017 . 29. Takata Y , Xiang Y-B , Villegas R et al. Dietary patterns and plasma adipokines and inflammatory markers in healthy Chinese men (40.3) . The FASEB Journal . 2014 ; 28 ( 1 Supplement ): 40.43 . 30. Franchini A , Perruzi E , Ottaviani E . Morphochemical age-related changes in the nematode Caenorhabditis elegans: immunoperoxidase localization of cytokine- and growth factor-like molecules . Eur J Histochem . 2003 ; 47 : 75 – 80 . Google Scholar CrossRef Search ADS PubMed 31. Michaud M , Balardy L , Moulis G et al. Proinflammatory cytokines, aging, and age-related diseases . J Am Med Dir Assoc . 2013 ; 14 : 877 – 882 . doi: 10.1016/j.jamda.2013.05.009 . Google Scholar CrossRef Search ADS PubMed 32. Wyczalkowska-Tomasik A , Czarkowska-Paczek B , Zielenkiewicz M , Paczek L . Inflammatory markers change with age, but do not fall beyond reported normal ranges . Arch Immunol Ther Exp (Warsz) . 2016 ; 64 : 249 – 254 . doi: 10.1007/s00005-015-0357-7 . Google Scholar CrossRef Search ADS PubMed 33. Ceron JJ , Eckersall PD , Martýnez-Subiela S . Acute phase proteins in dogs and cats: current knowledge and future perspectives . Vet Clin Pathol . 2005 ; 34 : 85 – 99 . Google Scholar CrossRef Search ADS PubMed 34. Kuribayashi T , Shimada T , Matsumoto M et al. Determination of serum C-reactive protein (CRP) in healthy beagle dogs of various ages and pregnant beagle dogs . Exp Anim . 2003 ; 52 : 387 – 390 . Google Scholar CrossRef Search ADS PubMed 35. Christensen MB , Langhorn R , Goddard A et al. Comparison of serum amyloid A and C-reactive protein as diagnostic markers of systemic inflammation in dogs . Can Vet J . 2014 ; 55 : 161 – 168 . Google Scholar PubMed 36. Gebhardt C , Hirschberger J , Rau S et al. Use of C-reactive protein to predict outcome in dogs with systemic inflammatory response syndrome or sepsis . J Vet Emerg Crit Care (San Antonio) . 2009 ; 19 : 450 – 458 . doi: 10.1111/j.1476-4431.2009.00462.x . Google Scholar CrossRef Search ADS PubMed 37. Boraschi D , Del Giudice G , Dutel C , Ivanoff B , Rappuoli R , Grubeck-Loebenstein B . Ageing and immunity: addressing immune senescence to ensure healthy ageing . Vaccine . 2010 ; 28 : 3627 – 3631 . doi: 10.1016/j.vaccine.2010.03.035 . Google Scholar CrossRef Search ADS PubMed 38. Kennedy LJ , Lunt M , Barnes A et al. Factors influencing the antibody response of dogs vaccinated against rabies . Vaccine . 2007 ; 25 : 8500 – 8507 . doi: 10.1016/j.vaccine.2007.10.015 . Google Scholar CrossRef Search ADS PubMed 39. Sharma R , Kapila R , Haq MR , Salingati V , Kapasiya M , Kapila S . Age-associated aberrations in mouse cellular and humoral immune responses . Aging Clin Exp Res . 2014 ; 26 : 353 – 362 . doi: 10.1007/s40520-013-0190-y . Google Scholar CrossRef Search ADS PubMed 40. Strasser A , Teltscher A , May B , Sanders C , Niedermüller H . Age-associated changes in the immune system of German shepherd dogs . J Vet Med A Physiol Pathol Clin Med . 2000 ; 47 : 181 – 192 . Google Scholar CrossRef Search ADS PubMed 41. Blomberg BB , Frasca D . Age effects on mouse and human B cells . Immunol Res . 2013 ; 57 : 354 – 360 . doi: 10.1007/s12026-013-8440-9 . Google Scholar CrossRef Search ADS PubMed 42. Frasca D , Diaz A , Romero M , Landin AM , Blomberg BB . Age effects on B cells and humoral immunity in humans . Ageing Res Rev . 2011 ; 10 : 330 – 335 . doi: 10.1016/j.arr.2010.08.004 . Google Scholar CrossRef Search ADS PubMed 43. Stadtman ER . Role of oxidant species in aging . Curr Med Chem . 2004 ; 11 : 1105 – 1112 . Google Scholar CrossRef Search ADS PubMed 44. Valavanidis A , Vlachogianni T , Fiotakis C . 8-hydroxy-2’ -deoxyguanosine (8-OHdG): A critical biomarker of oxidative stress and carcinogenesis . J Environ Sci Health C Environ Carcinog Ecotoxicol Rev . 2009 ; 27 : 120 – 139 . doi: 10.1080/10590500902885684 . Google Scholar CrossRef Search ADS PubMed 45. Nunomura A , Perry G , Pappolla MA et al. RNA oxidation is a prominent feature of vulnerable neurons in Alzheimer’s disease . J Neurosci . 1999 ; 19 : 1959 – 1964 . Google Scholar CrossRef Search ADS PubMed 46. Yamauchi Y , Matsuno T , Omata K , Satoh T . Relationship between hyposalivation and oxidative stress in aging mice . J Clin Biochem Nutr . 2017 ; 61 : 40 – 46 . doi: 10.3164/jcbn.16-79 . Google Scholar CrossRef Search ADS PubMed 47. Rofina JE , van Ederen AM , Toussaint MJ et al. Cognitive disturbances in old dogs suffering from the canine counterpart of Alzheimer’s disease . Brain Res . 2006 ; 1069 : 216 – 226 . doi: 10.1016/j.brainres.2005.11.021 . Google Scholar CrossRef Search ADS PubMed 48. Pálmai-Pallag T , Bachrati CZ . Inflammation-induced DNA damage and damage-induced inflammation: a vicious cycle . Microbes Infect . 2014 ; 16 : 822 – 832 . doi: 10.1016/j.micinf.2014.10.001 . Google Scholar CrossRef Search ADS PubMed 49. Schlesinger MJ . Heat shock proteins . J Biol Chem . 1990 ; 265 : 12111 – 12114 . Google Scholar PubMed 50. de Toda IM , Vida C , Ortega E , De La Fuente M . Hsp70 basal levels, a tissue marker of the rate of aging and longevity in mice . Exp Gerontol . 2016 ; 84 : 21 – 28 . doi: 10.1016/j.exger.2016.08.013 . Google Scholar CrossRef Search ADS PubMed 51. de Toda IM , De la Fuente M . The role of Hsp70 in oxi-inflamm-aging and its use as a potential biomarker of lifespan . Biogerontology . 2015 ; 16 ( 6 ): 709 – 721 . Google Scholar CrossRef Search ADS PubMed 52. Marini M , Lapalombella R , Canaider S et al. Heat shock response by EBV-immortalized B-lymphocytes from centenarians and control subjects: a model to study the relevance of stress response in longevity . Exp Gerontol . 2004 ; 39 : 83 – 90 . Google Scholar CrossRef Search ADS PubMed 53. Ambra R , Mocchegiani E , Giacconi R et al. Characterization of the hsp70 response in lymphoblasts from aged and centenarian subjects and differential effects of in vitro zinc supplementation . Exp Gerontol . 2004 ; 39 : 1475 – 1484 . doi: 10.1016/j.exger.2004.07.009 . Google Scholar CrossRef Search ADS PubMed 54. Romanucci M , Della Salda L . Oxidative Stress and Protein Quality Control Systems in the Aged Canine Brain as a Model for Human Neurodegenerative Disorders . Oxid Med Cell Longev . 2015 ; 2015 : 940131 . doi: 10.1155/2015/940131 . Google Scholar CrossRef Search ADS PubMed 55. Zheng Z , Kim JY , Ma H , Lee JE , Yenari MA . Anti-inflammatory effects of the 70 kDa heat shock protein in experimental stroke . J Cereb Blood Flow Metab . 2008 ; 28 : 53 – 63 . doi: 10.1038/sj.jcbfm.9600502 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2017. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: email@example.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)
The Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences – Oxford University Press
Published: Nov 6, 2017
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera