Effects of Age on Thermal Sensitivity in the Rat

Abstract

Age-dependent changes in thermal sensitivity were evaluated with reflex- and operant-based assessment strategies in animals ranging in age from 8 to 32 months. The impact of inflammatory injury on thermal sensitivity was also determined in animals of different ages. The results showed that operant measures of escape behavior are needed to demonstrate significant changes in thermal sensitivity across the life span of female Long-Evans rats. Increased escape from both heat (44.5°C) and cold (1.5°C–15°C) was observed for older animals, with a greater relative increase in sensitivity to cold. Physical performance deficits were demonstrated with aging but were not associated with changes in escape responding. Reflex responding to cold stimulation was impaired in older animals but was also influenced by physical disabilities. Reflex responding to heat was not affected by increasing age. Inflammation induced by formalin injections in the dorsal hindpaw increased thermal sensitivity significantly more in older animals than in their younger counterparts.

IT is well documented that non-nociceptive sensory capacities decline as a function of advancing age. Sensitivity decreases for hearing, taste, smell, touch, and vision due in part to a diminished number of specialized peripheral receptors and deterioration of supporting tissues (1).

Consistent with other sensory modalities, numerous age-related anatomical, chemical, and functional changes have been documented in the somatosensory system (2). Peripheral nerves of older animals and humans show a reduction of myelinated and unmyelinated fibers (4,5) as well as signs of damage, including axonal involution and Wallerian degeneration (5,6). The number and size of sensory nerve cells in lumbar dorsal root ganglia also increase through early adulthood, peak at midlife (13–18 months), and then decrease thereafter (7,8). Age-related peripheral afferent reduction, demyelination, and inflammation are similar to forms of axonal disruption that occur following nerve or tissue injury in younger animals (8–14).

In addition to peripheral changes, altered expression of neurotransmitters and receptors is observed in the spinal cords of old animals and human postmortem material. For example, calcitonin gene related peptide, substance P, somatostatin, and nitric oxide are diminished with advancing age, potentially decreasing pain sensitivity (15–19). Furthermore, the aging dorsal horn, especially lamina I, exhibits changes suggesting degeneration of descending modulatory circuits, including decreased levels of 5-serotonin and norepinephrine (12,20). Such changes could decrease inhibitory influences and enhance activation of spinal neurons by peripheral input. Accordingly, increased spontaneous activity and stimulus-evoked activation of spinal neurons has been observed in old animals (21). Also, modifications of expression and the functional state of spinal glial cells provide an evolving construct for increased pain sensitivity with age (22,23). Thus, numerous age-related changes within peripheral and central somatosensory pathways could create a predisposition to increased or decreased pain sensitivity.

Studies evaluating pain sensitivity in humans have reported a pattern of increased threshold but decreased tolerance with advancing age (11,24–26). These results have been interpreted in terms of the opposite influences of age on the sensory dimension of pain (pain detection threshold) versus the affective dimension (tolerance threshold) (27,28). Alternatively, reduced peripheral input with age may reduce sensitivity to non-nociceptive and mild nociceptive stimulation (at or below pain detection threshold), whereas increased excitability within intact pain pathways could increase the magnitude of suprathreshold pain sensations.

Understanding the biologic process of aging and the impact on the perception of nociceptive information requires development of animal models and behavioral strategies to monitor changing sensibilities across the life span of experimental animals. Animal studies examining the effects of age on pain sensitivity have resulted in conflicting observations, including increases, decreases, or no change in cutaneous sensibilities with advancing age (3-11). The approach used in these studies has been to employ reflex-based behavioral measures of nociceptive sensitivity, which do not appropriately reveal the effects of experimental treatments known to affect pain sensitivity of humans (Green et al., unpublished data, 2009) (29–37). Therefore, both operant escape and reflex tests were used in the present study to evaluate the impact of age on thermal sensitivity under normal conditions and during chronic inflammation following formalin injection in the dorsal hindpaws (38). Age-related changes in physical performance were also measured and related to the results of reflex and operant tests of nociception. Operant measures of pain assessment revealed an increased thermal sensitivity across the life span of the rat, but reflex responding was not similarly affected by age. Furthermore, inflammation induced by formalin injections increased thermal sensitivity significantly more in older animals than in their younger counterparts. A preliminary description of these data has been reported (39).

METHODS

Animals

Experiments were carried out using 36 female Fischer 344 × Brown Norway rats purchased from the National Institute on Aging. Animals were pair housed and maintained on a 12-hour light and 12-hour dark cycle, with food and water available ad libitum in a facility accredited by the American Association for Accreditation of Laboratory Animal Care. Animals were assessed daily for signs of health problems and were palpated weekly to monitor for symptoms of gross tumors. All experiments were approved by the Institutional Animal Care and Use Committee at the University of Florida.

Experimental Design

Four groups of animals were purchased at the following ages: (a) group 1: 6 months (n = 12), (b) group 2: 14 months (n = 6), (c) group 3: 22 months (n = 6), and (d) group 4: 30 months (n = 6). Following acclimation and 2 weeks of training on thermal escape and thermal preference tasks, baseline data collection was initiated. Collection of stable baseline data was completed when groups 1–4 were 8, 16, 24, and 32 months of age. At this time, each animal received bilateral subcutaneous injections of formalin into the dorsal hindpaws (38). Post-formalin testing was carried out for 3 months to monitor the duration of the formalin effect and to ensure recovery from the effects of injections. At the conclusion of this 3-month period, each animal was evaluated for reflex lick and guard (L/G) responding to thermal stimulation. During reflex testing, animals in groups 1–4 were 11, 19, 27, and 35 months of age, respectively. Nine months after purchase, animals in groups 1–3 began a repeat of baseline testing, thus increasing the number of animals to n = 18 at 16 months and n = 12 at 24 and 32 months, when collection of baseline data was completed. Groups 1–3 received second formalin injections in the dorsal hindpaws at 16, 24, and 32 months of age. In addition, a new group of animals 6 months of age (n = 6) was purchased, increasing to 18 the number of animals completing baseline testing and formalin injections at 8 months.

Formalin Injections

Once stable baselines were obtained for groups 1–4, animals received a 50 μL subcutaneous injection of dilute formalin (5% in saline) in the dorsal surface of each hindpaw (one injection per day). The formalin hindpaw injection model has been used successfully to examine effects over time of inflammatory sensitization of cutaneous nociceptors (38). Thermal sensitivity of the ventral surface of the paws after formalin injection into the dorsal surface was compared with baseline performance to determine if secondary hyperalgesia from an inflammatory injury occurred and varied as a function of age.

Behavioral Outcomes

Thermal escape.—

The escape test was similar to that used previously (29,30,34). Briefly, a two-compartment shuttle box, divided by a hanging partition, was used in this test. In a dark compartment, a thermally regulated floor delivered neutral, cold, or heat stimulation to the rat's paws. In the adjacent compartment, a thermally neutral platform provided escape from thermal stimulation. A 35-W halogen light illuminated the escape compartment. The escape test therefore presents a conflict between motivations to (a) escape bright light and (b) escape thermal nociception. Rats resolve this conflict by alternating occupancy between the two compartments, with time on the platform positively related to the intensity of the thermal stimulus (34).

Animals were trained over a 2-week period to become familiar with the testing protocol, to discriminate among different plate temperatures, and to learn to apportion their time on the platform to escape from thermal stimulation. Initial training included exposure to a neutral temperature followed by increasing temperatures (36°C, 42°C, 44°C, and 47°C). After sessions without light in the escape compartment, animals were exposed to the same series of temperatures in the presence of light. Rats were tested four times per week during two consecutive 15-minute trials per day. The first trial (36°C) standardized foot temperatures prior to the second trial with the plate temperature set to 10°C, 15°C, 42°C, or 44.5°C. Escape responses were assessed by platform durations (time spent occupying the escape platform). Data collection was automated, and therefore, there was no opportunity for experimenter bias. The same experimenter tested all animals throughout the study (blinded to all experimental conditions), thus eliminating any influence of different handling procedures.

Thermal preference.—

A thermal preference test was carried out as previously described (37). This operant paradigm permits animals to choose between two compartments with different floor temperatures. The test was used to determine if aging preferentially affected cold or heat sensitivity. Animals were trained over a 2-week period to become familiar with the testing protocol and discriminate between different plate temperatures. Baseline testing presented temperatures (10°C and 45°C) that produce a near-equal preference for the two compartments. A pretest (36°C, one compartment) was used to standardize foot temperatures prior to each test session. Pretest and test trials were 15 minutes in duration. Collection of thermal preference data was automated to avoid experimenter bias and was evaluated as the time spent occupying each compartment during a trial.

L/G testing.—

Licking and guarding responses were observed in an apparatus consisting of a single compartment with a thermally regulated floor and no escape platform. First, latencies to lick and to guard were measured. Temperatures and trial times were chosen so rats did not experience stimuli that produced damage to the feet. The test temperatures included 1.5°C and 44.5°C following a pretest exposure to 36°C. Trial times for each temperature were 15 minutes. Sessions of L/G testing with different temperatures were conducted on separate days. Cold stimulation of 1.5°C was utilized because licking/guarding behaviors occur rarely for cool temperatures that produce moderate levels of escape responding by young animals (eg, 10°C and 15°C).

Physical Performance Measures

Standard measures of physical performance were used in the assessment of animals at different ages (40,41).

Grip strength.—

Forelimb grip strength was determined using an automated grip strength meter (Columbus Instruments, Columbus, OH). The experimenter grasped the rat by the tail and suspended it above a grip ring. After approximately 3 seconds, the animal was lowered toward the grip ring and allowed to grasp the ring with its forepaws. The animal's body was lowered to a horizontal position until the grasp of the ring was broken. The force in grams was determined with a computerized electronic strain gauge fitted directly to the grasping ring, and the peak force was divided by body weight. The maximal measurement from three successful trials was taken as the final outcome. Successful trials were defined as those in which the animal grasped the ring with both forepaws without jerking.

Inclined plane.—

This test measures muscle tone and stamina. Rats were placed facing upward in one compartment of an apparatus that contained 60-degree tilted 1-cm mesh screens. The time taken for the animal to fall onto a 7.6-cm-thick foam pad was recorded with a maximum allowable latency of 10 minutes.

Data Analysis

Escape performance is displayed as cumulative platform durations for up to 10 platform occupancies within each trial. Escape responding by groups of animals at different ages typically plateaued within 9 or 10 responses for cold or heat stimulation (Figures 3 and 4). However, platform durations for individual animals under certain conditions could be limited to three responses, and therefore, all statistical comparisons are based on the first three escape responses.

Figure 3.

Stimulus–response functions for temperatures ranging from 10.0°C to 44.5°C for animals in different age-groups. Responses to cold and heat for animals 8 months of age are shown in (A) and (C), respectively. Responses to cold and heat for animals 32 months of age are shown in (B) and (D), respectively. Escape duration (ordinate) is summed across escape platform occupancies (eg, response numbers 1–10 on the abscissa represents the sum of escape durations for the first 10 responses in the trial). Compared with responses to nociceptive cold or heat stimulation, there was only a minimal response to neutral stimulation (36°C). Escape durations increased with increasing intensity of nociceptive cold or heat except for the enhanced responding of 32-month-old animals to cold, where there was a ceiling effect.

Figure 4.

Age-related responses of animals to different temperatures. Animals in four groups ranging in age from 8 to 32 months were evaluated on the thermal escape task for responses to 15°C (A), 10°C (B), 42°C (C), and 44.5°C (D). Accumulated escape durations in seconds are represented on the y-axis, and the number of summed response durations for each data point is represented on the x-axis. To control for possible effects of age on activity levels, differences between the cumulative escape durations for nociceptive stimulation and the cumulative escape durations for 36°C stimulation are plotted. For each temperature, the maximal response durations (ie, greatest sensitivity) were observed for animals 24–32 months of age, whereas the lowest response durations (ie, least sensitivity) were for the youngest animals.

Thermal preference is plotted as the cumulative difference in successive pairs of occupancies on the heated or cooled plates (Figure 5). Durations on the cold plate are subtracted from durations on the hot plate. Thus, positive values represent a heat preference and negative values represent a cold preference.

Figure 5.

Thermal preferences for animals of different ages. Animals ranging in age from 8 to 32 months were evaluated for their preference to heat (45°C) or cold (10°C). The graph is a plot of the difference in time spent on the hot plate minus time spent on the cold plate, accumulated for 1–15 pairs of responses (occupancies of the cold and hot plate). An increasing preference for time on the hot plate was observed with increasing age suggesting that the age-related increase in sensitivity to cold was greater than that for heat (see Figure 4).

Comparisons between groups determined differences attributable to age, and within-group comparisons were used for stimulus intensity and formalin injury. Thermal sensitivity was evaluated by the latency of reflex responding, the duration of escape responding, and the relative duration of heat versus cold plate occupancy (thermal preference). Repeated measures analysis of variance (ANOVA) was utilized for the thermal escape and thermal preference tests (StatSoft, Inc., Tulsa, OK). Reflex latencies were evaluated with one-way ANOVA. Pairwise multiple comparisons for evaluation of body weights and physical performance were carried out using the Student–Newman–Keuls method. A probability level of .05 was used for significance.

RESULTS

Age Effects on Physical Performance

Behavioral performance in older animals requires consideration of compromised physical performance as a potential confound in the interpretation of behavioral measurements. For this reason, physical performance of groups 1–4 was measured approximately every 6 weeks. Inclined plane, grip strength, and body weight measurements are shown in Figure 1. The body weights of animals were relatively stable from the ages of 6–18 months and then significantly increased and remained relatively constant from the ages of 26–32 months (Figure 1, left; F = 48.6, df = 4, p < .001). Pairwise multiple comparisons showed that 6-, 12-, and 18-month weights were significantly different from those of 26 and 32 months (p < .01). With increasing age, performance declined significantly for grip strength (Figure 1, middle; F = 22.2, df = 4, p < .001) and the inclined plane (Figure 1, right; F = 27.8, df = 4, p < .001). No significant difference in grip strength was observed between the ages of 6 and 18 months, but there was a significant decline in performance at 26 and 32 months (6, 12, and 18 months significantly different from 26 and 32 months, p < .01). Performance on the inclined plane was similar at 6 and 12 months and declined significantly at 18, 26, and 32 months (6, 12, and 18 months significantly different from 26 and 32 months, p < .01).

Figure 1.

Body weight and physical performance data obtained in rats ranging in age from 6 to 32 months. Little change was observed in body weight from the ages 6–18 months. Compared with this age range, a significant increase was observed in the range 26–32 months (left panel). Grip strength peaked at 12 months of age and then decreased with advancing age to a minimum at 32 months of age (center panel). Inclined plane performance was highest in young rats (6–12 months) and then declined with increasing age to a minimum at 26–32 months (right panel). The number of animals for each group is as follows: 6 months, 6; 12 months, 12; 18 months, 6; 26 months, 7; and 32 months, 6.

Effects of Age on L/G Behaviors

The first latency to lick or guard either hindpaw was evaluated for both cold (1.5°C) and heat (44.5°C) stimulations (Figure 2). Overall, latencies to the first reflex response to 44.5°C exceeded 3 minutes, and reflex latencies to 1.5°C averaged nearly 10 minutes for young animals. These values are considerably longer than those observed for escape responses to 44.5°C or 10°C (35). Between-group age effects on L/G latencies for 1.5°C stimulation were significant (F = 6.35, df = 3, p < .001), decreasing up to 27 months but reversing to longer latencies at 35 months (Figure 2, top). A post hoc test of the reversal from 27 to 35 months was significant (t = 3.92, df = 1, p < .001). L/G latencies did not differ significantly with age for 44.5°C stimulation (Figure 2, bottom; F = 0.72, df = 3, p = .54).

Figure 2.

First latencies to either lick or guard during exposure to 1.5°C (top) and 44.5°C (bottom). A decline in response latency from 11 to 27 months was observed for cold. This was followed by a dramatic increase in response latency (ie, decreased sensitivity) at 35 months. Little change in response latency was observed across different age-groups for 44.5°C.

Stimulus–Response Functions for Operant Escape Performance

Differential responding to nociceptive and non-nociceptive levels of stimulation provides important controls for changes in generalized activity and levels of physical performance as well as for development of avoidance tendencies with repeated testing. The stimulus–response functions in Figure 3 represent accumulated escape durations over the first nine responses in 15-minute trials, revealing clear discrimination between neutral (36°C) and nociceptive stimulus intensities for young and old animals. Throughout the study, animals of all ages infrequently escaped non-nociceptive 36°C stimulation (Figure 3A–D) but consistently escaped mildly nociceptive stimulation (15°C, Figure 3A and B; 42^°C, Figure 3C and D) or moderately nociceptive stimulation (10°C, Figure 3A and B; 44.5°C, Figure 3C and D). Within-group statistical comparisons of responses to 36°C, 15°C, and 10°C were significant at 8 months (Figure 3A; F = 39.99, df = 2, p < .001) and at 32 months (Figure 3B; F = 44.87, df = 2, p < .001). Within-group comparisons of sensitivity to 36°C, 42°C, and 44.5°C were significant at 8 months (Figure 3C; F = 61.02, df = 2, p < .001) and at 32 months (Figure 3D; F = 49.98, df = 2, p < .001). An important conclusion from the stimulus–response functions is that animals ranging in age from 8 to 32 months had no difficulty learning and performing the operant escape task used to assess thermal sensitivity. Furthermore, the low level of responsivity to 36°C at all ages evaluated indicates that increasing sensitivity to light in the escape compartment was not a concern with advancing age.

Thermal Escape as a Function of Temperature and Increasing Age

Age-related escape durations (ie, accumulated platform durations) for nociceptive temperatures are shown in Figure 4. Eight-, 16-, 24-, and 32-month-old animals were tested with the thermal plate at 15°C (Figure 4A), 10°C (Figure 4B), 42°C (Figure 4C), and 44.5°C (Figure 4D). For cold temperatures (10°C and 15°C), animals 32 months old were considerably more sensitive than the younger groups, accounting for a significant between-group effect of age (10°C: F = 1.41, df = 3, p < .001; 15^°C: F = 27.27, df = 3, p < .001). For nociceptive heat stimulation, hyperalgesia was evident for 24- and 32-month-old animals, relative to 8- and 16-month-old animals, producing an overall between-group significance for age (42°C: F = 3.79, df = 3, p = .02; 44.5^°C: F = 9.98, df = 3, p < .001). Animals 8 and 16 months of age did not differ in their sensitivity to nociceptive cold or heat stimulation. At the age of 24 months, thermal sensitivity had increased for all temperatures, and thermal sensitivity was greatest for animals 32 months of age.

Thermal Preference as a Function of Increasing Age

The thermal preference test (Figure 5) controls for age-related changes in physical performance and light aversion with advancing age. The two compartments did not differ in lighting or motoric demands. In tests of preference for cold (10°C) or heat (45°C) temperatures, sequential times on the cold plate were subtracted from times on the hot plate, and these difference scores were summed across successive pairs of hot and cold plate occupancies. The accumulated positive values in Figure 5 show that animals of all ages preferred 45°C stimulation instead of 10°C stimulation and a between-group ANOVA revealed a stronger preference for 45°C (ie, aversion to cold) with age (F = 5.48, df = 3, p = .002). As shown in Figure 4, older animals demonstrated a hyperalgesia for both heat and cold stimulation in the escape test, but the thermal preference test revealed that sensitivity to cold had a greater behavioral impact as a function of age than sensitivity to heat.

Effects of Inflammatory Injury on Escape Behavior

The effects of formalin injections on escape duration were averaged across 5 postinjection weeks for each temperature and for age-groups 8, 16, and 24 months. Due to nearly maximal baseline escape responding from thermal stimulation by the 32-month group (ie, a ceiling effect), their data are not shown. Figure 6 shows sums of the first three escape durations because animals with high sensitivity sometimes escaped no more than three times in a trial. Formalin injection significantly increased sensitivity to nociceptive thermal stimulation at 16 months (F = 35.8, df = 1, p < .001) and 24 months (F = 20.45, df = 1, p < .001) but not at 8 months (F = 0.01, df = 1, p = .93; within-group comparisons of baseline vs formalin across five temperatures). Thus, inflammatory injury had no effect on thermal sensitivity for the youngest animals (8 months), but inflammation summated with age to produce high levels of thermal sensitivity for the older animals (16 and 24 months) for both cold and heat temperatures.

Figure 6.

Effects of formalin injection in the dorsal hindpaw on escape durations for animals of different ages. Accumulated escape durations for the first three responses of each trial are shown. These values were averaged across five postinjection weeks for each temperature and age-group. Formalin injections significantly increased sensitivity to nociceptive thermal stimulation at 16 and 24 months but not at 8 months. The three age-groups along with temperatures tested are shown at the bottom of the figure. Post-formalin data were compared with data obtained pre-formalin (baseline).

DISCUSSION

Pain in the elderly individuals is a complex condition biologically, clinically, and therapeutically (15). Elucidation of relationships between physiological changes and pain sensitivity that occur during aging will benefit from behavioral methods of testing laboratory animals that translate to human studies of pain processing. Previous studies in rodents have attempted to relate age-dependent changes in reflexive behaviors (eg tail-flick, hindpaw withdrawal, and licking/guarding) to human pain sensitivity. However, L/G and other reflex responses are subserved by spinal or spinal–brain stem–spinal circuits in contrast to operant-based behavioral tasks that rely on processing of sensory information throughout the neuraxis, including the cerebral cortex (32,33,42). Cortically dependent strategies of behavioral assessment rely on neural structures important to the general perception of pain and measure the impact of peripheral stimuli on clinically relevant measures of cutaneous sensibilities (31–33,42).

In order for reflex tests to substitute for operant tests, a number of criteria must be met, including similar sensitivities of these methods to nociceptive stimulation. The present study demonstrated stimulus–response functions for operant responding to nociceptive cold and heat stimulation that are consistent with human psychophysical data (31–33). However, stimulus–response functions differ for operant escape versus reflex responding, especially for cold. Human pain ratings and escape responding of rats gradually increase from approximately 20°C to 0°C ((43); present study), but L/G responses are rarely observed above several degrees centigrade (35). Also, latencies for L/G responses to 1.5°C in the present study were considerably longer than those for escape latencies for either 10°C or 15°C. Eight-month-old animals escaped 10°C stimulation six times in 10 minutes, which was the latency to first response for the same animals to lick or guard in response to 1.5°C.

An important criterion for validity of reflex measures in the assessment of age effects on thermal pain sensations is that they should reveal effects similar to those shown by an escape task. Increasing sensitivity with age was observed in the present study for escape responding to heat stimulation, but there was no effect of age on L/G responding to 44.5°C. In the case of cold sensitivity, operant escape testing revealed increased sensitivity from 8 to 32 months. Similarly, decreased latencies for licking/guarding to a stimulus of 1.5°C were observed for animals progressing from 11 to 27 months of age (Figure 2). However, a significant reversal of behavioral effects was observed at 35 months. Thus, in direct comparisons with operant escape, reflex-based assessment tasks did not appropriately assess thermal sensitivity across the life span of the Long-Evans rat.

The escape test was not sensitive to muscle weakness or other motoric influences of aging because the primary measure of sensitivity was the relative occupancy of the escape compartment, regardless of the speed of movement between compartments. However, reflexive licking and guarding during 1.5°C stimulation may have been attenuated by physical impairments. Grip strength and endurance on an inclined plane were substantially impaired for these animals at 32 months of age. These deficits and associated physical impairments at 35 months appeared to have reversed a trend toward decreased reflex latencies of response to 1.5°C at 27 months. Cold sensitivity in aging humans has been evaluated in terms of pain detection thresholds or pain tolerance thresholds (eg, the cold pressor test). Threshold tests for cold pain detection involve brief activation of A-delta afferents and have revealed little or no effect of age (44,45). In contrast, thresholds for cold pain tolerance of elderly individuals (24,46–48) reveal an increased sensitivity to prolonged activation of cold nociceptors. The present study is consistent with the latter result and indicates that suprathreshold cold sensations are enhanced with aging. Escape responding to mildly nociceptive cold stimulation (10°C and 15°C) increased with advancing age (Figure 4). In addition, we evaluated whether hyperalgesia for cold and heat stimulation represented a generalized exaggeration of pain sensitivity or depended in part on the afferent channel stimulated. To this end, young and old animals were tested for their preference for cold or heat pain. The thermal preference test revealed increased occupancy of the hot plate relative to the cold plate with advancing age (Figure 5). One explanation for the greater aversion to cold may be that female rats are prone to autonomic dysfunction, with a high degree of peripheral vasoconstriction secondary to a hyperactive sympathetic nervous system (49).

Heat pain thresholds have been found to be higher for old people than for younger individuals in studies using brief stimulation with a fast onset (50–53). Rapid-onset heat stimulation with a contact thermode, radiant heat source, or laser elicits sensations dominated by input from A-delta nociceptors (50,54,55). Accordingly, a plausible explanation for elevated heat pain thresholds is that myelinated peripheral afferents are preferentially lost or disrupted with advancing age, including lightly myelinated A-delta nociceptors (4,56). This conclusion is supported by a demonstration that blockade of conduction in A-delta afferents eliminated differences in threshold for young and old individuals (50). Thus, it is possible that elevated heat pain thresholds of older humans are related to preferential activation of myelinated nociceptors by brief fast-onset stimulation. Also, elderly humans appear to employ a conservative approach to identify sensations as painful (57,58), which would drive pain thresholds of elderly individuals toward those of younger individuals. Parametric variations of nociceptive heat stimulation in human studies are needed to evaluate these possibilities.

In the present study, 42°C and 44.5°C stimulation from a stable heat source preferentially activated unmyelinated (C) afferents (34,55) and escape from long exposures to this stimulus was enhanced in older rats. This result is supported by demonstrations that temporal summation of heat pain, which depends upon repetitive activation of C nociceptors, increases with age for humans (53,59). Thus, pain from activation of C nociceptors, which is enhanced by inflammation and therefore during many forms of chronic pain, appears to be enhanced with age.

A potential mechanism for the increase in thermal pain sensitivity in the present study is that inflammation increases with age (60–62). Chronic inflammation sensitizes peripheral nociceptors, is a factor in central sensitization by activated microglia (63), and activates central stress circuits (64). One way to evaluate the influence of inflammation with age is to provide an inflammatory challenge. This was done by evaluating effects of formalin injury on thermal pain sensitivity for a testing period of 5 weeks. Secondary hyperalgesia was not observed for animals 8 months of age, but significant hyperalgesia was obtained for cold and heat stimulation of 16- and 24-month-old animals. The present results are consistent with those showing that paw injections of complete Freund's adjuvant in 18-month-old rats increased expression of the peptide dynorphin (DYN) in the spinal cords of 18-month-old rats compared with 3-month-old rats. Spinal DYN has been shown to be pronociceptive and its upregulation is required for the maintenance of neuropathic pain (65,66).

Evidence for similarities between aging and conditions associated with neuropathic pain (see introductory paragraphs) suggests that there could be similarities in the profile of behavioral changes during aging and after chronic constriction injury (CCI) of the sciatic nerve. Hypersensitivity to cold stimulation is characteristic of neuropathic pain models (36), and increased sensitivity to cold was detected by L/G and operant escape testing of older animals. Thus, similar peripheral and spinal abnormalities could underlie the effects of CCI and aging on reflex and operant responses to cold. However, operant responsivity to heat increased with age but not after CCI, and hypereflexia for heat was not observed with age (present results and (36)). Development of heat hyperalgesia with age is therefore likely to depend upon changes within supraspinal pain pathways.

In conclusion, operant escape from long-duration nociceptive cold and heat stimulations increases with age for female rats, in contrast to elevated pain thresholds for brief thermal stimulation of aged humans. The present results are consistent with a reduction in myelinated afferent input with age (ie, elevated thresholds) but sensitizing peripheral and central influences on sustained suprathreshold nociceptive input. In addition, hyperalgesia in response to an inflammatory challenge was greater for older animals that are presumed to have higher baseline levels of inflammation. Age-related hypersensitivity for nociceptive heat and cold stimulation was revealed by operant escape testing but not by more traditional reflex tests. Escape responding reveals the sum total of sensory processing throughout the neuraxis including a reflection of the sensory and affective components of pain, in contrast to a direct sensory–motor linkage of nociceptive reflexes dependent on spinal and spinal–brainstem–spinal circuits.

FUNDING

This work was supported by the National Institute on Aging with funds from RAG031821 (to R.P.Y.), the University of Florida Institute on Aging, and by the Claude D. Pepper Older Americans Independence Center (1 P30 AG028740).