Modulation of the human immune status by spinal thermal massage: a non-randomized controlled study

2.1 Participants

One hundred and forty individuals above 60 years of age with a disability caused by damage to the central nervous system for more than 6 months were recruited for the study. All recruited patients had symptoms of neck, shoulder, buttock, or knee pain and back muscle pain for more than 3 months. As the medical history record of each patient is important, patients with a stroke and with similar histories were included, and these patients were expected to have a reduced immune function because they cannot maintain normal activities for a long time due to central nervous system damage for more than 6 months. Participants with an immunological disease, a cardiac pacemaker, a transplanted electrical device, a malignant tumor, a spinal cord infection, thrombosis, a suspected skin disease, skin irritation, a vertebral deformity, scoliosis with Cobb’s angle of 20${}^{◦}$ or more, a history of vertebral fractures due to osteoporosis, severe osteoporosis, vertebral fractures, myopathies, and back syndrome after spinal surgery were excluded from the study. Therefore, 130 participants were enrolled after the exclusions. Of the 130 participants, 65 were assigned to the control group receiving rehabilitative regular care (RRC) and 65 were assigned to the experimental group receiving STM and RRC. RRC mainly consists of occupational therapy performed by occupational therapists to improve upper extremity and daily life functions, and physical and occupational therapies consisting of balance and gait training and gait posture to improve lower extremity functions. It was not associated with the improvement in the immune function and it was ethically necessary. During the 1-week adaptation process, four participants were excluded from the experimental group. During treatment, 4 participants from the control group and 6 from the experimental group were excluded, and a total of 116 participants completed the treatments. However, 57 blood samples were used for laboratory calibration to establish assay conditions for 15 immunological measurements, such as appropriate cell types, dilution rate with medium for whole blood culture, period of culture, concentration of stimulants, and dilution rate of samples for enzyme-linked immuno-sorbent assay. Finally, data derived of 29 control subjects (14 males and 15 females) and 30 experimental subjects (20 males and 10 females) were analyzed (Fig. 1). The mean duration of pain ( $\pm$ standard error [SE]) was 28.10 months (SE, $\pm$ 28.97 months) for the control group and 15.70 months (SE, $\pm$ 16.71 months) for the experimental group.

Fig. 1.Procedure followed in the study. One hundred and forty individuals applied to participate in this study, but 10 individuals dropped out in the screening stage; therefore, 130 individuals were assigned to two groups (Exp. n = 65, Con. n = 65). During the 1-week adaptation process, four participants were excluded from the experimental group. During treatment, 4 participants from the control group and 6 from the experimental group were excluded, and a total of 116 participants completed the treatments. fifty-seven blood samples were used for laboratory calibration to establish assay conditions for nine immunological measurements. Ultimately, data of 30 individuals in the experimental group and 29 in the control group were analyzed. RMI, Rehabilitative Medical Indicators; CLS, Changes in Leukocyte Subpopulations; CCP, Changes in Cytokine Production.

Written informed consent was obtained from all subjects before the study. This study adhered to the principles and recommendations of the Declaration of Helsinki. The study protocol was reviewed and approved by the Institutional Review Board at the Presbyterian Medical Center (IRB no. 2017-06-022).

2.2 STM treatment

This study was conducted using a personal thermal device for relieving muscle pain (CGM MB-1401; Ceragem Co., Ltd., Cheonan, Korea), an instrument approved by the Food and Drug Administration of Korea and the United States as a personal thermal device for relieving muscle pain. The STM device provides a 36.5-min preprogrammed chiropractic massage sequence. During this sequence, a roller massages the muscles along the spine by moving cranially, caudally, and vertically for paraspinal massage and applies segmental acupressure at the basic strength presented in the program algorithm (Fig. 2). For thermal massage, heat is added to the program using a heating source inside the roller that can be set at 30 ${}^{\circ }$C–60 ${}^{\circ }$C. However, as the temperature suggested for the induction of immune response in previous studies related to hyperthermia treatment is 39 ${}^{\circ }$C–41 ${}^{\circ }$C [20], it was set to 43 ${}^{\circ }$C considering the clothes the users were wearing. A trial session allowed participants to become familiar with the massage device.

Fig. 2.Structural image of the spinal thermal massage device. A roller massages the muscles along the spine by moving cranially, caudally, and vertically for paraspinal massage and applies segmental acupressure at the basic strength presented in the program algorithm.

2.3 Study design

Data were collected and blood was drawn at the following three time points: before the massage, 4 weeks after the massage, and 8 weeks after the massage. Therefore, we conducted a non-randomized controlled study with repeated measures.

2.4 Rehabilitative medical indicators

Questionnaires such as the pain numeric rating scale (PNRS), the Oswestry disability index (ODI), Roland and Morris disability questionnaire (RMDQ), and 3-level EuroQol Five-Dimensional Questionnaire (EQ) were used to obtain data regarding pain and the quality of life of the participants.

2.5 Source of blood samples

Blood samples for the culture and immunofluorescence assay were collected in evacuated Vacutainer blood collection tubes containing 75 USP sodium heparin (ref. 367871) (Becton Dickinson and Co., Franklin Lakes, NJ, USA). Blood for serum samples was drawn into BD Vacutainer SST II Advance blood collection tubes (ref. 367955) (Becton Dickinson and Co.). The samples were collected at approximately 9:00 AM and processed immediately.

2.6 Quantitation of leukocyte subpopulations and flow cytometry

Total white cell blood counts and differential counts (lymphocytes and monocytes) were estimated from whole blood using an autoanalyzer (NX-3000; SYSMEX Co., Kobe, Japan). Immunofluorescence was performed for the quantitation of lymphocyte subpopulations using whole blood and monoclonal antibodies according to the manufacturer’s instructions. The antibodies (BioLegend, San Diego, CA, USA) against leukocyte surface molecules used in this study were FITC-anti-CD3 (clone UCHT1), FITC-anti-CD4 (clone RPA-T4), PE-anti-CD8a (clone RPA-T8), FITC-anti-CD19 (clone HB19), and PerCP-anti-CD56 (clone HCD56). The erythrocytes were lysed using red blood cell lysis buffer (BioLegend) and flow cytometric analysis was performed using an EPICS V flow cytometer (Coulter Corp., Hialeah, FL, USA).

2.7 Blood culture

Iscove’s modification of Dulbecco’s medium (IMDM) (GIBCO Life Technologies, Grand Island, NY, USA) was supplemented with 1% antibiotic/antimycotic solution (penicillin 100 units/mL, 100 $\mu$g/mL streptomycin, and 250 ng/mL amphotericin B; Sigma Chemical Company, St. Louis, MI, USA), 5 mM 2-mercaptoethanol, and 200 mM L-glutamine (Sigma Chemical Company). Heparinized blood was mixed with supplemented IMDM at a ratio of 1:10. The blood suspension was incubated with purified anti-CD3 (1 $\mu$g/mL) and anti-CD28 (1 $\mu$g/mL) antibodies (Miltenyi Biotec GmBH, Bergisch Gladbach, Germany) for T cell stimulation, and 100 ng/mL purified human interleukin (IL)-2 (Miltenyi Biotec GmBH) or 1 $\mu$g/mL lipopolysaccharide (LPS) (Sigma Chemical Company) for NK cell or inflammatory cell stimulation. After stimulation for 1 or 3 days in 48-well culture plates (Nunc, Roskilde, Denmark) at 37 ${}^{\circ }$C and with 5% CO${}_2$, the supernatants were recovered and frozen at -80 ${}^{\circ }$C until analysis.

2.8 Enzyme-linked immunosorbent assay

The cytokine levels in the culture supernatants were measured using an enzyme-linked immunosorbent assay set (MAX Deluxe Set; BioLegend) according to the manufacturer’s instructions.

2.9 Statistics

To control the difference between the groups at baseline, the data were processed using the analysis of covariance (ANCOVA) using the baseline value as a covariate. Two main effects (group and time effects) and an interaction effect (group $\times$ time) were analyzed using ANCOVA, and post-hoc analysis was performed to evaluate the differences between the groups at each instance of measurement to check if any of the effects were significant. Data are presented as mean and standard error adjusted for the baseline. The criterion for statistical significance ($\alpha$) was 0.05.

3.1 Pain and quality of life

First, we examined the pain and quality of life perceived by the participants using four self-reported tests. For all measurements, the ANCOVA was performed using the baseline value as a covariate. In PNRS, the group effect was significant (p $<$ 0.01), but the time and interaction effects were not. The PNRS score in the experimental group was lower than that in the control group after both 4 weeks (p $<$ 0.01, ES = 1.08) and 8 weeks (p $<$ 0.01, ES = 1.22). The ANCOVA for ODI showed statistical significance in the group, time, and interaction effects (p $<$ 0.01). The ODI score was lower in the experimental group than in the control group after both 4 weeks (p $<$ 0.05, ES = 0.69) and 8 weeks (p $<$ 0.01, ES = 1.08). In RMDQ, the group effect was significant (p $<$ 0.05), but the time and interaction effects were not. The RMDQ score was lower in the experimental group than in the control group after 4 weeks (p $<$ 0.05, ES = 0.54) and 8 weeks (p $<$ 0.01, ES = 0.73). In EQ, the group (p $<$ 0.01) and interaction effects (p $<$ 0.05) were significant, but the time effect was not. The EQ score was higher in the experimental than in the control group after both 4 weeks (p $<$ 0.01, ES = 0.73) and 8 weeks (p $<$ 0.01, ES = 0.94) (Table 1, Fig. 3).

Fig. 3.Changes in pain-related measures. Values are expressed as mean (standard error) adjusted for the baseline value. Asterisks indicate significant group differences each time points (*p $<$ 0.05, **p $<$ 0.01).

3.2 Changes in leukocyte subpopulations

The changes in the count of peripheral blood leukocytes and the distribution of their subpopulations were evaluated for CD3${}^{+}$, CD4${}^{+}$, and CD8${}^{+}$ T cells, CD19${}^{+}$ B cells, and CD56${}^{+}$NK cells, which are reportedly low in elderly individuals [21, 22, 23, 24, 25, 26]. The ANCOVA for leukocyte, lymphocyte, CD3${}^{+}$ T cell, CD4${}^{+}$ T cell, CD8${}^{+}$ T, and CD19${}^{+}$ B cell showed that there was no significant group, time, and interaction effects. For monocytes and CD56${}^{+}$ NK cells, the group effect was significant (p $<$ 0.05), but the time and interaction effects were not. Monocyte and CD56${}^{+}$ NK cell proportions were higher in the experimental group than in the control group after 8 weeks (p $<$ 0.05, ES = 0.68, 0.53, respectively) (Table 2, Fig. 4).

Fig. 4.Changes in leukocyte subpopulations. Values are expressed as mean (standard error) adjusted for the baseline value. Asterisks indicate significant group differences each time points (*p $<$ 0.05, **p $<$ 0.01).

3.3 Changes in cytokine production

The changes in cytokine production by T cells were evaluated; specifically, the IL-2, IL-4, and IFN-$\gamma$ levels were evaluated based on the findings of a previous study, which reported that their levels are low in elderly individuals [27]. For IL-2 production, the interaction effect was statistically significant (p $<$ 0.05), but the group and time effects were not. The IL-2 level was lower in the experimental group than in the control group after 4 weeks (p $<$ 0.05, ES = 0.56). There was no significant effect on IL-4 and IFN-$\gamma$ production.

Based on previous reports indicating that NK cell activity declines in elderly individuals [25, 28], the changes in IFN-$\gamma$ production following NK cell stimulation by IL-2 were measured. Although there was no significant effect on the production of IFN-$\gamma$, the group effect was close to the significance level (p = 0.058). The IFN-$\gamma$ level was higher in the experimental group than in the control group after 4 weeks (p $<$ 0.05, ES = 0.61) (Table 3, Fig. 5).

Fig. 5.Changes in cytokine production. Values are expressed as mean (standard error) adjusted for the baseline value. Asterisks indicate significant group differences each time points (*p $<$ 0.05, **p $<$ 0.01).

An increase in the production of proinflammatory cytokines by stimulated mononuclear cells has been observed in elderly subjects, and it may be related to several features of age-associated pathological events [29]. In this study, the production of tumor necrosis factor (TNF)$\alpha$, IL-1$\beta$, and IL-6 after LPS stimulation of whole blood cells was also analyzed. For TNF$$ secretion, the interaction effect was statistically significant, but the group and time effects were not. This resulted from a significant reduction in the level of TNF$\alpha$ in the control group at 8 weeks compared with that at 4 weeks. There was no significant effect on IL-1$\beta$ and IL-6 production.