Extended Abstract
1. Introduction
Physical health of elderly people is very important. Knee deformities such as genu valgum in the elderly cause pain and instability of the patellofemoral joint [1]. The main factors in the treatment of lateral patellar dislocation include activity modification, and physiotherapy for stretching and correction of quadriceps [2]. The most commonly used corrective method for people with knee deformities is the corrective exercise program. This exercise program focuses on strengthening the weak muscles in all three lower extremity joints. To our knowledge, there is no study on examining effect of corrective exercise programs on the frequency spectrum of Ground Reaction Force (GRF) during double-leg drop-landing task in older adults with genu valgum. In this regard, this study aimed to examine the effect of a corrective exercise protocol on the GRF frequency spectrum of these patients during a double-leg drop-landing task.
2. Materials and Methods
The is a quasi-experimental study (a registered clinical trial with code: IRCT2016110230657N1). Participants were 26 elderly men with genu valgum (13 in intervention group with a mean age of 63.92±2.21 years and 13 in control group a mean age of 64.01±2.62 years). The GRF data were measured by two force plates (Kistler Group, Winterthur, Switzerland). Data sampling rate of the force plate was at 1000 Hz. In the pre-test phase, the subjects were asked to perform five correct double-leg drop-landing task.
 A 4th order Butterworth low pass filter with a cutoff frequency of 5 Hz was used for GRF data. After filtering vertical, anterior-posterior, and mediolateral GRF data in the time function, they were converted from time function to frequency function using Fourier transform or harmonic analysis in MATLAB software. In the first two weeks, stretching exercises were performed for the muscles of hip adductor, biceps femoris and tensor fasciae latae in four 3-s sets for each movement [3]. After the 2-week stretching protocol, the subjects performed resistance training with Theraband for 14 weeks, 3 sessions per week. Each training session consisted of 10-min warm-up, 35-40 min resistance training, and cooling down in the end. Subjects were asked not to use any exercise programs or orthodontic tools during training sessions. After completing the intervention, post-test measurements similarly to the pre-test phase were carried out. The control group did not participate in any exercise program during the study period and were evaluated only in pre- and post-test phases. Two-way repeated measures ANOVA was used for analyzing data.
3. Results
Results of Shapiro-Wilk test reported the normality of data distribution in the variables of GFR frequency. The results did not show a statistically significant difference between the control and experimental groups in most of the dependent variables in the pre-test phase, indicating that the groups were homogenous before intervention. Based on the ANOVA results, the effect of time factor on the frequency with a power of 99.5% in vertical GRF (P=0.003) and the number of essential harmonics in anterior-posterior direction (P=0.006) was significant in the left leg; however, its effect on frequency bandwidth and mid-frequency in the left leg in three directions of vertical, anterior-posterior, and mediolateral was not statistically significant (P>0.05). The effect of group factor on the 99.5% frequency power (P=0.006) and the number of essential harmonics of vertical GRF (P=0.047) in the left leg was also significant; however, its effect on frequency bandwidth and mid-frequency in the left leg in three vertical, anterior-posterior, and mediolateral directions was not statistically significant (P>0.05). The interaction effect of group and time on the 99.5% frequency power (P<0.001) and the number of essential harmonics of vertical GRF (P=0.001) in the left leg was also significant. The results of paired t-test showed that the number of essential harmonics in vertical GFR in the left leg during drop-landing task in the intervention group was significantly greater in the post-test phase than in the pre-test phase (P=0.018). No significant difference was found in other variables (P>0.05). 
In the right leg, no significant effect of time, group, and interaction effect of time and group on 95.5% frequency power, number of essential harmonics, frequency bandwidth and mid-frequency of GFR was observed in three vertical, anterior-posterior, and mediolateral directions (P>0.05) (Table 1). Moreover, the results of paired t-test showed no significant difference between groups in GRF frequency spectrum during a double-leg drop-landing task before and after intervention in the right leg (P>0.05) (Table 1). The odds ratios for the effect of time, group, and interaction effect of time and group on components of GRF frequency spectrum were low in the right leg. Despite the similarity of the corrective exercise program for both right and left legs, the effect of intervention on GRF frequency spectrum in the left leg was higher than in the right leg.

4. Conclusion
In the intervention group, the number of essential harmonics of vertical GRF in the left leg of older adults with genu valgum during the drop landing was significantly higher in the post-test phase than in the pre-test phase. However, the GRF frequency spectrum did not change in both mediolateral and anterior-posterior directions after the corrective exercises. Despite the similarity of the corrective exercise program for both right and left legs, the effect of intervention on GRF frequency spectrum in the left leg was higher than in the right leg of older adults with genu valgum. Further studies are recommended for evaluating electrical activity of muscles in elderly men and women with knee osteoarthritis before and after a corrective exercise program. Moreover, it is suggested that seniors perform corrective exercises to reduce the severity of the damage caused by the GFR frequency spectrum during tasks such as drop landing.

Ethical Considerations
Compliance with ethical guidelines
The research project was approved by the Ethics Committee on Medical Studies of Ardabil University of Medical Sciences (IR.ARUMS.REC.1395.77) and was registered at the Iranian Clinical Trial Site (IRCT2016110230657N1). For the sake of ethical considerations, while familiarizing subjects with all stages of research, such as exercise programs and laboratory evaluations, they could be excluded if desired at any stage of the study. In addition, throughout the laboratory phase and training, a trained person was with the subjects to prevent falls or possible injury.
Funding
The study received no financial support from government and nonprofits.
Conflict of interest
According to the authors, there is no conflict of interest between them.
Authors' contributions
According to the authors, there is no conflict of interest between them.
Conflicts of interest
The authors declared no conflict of interest.