Article Text

This article has a correction. Please see:

Download PDFPDF

Stair climbing/descending exercise for a short time decreases blood glucose levels after a meal in people with type 2 diabetes
  1. Hiroto Honda1,2,
  2. Makoto Igaki1,
  3. Yuki Hatanaka1,
  4. Motoaki Komatsu1,
  5. Shin-ichiro Tanaka1,
  6. Tetsuo Miki1,
  7. Taiga Suzuki2,
  8. Tetsuo Takaishi3,
  9. Tatsuya Hayashi2
  1. 1Toyooka Hospital Hidaka Medical Center, Toyooka, Japan
  2. 2Laboratory of Sports and Exercise Medicine, Graduate School of Human and Environmental Studies, Kyoto University, Kyoto, Japan
  3. 3Graduate School of Natural Sciences, Nagoya City University, Nagoya, Japan
  1. Correspondence to Professor Tatsuya Hayashi; tatsuya{at}

Statistics from

Key messages

  • Stair climbing–descending exercise (ST-EX) is a convenient method to increase physical activity without any special equipment.

  • Lean, moderately active older people with type 2 diabetes performed 3 min ST-EX 60 and 120 min after a meal.

  • ST-EX significantly hastened the decrease in postprandial blood glucose levels.


Postprandial hyperglycemia is recognized as an independent risk factor for cardiovascular events,1–3 and is highly prevalent throughout the day in people with type 2 diabetes (T2D), even among those with apparently good glycemic control according to their glycated hemoglobin (HbA1c) level.4 Physical exercise has been widely prescribed as part of the treatment of hyperglycemia, and recent studies have shown that high-intensity exercise (HIE) effectively improves postprandial glucose metabolism in people with T2D. Gillen et al5 demonstrated that 10 bouts of 60 s high-intensity cycling exercise after a meal reduced the postprandial peak of blood glucose (BG) level and the area under the curve (AUC) of BG in people with T2D. Karstoft et al6 demonstrated that premeal 1-hour interval walking in people with T2D (repeated cycles of 3 min of slow and fast walking) decreased the postmeal incremental BG levels. Francois et al7 examined the effect of six bouts of 1 min high-intensity incline walking (90% of maximal heart rate) before each meal in individuals with insulin resistance, and found that the postdinner and subsequent 24-hour BG levels were significantly improved. These studies have clearly shown the clinical benefits of HIE for the management of T2D.

On the other hand, the lack of time, the lack of access to an exercise facility and the perceived difficulty in performing exercise are important barriers to regular participation in physical activity in people with T2D.8 Thus, it would be desirable to develop a time-saving and non-strenuous way of HIE that assures substantial improvements in glycemic control regardless of weather conditions without the need for dedicated exercise equipment. To address this issue, we examined the acute hypoglycemic effect of two separate 3 min bouts of stair climbing–descending exercise (ST-EX), an easy-to-perform HIE in daily life. One can increase the overall exercise intensity without much effort by alternately climbing and descending stairs on a flight of stairs, because the subjective intensity is alleviated when descending the stairs.9–11



Sixteen Japanese people with T2D but no macrovascular or microvascular complications (13 men and 3 women, under the age of 75), who regularly visited Toyooka Hospital Hidaka Medical Center (Toyooka, Japan), volunteered for this study. Their clinical characteristics were determined in the outpatient clinic within 4 weeks prior to the experiment (table 1). All participants were under medical nutritional therapy (energy intake: 25–30 kcal/kg body weight/day) and exercise therapy (low-intensity to moderate-intensity aerobic exercise including walking, cycling and/or calisthenics for 20–60 min/day), which were discontinued on the experimental days. No participants regularly climbed stairs in their daily life. They were taking oral hypoglycemic agents (glimepiride, metformin, and voglibose (n=2); glimepiride and metformin (n=2); voglibose (n=2); alogliptin (n=2); glimepiride, miglitol, and vildagliptin (n=1); voglibose, nateglinide, and vildagliptin (n=1); metformin and voglibose (n=1); metformin and miglitol (n=1); voglibose and nateglinide (n=1); miglitol and alogliptin (n=1); metformin (n=1); sitagliptin (n=1)). No participant was taking β-blockers or antihypertensive drugs that affect heart rate responses to exercise. Written informed consent was obtained from all participants before the experiments. The Institutional Review Board of Toyooka Hospital Hidaka Medical Center approved the study protocol.

Table 1

Characteristics of study participants

Experimental protocols and analytical methods

This study was a cross-over design with allocation to two different interventions in random order. On separate days with an interval of 1–2 weeks, after an overnight fast from 21:00 the participants consumed a test meal for breakfast (E460F18; Kewpie, Tokyo, Japan) consisting of 19 g crackers, 62 g pudding, and 280 g chicken cream stew (56.5 g carbohydrate, 18 g protein, 18 g fat, 460 kcal) in 10–15 min between 7:00 and 8:00. The participants took medications including any oral hypoglycemic agents before or after breakfast in their usual way, as prescribed by their physicians. Then the participants sat on a chair for 180 min (REST session) or sat on a chair for 180 min except when they performed an ∼3 min bout of ST-EX 60 and 120 min after the meal (ST-EX session). Each bout of ST-EX comprised six repetitions of climbing and descending stairs. The participants climbed to the second floor of the clinic (21 steps, each 17 cm in height) at a rate of 80–110 steps/min, made a turn at the top of the flight, and then slowly walked back down the stairs to the first floor at a free step rate. The participants made a turn at the bottom of the flight, and repeated stair climbing and descending for a total of six times without rest. Their heart rate was recorded using a Polar Accurex Plus monitor (Polar Electro, Kempele, Finland). Borg's ratings of perceived exertion (RPE) scores12 were recorded immediately after the first and second ST-EX.

For the measurement of glucose, lactate, C peptide, and non-esterified fatty acid (NEFA), capillary blood samples (50–60 µL) were collected from a fingertip before (0 min) and 60 (immediately before the first ST-EX), 90, 120 (immediately before the second ST-EX), 150, and 180 min after the meal. Capillary blood samples were also collected after the first and second ST-EX for the measurement of lactate. Glucose and lactate concentrations were measured using a glucose analyzer (Glutest Ace; Arkray, Kyoto, Japan) and a lactate analyzer (Lactate Pro; Arkray), respectively. The blood samples were then centrifuged, and plasma was collected and stored at −20°C until C peptide and NEFA concentrations were measured using Ultrasensitive C peptide ELISA (Mercodia, Uppsala, Sweden) and LabAssay NEFA (Wako, Osaka, Japan), respectively.


All values are reported as the mean±SE. The time-course changes of BG, C peptide, and NEFA were analyzed with two-way repeated measures analysis of variance (ANOVA). For BG and C peptide levels, post hoc analysis was then performed with Tukey's test to assess differences between each data point. Differences in the AUCs between the two sessions and parameters in table 2 were analyzed using paired Student's t-test. The AUCs were calculated using the trapezoid method as follows: 0.5×δ(60)×60 min+(0.5×δ(60)(90)(120)(150)+0.5×δ(180))×30 min, where δ(n) is the increase from the premeal value (time=0) at the time point of n min. Significance was set at p<0.05.

Table 2

Profile and physical response to ST-EX


The BG at 60 min after the meal during the ST-EX session (immediately before the first ST-EX) did not differ from that during the REST session; however, it decreased more rapidly during the ST-EX session than during the REST session. The ANOVA revealed a significant interaction between time and treatment on BG (p<0.01; figure 1A). The BG at 150 min during the ST-EX session (30 min after the second ST-EX) was significantly lower than that during the REST session (p<0.01; figure 1A). The decreases in BG were also greater during the ST-EX session than during the REST session at any sampling period (p<0.05; figure 2). Furthermore, the AUC for BG (0–180 min) during the ST-EX session was 18% lower than during the REST session (ST-EX 428.1±67.0 mmol/L×min vs REST 521.0±50.6 mmol/L×min, p<0.05).

Figure 1

Time-course changes in blood glucose (BG; A), C peptide (B), and non-esterified fatty acid (NEFA; C) levels. Participants kept resting for 180 min except when performing each 3 min bout of stair climbing–descending exercise (ST-EX) at 60 and 120 min postmeal (ST-EX session), or kept resting for 180 min (REST session). There were significant interactions between time and intervention on BG (p<0.01) and C peptide (p<0.01), but not NEFA. Values are the mean±SE. **p<0.01 versus corresponding REST. †p<0.05, ††p<0.01 vs 60 min. N=16.

Figure 2

Changes in BG levels from 60 min (immediately before the first stair climbing–descending exercise (ST-EX)) to 180 min postmeal. Participants kept resting for 180 min except when performing each 3 min bout of ST-EX at 60 and 120 min postmeal (ST-EX session), or kept resting for 180 min (REST session). Values are the mean±SE. *p<0.05, **p<0.01 versus corresponding REST. N=16.

The ANOVA showed a significant interaction between time and treatment on C peptide (p<0.01), but there was no significant difference between sessions at any sampling time (figure 1B). The interaction was not significant in NEFA (figure 1C). The AUCs for C peptide and NEFA (0–180 min) were not different between sessions (C peptide: SE-EX 64.5±7.7 nmol/L×min vs REST 64.8±6.1 nmol/L×min, NEFA: ST-EX −28.1±4.2 mmol/L×min vs REST −29.9±6.9 mmol/L×min).

All the participants completed the ST-EX session without adverse clinical manifestations indicating cardiovascular, respiratory, or orthopedic complications. The percentage of age-predicted maximal heart rate (% HRmax; table 2) indicated that ST-EX was indeed an HIE, because activity is classified as ‘hard’ when the heart rate is between 70% and 89% HRmax.13 Moreover, the blood lactate level was robustly increased at the end of ST-EX (table 2). Nevertheless, the participants performed ST-EX without serious symptoms such as dyspnea or leg exhaustion, and the overall extent of physical effort estimated by the RPE for ST-EX was at the ‘moderate’ level.13


We reported previously that a continuous 6 min bout of ST-EX starting 90 min after a meal accelerated the decrease in postprandial BG levels in people with impaired glucose tolerance (IGT).9 We also demonstrated that a continuous 6.5 min bout of ST-EX starting 90 min after ingestion of a carbohydrate solution hastened the decrease in BG levels in people with T2D.11 However, we have realised that even the 6–6.5 min bout of ST-EX is too strenuous for some unfit people to perform regularly in daily life. Therefore, we conducted this study to clarify whether ST-EX for a shorter duration (∼3 min) would be sufficient to reduce the postprandial response in people with T2D. We employed two separate bouts of ST-EX after a meal, on the basis of our preliminary experiments, indicating that a single 3 min bout of ST-EX had a limited effect on BG levels (data not shown).

In the present study, we chose the timing for the first ST-EX (60 min after the meal) so as not to increase the BG levels again after exercise. Larsen et al14 ,15 demonstrated that moderate-intensity to high-intensity cycling exercise during the rapid rising phase of BG after a meal decreased the BG levels during exercise, but resulted in a rapid rebound after exercise. Similarly, we found in our preliminary experiments that ST-EX during the rapid rising phase of BG (<60 min) induced a rebound in BG levels after exercise (data not shown). As shown in figure 2, the first ST-EX was sufficient to hasten a decrease in the BG level after the meal. We added the second ST-EX (120 min after the meal) to boost this decrease during the declining phase in BG levels. As a result, our protocol clearly maintained a hypoglycemic effect up to 180 min (figure 2). However, it is possible that addition of ST-EX during the rapid rising phase (eg, at 30 min), after the first ST-EX (eg, at 90 min) and/or after the second ST-EX (eg, at 150 min), might improve postprandial hyperglycemia more significantly than the current protocol.

An acute bout of exercise is a physiologically relevant stimulus to promote glucose uptake in skeletal muscle. Exercise exerts an insulin-like effect on contracting skeletal muscle by stimulating translocation of the glucose transporter 4 (GLUT4) to cell surface membranes independently of insulin.16 ,17 The rate of glucose uptake increases rapidly (<5 min) after the initiation of exercise, and depends on the intensity of exercise performed.18 ,19 It has also been shown that the rate of glucose uptake reaches a near-maximal level (∼80% of the maximum) within 10 min after the start of exercise.20 Importantly, exercise-induced GLUT4 translocation and glucose uptake in skeletal muscle is intact in people with T2D.21–23 Thus, it seems reasonable to speculate that a 3 min bout of ST-EX is sufficient to substantially increase glucose uptake in working muscles. In addition, the postexercise period has been characterised by enhanced insulin sensitivity in skeletal muscle, which leads to a prolonged increase in insulin-stimulated glucose uptake.24 These insulin-dependent and insulin-independent mechanisms might contribute to the acute hypoglycemic effect of ST-EX in people with T2D.

C peptide and NEFA levels often reflect the intensity and/or duration of exercise. Larsen et al14 ,15 demonstrated that moderate-intensity to high-intensity cycling exercise for 45 min after a meal acutely blunted the postprandial increases in insulin and C peptide levels and postprandial decreases in NEFA levels in people with T2D. The reduction in insulin secretion might occur in response to the decreased BG level induced by exercise, as well as from the induction of a counter-regulatory response such as increased sympathetic nervous activity during exercise,25 which also causes NEFA release.26 In this study, our exercise protocol did not change NEFA levels (figure 1C), suggesting strongly that it does not elicit a clinically unfavorable counter-regulatory effect, and that insulin secretion might have been slightly reduced as a secondary response to the decreased BG levels induced by ST-EX (figure 1A, B).

It is notable that falls on stairs are common among older and/or obese people, particularly those with orthopedic disorders, and that stair descent seems to be more hazardous than stair ascent despite the low exercise intensity during descent. A survey of falls in community-living older people at least 75 years of age demonstrated that 80% (20/25) of falls on stairs occurred during descent.27 Aging is accompanied by deteriorations in musculoskeletal capacity for safe stair descent, such as decreased eccentric strength at the knee and ankle. In people with diabetes, visual dysfunction due to retinopathy may also increase the risk of accident. Furthermore, neuropathy is often accompanied by deterioration in sensory, motor, and autonomic nervous systems that may be critical to safe locomotion on stairs. Thus, the use of ST-EX in practice should be individualised with careful consideration of the risk of falling, particularly during descent.

There were some limitations to this study. First, the study participants were lean, maintained optimal HbA1c values, and already under exercise therapy, so the results might not be directly applicable to obese, poorly controlled, or sedentary patients. In particular, physical fitness and body mass index are closely linked in people with T2D: the heaviest people are the least fit,28 and thus the decrease in exercise intensity during the descending phase of ST-EX might not be generalised to severely obese patients. Second, no comparison of the exercise intensity between ST-EX and other exercise modalities such as level walking was made. In this regard, we reported previously that a single bout of ST-EX for 6–6.5 min was more effective than level walking for the same duration in decreasing postprandial BG in people with IGT9 and T2D.11 Third, an ST-EX protocol that could attenuate the increase in BG after a meal and blunt the peak postprandial BG, which should be lower than 10 mmol/L,29 remains to be determined. Thus, further studies are needed prior to clinical applications of this regimen to a variety of patient populations.


In lean, moderately active older people with T2D, 3 min ST-EX starting 60 and 120 min after a meal hastened the decrease in postprandial BG levels without the need for excessive effort. Although further research is required, ST-EX may be a potentially useful method for increasing physical activity in daily life, with efficient and acute postprandial BG reduction in people with T2D.


The authors thank the study participants for their time and effort.


View Abstract

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Correction
    American Diabetes Association