Walking

What it is

Most cancer trials used moderate-intensity walking [1]. Trials also used brisk walking and daily step goals, such as 10,000 steps a day in pregnancy [2][3]. Blood sugar studies used everyday walking: 10-minute walks after main meals, walking in place of sitting, and light 3.2 km/h pace breaks [4][5].

What the research says

Walking after meals lowers the post-meal blood sugar rise. A 2023 meta-analysis pooled eight crossover trials (116 people, 47 with type 2 diabetes) and found post-meal exercise, mostly walking, cut post-meal glucose excursions compared with sitting (standardised effect 0.55) and compared with exercise before the meal (0.47); the sooner the walk started after eating, the bigger the effect, and the review rated the trials at high risk of bias [6]. A 2026 meta-analysis of 39 crossover trials found activity breaks lowered post-meal glucose and insulin, with walking breaks having the largest effect of any activity type and the biggest reductions when sitting was interrupted every 15 to 20 minutes [7]. In seven one-day crossover trials, short light walks lowered post-meal glucose (change of -0.72) and insulin (-0.83) compared with staying seated, and were better than standing breaks [8]. A 2026 meta-analysis of 29 study cohorts (573 people) found short walking bouts lowered post-meal glucose and insulin areas, while standing alone did not [9].

In type 2 diabetes, advice to walk 10 minutes after each main meal lowered the post-meal glucose area by about 12% compared with 30 minutes of walking once a day, with a 22% drop after the evening meal, in a two-week crossover study of 41 adults [4]. Three-minute light walks (3.2 km/h) every 30 minutes during prolonged sitting cut post-meal glucose (area 14.8 vs 24.2 mmol x h/L) and, in the same participants, 22-hour and overnight glucose stayed lower, remaining lower until the next morning [5][10]. In 10 people taking metformin, post-meal treadmill walking lowered the breakfast glucose peak from an average of 12.0 to 9.7 mmol/L when combined with metformin [11]. A review of 16 randomised trials (1,820 people with type 2 diabetes) found walking alongside diabetes medication lowered fasting glucose by about 12 mg/dL and HbA1c by 0.35% compared with no regular exercise [12].

In healthy young adults, a 10-minute walk straight after a sugary drink lowered the 2-hour glucose area (15,607 vs 16,605 mg x min/dL) and the peak (164 vs 182 mg/dL) compared with resting [13]. A series of crossover studies found post-meal but not pre-meal exercise improved glucose, 10 to 20 minutes of accumulated walking was enough, and starting 15 minutes after the meal blunted the peak best [14]. In free-living young women, three brisk post-meal walks lowered 3-hour post-meal glucose, most after dinner (6.71 vs 7.55 mmol/L), but not 24-hour glucose [15].

Results are less clear in pregnancy and over the long term. A single post-meal walk lowered glucose compared with sitting in women with gestational diabetes [16], but a 7-week trial of 40 women found advice to take three 10-minute post-meal walks did not improve post-meal glucose beyond standard advice to walk 30 minutes continuously [17], and a shorter crossover trial found the two approaches about equal [18]. A 12-week randomised trial in 56 people with hyperglycaemia or high risk of it found no change in any glucose measure from daily post-prandial activity [19].

In people with cancer, a 2015 meta-analysis of nine randomised trials (599 people) found walking significantly improved sleep, with a standardised effect of 0.52 [1]. In a 12-week trial of 111 people with lung cancer, home walking improved the Pittsburgh Sleep Quality Index and total sleep time at 3 and 6 months [20].

In adults with insomnia, a 2026 network meta-analysis of 22 trials (1,348 people) found walking or jogging may reduce insomnia severity scores by about 9.6 points on average; the certainty of that evidence is low [21]. In a 24-week trial of 75 students with insomnia, brisk walking lowered Pittsburgh scores more than the control group did [2].

Across 22 trials of exercise in healthy adults, self-rated sleep quality improved by 2.19 points on the Pittsburgh scale; few measured sleep outcomes changed [22]. Exercise also reduced insomnia severity by about four points in a 6-month trial of 41 inactive adults with insomnia [23].

In a 4-week study of 59 adults, days when people were more active than their own average came with better self-reported sleep quality and duration; this is an association, not proof of cause [24].

Doses used in studies

Blood sugar: 10 minutes of walking after each main meal (three times a day) in type 2 diabetes and gestational diabetes [4][18][17]; a 10-minute walk right after a glucose drink in healthy young adults [13]; 10 to 30 minutes after the meal, starting within about half an hour, in healthy adults [14][6]. Walking breaks: 3 minutes of light walking (3.2 km/h) every 30 minutes during prolonged laboratory sitting [5][10]; short walking bouts of under 10 minutes, with the largest effects when repeated every 15 to 20 minutes [7][9]; 5 x 10-minute treadmill walking bouts at 60% VO2max in people on metformin [11]. Sleep studies: brisk walking 60-minute sessions three times a week for 24 weeks [2]; home-based walking with weekly counselling for 12 weeks in lung cancer [20]; pedometer goal of 10,000 steps a day in pregnancy [3]; at least 150 minutes a week of moderate-to-vigorous activity for 6 months in inactive adults with insomnia [23].

Safety and interactions

Blood sugar medicines: post-meal walking lowers glucose, so people taking insulin or sulfonylureas (medicines that can cause low blood sugar) should ask their care team whether doses need adjusting; in a small trial of 8 adults with type 1 diabetes, 30 minutes of walking an hour after a meal lowered glucose during exercise and more overnight low-glucose events followed exercise than rest, and some people also needed insulin correction for high glucose right after exercise [25].

Side effects: the abstracts of the walking trials cited here do not describe injuries or other harms [1][2][20]. Upper limits: no upper limit or toxic amount of walking was found in these studies [1][21]. Medicines: none of the abstracts reports an interaction with sleep medicines or other drugs [20][2]; people taking sleep medicines should check with a doctor before changing their activity. Pregnancy and breastfeeding: in a trial of 270 pregnant women, a walking programme did not prevent insomnia, and the abstract does not mention harms [3]; a walking programme did not lower post-meal glucose beyond standard care in a trial of 40 pregnant women [17]. No breastfeeding data were found in the studies read. Conditions: the trials were in people with cancer, adults with insomnia, students, pregnant women, healthy adults, people with type 2 diabetes and people with high blood sugar [1][21][2][3][4][9]. People with heart, joint or other health conditions should ask a doctor before starting a new routine.

Caveats and open questions

The blood sugar trials are small and acute: the 2023 timing meta-analysis pooled only 116 people across eight crossover trials and rated them at high risk of bias [6], and a 12-week randomised trial of daily post-prandial activity found no change in blood glucose [19]. Lowering a single post-meal glucose spike is not the same as preventing diabetes complications. In gestational diabetes, three 10-minute post-meal walks did not beat standard 30-minute continuous walking over 7 weeks [17].

The walking trials cited here are small: nine cancer trials totalled 599 people, the lung cancer trial had 111 and the student trial had 75 [1][20][2]. The 2026 network meta-analysis rated its evidence low certainty, and 68% of its 22 trials had some concerns about risk of bias [21]. A 2023 meta-analysis found no significant overall effect of physical activity on sleep quality [26]. In older adults, walking alone is not among the regimens the abstract reports as significantly better than usual care [27]. The same 2026 analysis ranks walking or jogging with yoga and Tai Chi as more effective than other exercises, again with low-certainty evidence [21].