Table 1.

Characteristics of included studies

Study ID	Country / number of participants	Intervention	Comparator	Duration	Outcome measures	Summary
Heat
Akin 2004	USA / 344 women	Active heat wrap (40 degrees C) for eight hours	Oral acetaminophen (paracetamol) 500 mg 4 times per day for one day	1 day, starting on the first day of moderate or greater menstrual pain	Pain relief (categorical 0–5 score, 0 = no pain relief, 5 = complete pain relief) during the eight hours Adverse events	Heat was superior to oral acetaminophen in reducing menstrual pain. One mild AE in the heat group (pink skin)
Akin 2001	USA / 81 women	Heat patch (180cm2, 38.9 °C) for 12 h each day, plus placebo pills.	3 comparator groups: 1- Heat patch (180 cm2, 38.9 °C) plus ibuprofen 400 mg, 3 times per day 2- Unheated patch (180 cm2) plus ibuprofen 400 mg, 3 times per day 3- Unheated patch plus placebo pills	2 days – from onset of menstrual pain	Pain relief (categorical 0–5 score, 0 = no pain relief, 5 = complete pain relief) during the 12 h Adverse events	Heated patch plus ibuprofen, heated patch plus placebo pills and unheated patch plus ibuprofen all showed similar pain reductions. The time to onset of pain reduction was shorter in the heated patch plus ibuprofen group compared to the unheated patch plus ibuprofen. No adverse events reported
Potur 2014	Turkey / 193 women	Heat patch (180cm2, 38.9 °C, 8 h)	2 comparator groups: 1. Self-medication group (SMG) – single dose of analgesics of participants choice 2. Control group – no medication but could use single dose of oral analgesic if pain was unbearable.	1 day, first day of menses	Pain intensity (VAS) at 4 h and 8 h after starting treatment	Quasi-randomised based on presentation order. The group receiving the heat patch reported the greatest pain reduction at 4 h and at 8 h after the start of the intervention.
Rigi 2012	Iran / 150 women	Heat patch (84 cm2, 40 °C, 8 h)	Ibuprofen (400 mg) every 8 h as needed	1 Day. First day of menses, once menstrual bleeding started	SF-MPQ at 24 h after starting treatment.	Total pain score as measured by SF-MPQ was not significantly different between groups at 24 h after onset of menses. There was no pre-post comparison. Fchar
Acupressure
Aghamiri 2005	Iran / 100 women	Acupressure at unspecified points on the abdomen, near the waist and in the leg. 15 min pressure, 15 min rest, then 15 min pressure	Sham acupressure: Same as active group but using ‘off channel’ points.	Frequency and timing of treatment not reported	Pain intensity (VAS) at 180 min following treatment.	Article translated from Farsi. Acupressure group had significantly lower pain scores at 180 min post treatment than sham.
Behbahani 2016	Iran / 120 women	Acupressure at CV4 and CV2. Pressure was applied for 15 s and then 15 s of rest for a total of 20 min.	Two comparator groups: 1. Self-care behaviours. Four 60–90 min sessions, once per week. Isometric exercises were also taught. 2. Ibuprofen group: 400 mg, three times per day.	Acupressure applied over the first two days of menses for two cycles. Comparator groups: Unclear about timing in relation to menses.	McGill pain questionnaire (MPQ) at the end of the 2 month intervention.	Both training and acupressure groups had greater reductions in pain than the ibuprofen group, however no between group comparisons were performed.
Bazarganipour 2010	Iran / 194 women	Acupressure at LR3. Two minutes of firm pressure then two minutes of ‘massage’ on the point. Twenty minutes in total per session. Unclear if alternating sides were used.	Sham Acupressure. Point located at the same level as LR3 but between the 3rd and 4th toes.	Starting three to seven days before menses, for two menstrual cycles.	Andersch and Milsom Scale (0 = no inhibition by pain 3 = work clearly inhibited by pain) during the first menstrual cycle post intervention.	Pain scores decreased in both groups from baseline but the reduction was significantly greater in the acupressure group
Charandabi 2011	Iran / 72 women	Acupressure at SP6. Four seconds of pressure and two seconds without pressure, for five minutes. After five minutes the alternate leg was used, for a total of 20 min per day (4 × 5 min cycles).	Ibuprofen (400 mg) every 6 h as needed	Starting from onset of menstrual pain for two menstrual cycles.	Menstrual symptom severity (5 point Likert scale) for eight symptoms (cramp, headache, back pain, leg pain, depression, irritability, general pain and abdominal pain) Ibuprofen consumption	Menstrual symptom severity had a greater reduction in the acupressure group compared to the ibuprofen group.
Chen 2004	Taiwan / 69 women	Acupressure at SP6. Alternate legs used, two five minute cycles of pressure on each leg (20 min total).	Rest: Participants rested in the school health centre for 20 min.	Once during menses when participants were experiencing cramping	Pain intensity (VAS) at the end of the intervention	No significant difference in pain intensity scores between groups.
Chen 2010	Taiwan / 134 women	Acupressure at LI4 + SP6. Six seconds of pressure, then two seconds without pressure for five minutes per cycle. Two cycles per acupressure point for four cycles (20 min) in total.	Three comparator groups: 1. Acupressure at LI4: Same as active intervention but on LI4 point only. 2. Acupressure at ST36: Same as active intervention but on LI4 point only. 3. Rest: Participants rested in the school health centre for 20 min.	Every day, for the first three days of menses for six menstrual cycles.	Pain intensity (VAS) at the end of the six month follow up period. Menstrual symptoms and distress as measured by the Moos MDQ at the end of the six month follow up period	Active group (LI4 + SP6) vs rest was included in the meta-analysis. LI4 + SP6 group had the greatest reduction in pain intensity and menstrual distress compared to all three comparator groups.
Chen 2015	Taiwan / 129 women	Acupressure at SP6, BL32 and LR3. Four seconds of pressure, then two seconds without pressure, repeated 10 times per minute. After five minutes the alternate side of the body was used for another five minutes. Two five-minute cycles per point, for 30 min in total.	Education control: Education on the use of supplements and dietary therapy for dysmenorrhea.	Three times per week for twelve months.	Pain intensity (VAS) at the end of the 12 month follow up period. Menstrual symptoms and distress as measured by the Moos MDQ at the end of the 12 month follow up period	The acupressure group had greater reductions in pain intensity and menstrual distress compared to the education only group.
Kashefi 2010	Iran / 86 women	Acupressure at SP6 for 30 min, once per cycle.	Sham acupressure: Same as active group but using an ‘off channel’ points.	Frequency and timing of treatment not reported	Pain intensity (VAS) at 180 min following treatment.	Article translated from Farsi. Acupressure group had significantly lower pain scores at 180 min post treatment than sham.
Mirbagher-Ajorpaz 2011	Iran / 30 women	Acupressure at SP6. Eight seconds of pressure and two seconds for rest for twenty minutes in total.	Sham acupressure at SP6 (“light touch”), where no pressure was applied to SP6.	Once during menses, timing unclear.	Pain intensity (VAS), at end of intervention, 1,2 and 3 h post intervention.	Acupressure caused a much greater reduction in pain scores from baseline than sham acupressure.
Pouresmail 2002	Iran / 216 women	Acupressure at SP6, LR3, LI4, SP16, ST36. Pressure for two minutes for each point. Total number of minutes unclear.	Two comparator groups: 1. Sham acupressure at two locations on the arms and two on the legs (exact locations not given). Pressure for two minutes for each point. 2. Ibuprofen (400 mg) three times per day for three days.	Starting 24 h before onset of menses for one cycle.	Andersch and Milsom Scale (0 = no inhibition by pain 3 = work clearly inhibited by pain) at end of intervention Pain intensity (VAS), at end of intervention	Pain measured on both VAS and Andersch and Milsom Scale decreased in all groups after intervention, however ibuprofen and acupressure had the greatest reduction.
Wang 2009	Taiwan / 74 women	Auricular acupressure at Liver (CO12), Kidney (CO10) and Endocrine (CO18) points using acupressure seeds under adhesive patch. Points were stimulated 15 times, three times per day.	Sham adhesive patch: Patches placed on same points as active group but no acupressure seed. Points were stimulated 15 times, three times per day in same manner as active group.	20 days.	Moos MDQ at end of intervention (20 days)	Auricular acupressure group had lower MDQ scores at the end of intervention compared to sham.
Wong 2010	Hong Kong / 40 women	Acupressure at SP6. 15 s of pressure followed by 15 s of rest. Repeated 10 times, totalling five minutes per cycle. Two cycles per leg for a total of 20 min per treatment. Performed twice per day upon waking and at bedtime	No treatment – participants were told to rest for 20 min upon waking and at bedtime.	First three days of menses, for three menstrual cycles.	Pain intensity (VAS) at end of intervention McGill pain questionnaire (SF-MPQ) at end of intervention Moos MDQ (MMDQ) at end of intervention	The acupressure group had a significantly greater reduction from baseline in pain intensity, SF-MPQ and MMDQ scores at the end of the three month intervention.
Yeh 2013	Taiwan / 113 women	Auricular acupressure at Shenmen, Liver, Kidney, Internal Genital, Central Rim and Endocrine acupressure seeds under adhesive patch. Points were stimulated for 1 min each point, four times per day	Sham acupressure: Six auricular acupressure points with no expected effects on dysmenorrhea (wind stream, tonsils, trachea, esophagus, internal nose, pharynx and larynx) using acupressure seeds under adhesive patch. Points were stimulated for 1 min each point, four times per day	2 days from onset of menstrual pain	Pain intensity (VAS) at the end of the intervention Moos MDQ at end of intervention	Auricular acupressure group had greater reductions in both pain and MDQ scores from baseline than sham.
Zafari 2011	Iran / 296 women	Acupressure at SP6. Four seconds of pressure and two seconds without pressure, for five minutes. After five minutes the alternate leg was used, for a total of 20 min per day (4 × 5 min cycles).	Ibuprofen (400 mg) at onset of pain and taken every eight hours if needed .	From onset of menstrual pain, number of days per cycle unclear. Two menstrual cycles.	Pain intensity (self-reported), scale unclear.	Both acupressure and ibuprofen groups had reductions in pain from baseline and there was no between group differences.
Exercise
Abbaspour 2006	Iran / 150 women	Stretching exercises: 1- Lie face up with legs and knees bent perform abdominal breathing about 10 times. 2- Stand holding backs of chair; lift one heel off the floor, then the other, repeat 20 times. 3- Stand holding back of chair then does 5 deep knee bends. 4- While lying on back lift and bring knees to touch chin, 10 times. Performed twice a day for 20 min. Exercise was not performed during menstruation	No treatment	Two menstrual cycles	Pain intensity (VAS), at end of intervention Pain duration (hours) at end of intervention	Both severity and duration of pain decreased significantly from baseline in the stretching group. No change in the control group.
Motahari-Tabari 2017	Iran / 122 women	Stretching exercises: included a five-minute warm up in a standing position and then six belly and pelvic stretching exercise for 10 min. This program was performed for 15 min, three times a week. Exercise was not performed during menstruation	Mefenamic acid (250 mg) every eight hours when in pain.	Two menstrual cycles (8 weeks)	Pain intensity (VAS), at end of intervention Pain duration (days) at end of intervention	Pain intensity and duration significantly decreased in both groups from baseline, with no differences between groups.
Rakhshaee 2011	Iran / 92 women	Yoga postures: Cobra, Cat and Fish postures, one 20 min session per day,	No treatment	Luteal phase of menstrual cycle (14 days) for two menstrual cycles	Andersch and Milsom Scale (0 = no inhibition by pain 3 = work clearly inhibited by pain) at end of intervention Pain duration (hours) at end of intervention	Quasi-randomised based on location Pain severity and duration decreased significantly from baseline in the Yoga group, and this was significantly different to the no treatment group.
Shahr-Jerdy 2012	Iran / 179 women	Stretching exercises: Six exercises in the abdominal, pelvic and groin region. Three days per week and two times per day for 10 min. Exercise was not performed during menstruation	No treatment	Two menstrual cycles (8 weeks)	Pain intensity (VAS), at end of intervention Pain duration (hours) at end of intervention Analgesic usage (number of tablets) at end of intervention	Quasi-randomised based on location. Stretching exercise group pain intensity, duration and analgesic usage decreased significantly from baseline and this was significantly different to the no treatment group.
Yang 2016	Korea / 40 women	Yoga postures: 10 cycles of surya namaskara for 15 min and then performed shavasana for 5 min of relaxation. Next, five cycles of cat, cobra, and fish yoga poses were performed for 10 min. Finally, the participants performed yoga nidra for 30 min. Yoga nidra was performed in shavasana. This was done for 1 h once a week.	No treatment	12 weeks	Pain intensity (VAS), at end of intervention Pain duration (hours) at end of intervention Moos Menstrual Distress Questionnaire (MMDQ) at the end of the intervention. Adverse events	Menstrual pain intensity and distress decreased significantly in the yoga group compared to the no-treatment group. No participant reported adverse events.