Clinical trials

Table 2 describes clinical trials conducted among postmenopausal women in various populations over diverse time frames. The largest clinical trial to date examining the association between HT and dementia is Women’s Health Initiative Memory Study (WHIMS), which included women >65 years of age and it was intended to estimate the risk of global dementia as a secondary outcome. The overall use of postmenopausal HT increased the dementia risk and the trial was terminated before completion. In WHIMS, the use of CEE+MPA based HT was associated with an increased risk of dementia and cognitive decline (Rapp et al., 2003b, Shumaker et al., 2004a, Shumaker et al., 2003a, Resnick et al., 2006). In more detail, the use of CEE alone was associated with the following outcomes:

adverse cognition among those women who had low cognitive function at baseline (Espeland et al., 2004b); no significant increase in probable dementia (Espeland et al., 2004c);

and no significant effect on verbal memory (Resnick et al., 2009, Espeland et al., 2013a).

Similarly, no protective association of postmenopausal HT with cognition or dementia has been detected in some other trials (Yaffe et al., 2006, Gleason et al., 2015, Grady et al., 2002, Maki et al., 2007, Almeida et al., 2006), although beneficial effects of HT on selective cognitive domains either directly or through relief of VMS have been observed in a few trials (Sherwin and Grigorova 2011, Marinho et al., 2008, Alhola et al., 2010, Joffe et al., 2006, Asthana et al., 2001, Shaywitz et al., 2003, Wharton et al., 2011, Berent-Spillson et al., 2015, Kocoska-Maras et al., 2011, Baker et al., 2012, Albertazzi et al., 2000).

The negative findings emerging from the WHIMS trial can be explained by considering certain limitations; women in the WHIMS trial were rather old (65-79 years) with mean age of 72; either oral CEE+MPA or CEE regimen were used; the women in the trial had a high rate of co-morbidities at baseline, 55% were hypertensive, 11% diabetic, and 23% were

morbidly obese; and women showing structural brain changes in these trials had low Modified Mini Mental Scale (3MS) examination scores at baseline (Sherwin 2007, Coker et al., 2010).

The individual’s age, reproductive age or both are important determinants in the critical window theory (Maki 2013b) and health and cognitive status at baseline are important predictors of how HT can influence cognition. Women with age-independent low cognitive status at baseline perform worse after HT use than those women with high cognitive scores at baseline. This effect was observed in WHIMS where HT related loss of brain volume was higher among those women with low 3MS scores at baseline (Maki 2013b).

These findings mean that there are serious limitations to generalizing the WHIMS findings to young peri-menopausal women, early surgically menopausal women, and those using different types, doses and routes of HT and who are healthy at baseline. Moreover, the findings from WHIMS among young women that there were no significant detrimental effects with short and long term HT use as well as in the Kronos Early Estrogen Prevention Study-cognitive and affective (KEEPS-cog) trials should reassure younger women who have recently become menopausal that HT, especially if administered for a shorter duration of time, if not actually preventing, should not increase their risk of developing AD (Cesaroni and Rossi 2015).

21 Table 2: Randomized controlled trials examining the effect of postmenopausal HT on risk of cognitive decline, demetia, and AD Study, Country, Follow-up timeParticipant’s characteristicsIntervention (HT type and dosage) OutcomeMain results WHIMS (randomized, double blind, placebo controlled clinical trial), USA Rapp et al., 2003 4.2 years

4381 ≥65 years with intact uterus

1 daily tablet of 0.625 mg CEE + 2.5 mg of MPAGlobal cognitive function (MMSE) Overall no clinically significant effe on cognition observed in interventio group compared to placebo. Shumaker et al. 2003 4.05 years

4532 ≥65 years With intact uterus 1 daily tablet of 0.625 mg CEE + 2.5mg MPAIncidence of probable dementia (primary outcome) and MCI (secondary outcome)

CEE+MPA use increased the comb risk of probable dementia and MCI compared to placebo. Espeland et al., 2004 5.4 years

2808 65-79 years prior hysterectomy

1 daily tablet of 0.625 mg CEEGlobal cognitive function (MMSE) CEE use had an adverse effect on cognition especially among those women with lower cognitive functi at baseline. Shumaker et al., 2004 7-9 years

7479 65-79 year With and without hysterectomy E alone trial: 1 daily tablet of 0.625 mg CEE E+P trial: 1 daily tablet of 0.625 mg CEE + 2.5 mg MPA

Probable dementia and MCI In combined analysis of estrogen alone and estrogen + progestin trial increased risk of dementia and M was observed compared to placebo Resnick et al. 2006 1.35 years

1416 ≥65 years Intact uterus 1 daily tablet of 0.625 mg CEE + 2.5 mg MPACognitive functions (attention, working memory, figural memory, fine motor speed, verbal fluency) and affect

Intervention group displayed a declin in verbal memory, a positive trend figural memory, and no significan effect on affect compared to placeb Resnick et al., 2009 2.7 years

886 ≥65 years prior hysterectomy

1 daily tablet of 0.625 mg CEE Rate of change in cognitive functions and affectNo significant effect of CEE based HT was observed on cognitive domain and affect compared to placebo. Espeland et al. 2013 7.2 years

1326 50-55 years With and without hysterectomy 1 daily tablet of 0.625 mg CEE with or without 2.5 mg MPA Global cognitive function and various cognitive domains (verbal memory, attention, executive function, verbal fluency, working memory) No overall beneficial or harmful eff on global cognitive function or cognitive domains was seen with CE based therapy compared to placeb

22 Other trials Albertazzi et al. 2000, Italy Single-blind RCT 6 months

22 (14 completed) 51-57 years old Tibolone versus combination of norethisterone acetate 1mg + estradiol valerate 2mg Memory (recognition and semantic memory), mood and libido

Estradiol and norethisterone combination was slightly more effective than tibolone in improvin cognition. Asthana et al. 2001, USA Double-blind RCT 8 weeks

20 with mild- moderate AD 61-90 years old Skin patch of 0.10 mg 17β estradiol per day versus placeboNeuropsychological tests for attention, verbal, visual, and semantic memory

Estrogen therapy significantly improved attention, visual and verba memory, and semantic memory compared to placebo. Graddy et al. 2002, USA RCT 4.2 years

1063 <80 years with heart disease and intact uterus 1 daily tablet of 0.625 mg CEE + 2.5 mg MPA (2.5mg) versus placebo Cognitive functions using 3MS, Verbal Fluency, Boston Naming, Word List Memory, Word List Recall, and Trails B tests

No significant effect of HT observed on cognitive scores except for a decline in verbal fluency test compared to placebo group. Shaywitz et al. 2003, USA Double-blind RCT 21 days

60 Mean age= 51.2 years 1 daily tablet of 1.25 mg CEE versus placeboOral reading (Grey oral reading test) and tests for verbal memory

CEE users performed better in oral reading and verbal memory tests compared to placebo. Yaffe et al. 2006, USA Double-blind RCT 2 years

417 60-80 yrs with intact uterus Weekly transdermal patch delivering 0.014 mg estradiol per day versus placebo Cognitive functions (language, visuospatial memory, executive function, semantic memory), HRQOL by SF-36

No statistically significant difference cognitive test scores or HRQOL was observed in estradiol compared to placebo group. Almeida et al. 2006 Double-blind RCT 20 weeks

115 ≥70 years oldEstradiol 2mg per day versus placeboMood, quality of life, various cognitive functions (e.g. verbal fluency, memory, learning)

No significant improvement in cognition, mood and quality of life observed with estradiol compared placebo group. Joffe, et al. 2006, USA Double-blind RCT 12 weeks

52 40-60 years peri and early postmenopausal Transdermal estradiol 0.05 mg/day patch versus placebo Various cognitive functions (e.g. verbal recall, spatial memory, executive functions), functional MRI during cognitive tasks Estradiol improved verbal recall an showed significantly higher activatio of prefrontal cortex during verbal an working memory tasks compared to placebo.

23 Maki et al. 2007, USA Double-blind RCT 4 months

180 45-55 years 1 daily tablet of 0.625 mg CEE + 2.5 mg MPA versus placeboMemory, attention, subjective cognition, affect, sleep quality, quality of life

Negative effect of HT use on short long term verbal memory while improvement in VMS and general quality of life observed compared placebo. Marinho et al. 2008, Brazil Clinical trial 12 weeks

74 48-65 years 1 daily tablet of 2mg 17 beta- estradiol versus placeboMMSE, attention, language, short-term memory, global cognitive function, visual search

Estrogen improved vasomotor symptoms but no beneficial effect cognition was observed compared placebo. Alhola et al. 2010, Finland Double-blind clinical trial 6 months

Group1: 16 premenopausal 45-51 years Group 2: 16 postmenopausal 58-70 yrs Group 1: Cyclic 2mg estradiol and 1mg norethisterone acetate versus placebo Group 2: Continuous 2mg estradiol valerate+0.7 mg norethisterone acetate versus placebo Various cognitive functions (e.g. verbal and visuomotor function, verbal memory, cognitive attention, visual memory)

Group 1: HT use improved cognitiv attention but placebo showed bet performance in shared attention and auditory attention tests. Group 2: HT improved verbal episo memory compared to placebo. Sherwin and Grigorova 2011, Canada Double-blind RCT 12 weeks

24 50-55 years with intact uterus Women assigned to one of following groups 1. CEE+placebo OR 2. CEE+MPA OR 3. CEE+micronized progesterone Mood, verbal memory, visuospatial sequencing, working memory

CEE+micronized progesterone gro had better working memory score than other 2 groups. No changes in cognitive scores observed in CEE+MPA or CEE+placeb groups. Mood improved in all groups. Wharton et al. 2011, USA Double-blind RCT 3 months

43 Mean age 74 years With mild- moderate AD Low and high doses of 17β estradiol patch+ placebo or MPA tablet versus placebo patch and placebo tablet Semantic memory, visual memory, verbal fluency, verbal memory, complex figure test, attention

Estradiol + MPA group improved v memory compared to estradiol alo group. HT use groups showed better visu and semantic memory compared t placebo group. Baker et al. 2012 RCT 8 weeks

39 56-84 years Transdermal estradiol 0.10 mg/dL versus placebo + 90 mg/d of oral hydrocortisone in last 4 days in both groups Verbal memory, selective attention, working memory, word fluency, stress Estradiol alone improved verbal an working memory compared to placebo. Cortisol administration diminished positive cognitive effects of estrad

24 Bernet-Spillson et al. 2015, USA Cross-over RCT 90 days

29 45-55 years Oral estradiol Oral progesterone counterbalanced with placebo

Functional MRI, serum hormone levels, verbal and visual memoryHT use increased prefrontal cortex activation during cognitive tasks and may have potential to improve cognition Gleason et al. 2015, USA Double-blind RCT Mean length of follow-up 2.85 years

693 Mean age 52.6 years

Group 1: Oral CEE (0.45 mg/day) + micronized progesterone (200 mg/day) Group 2: Transdermal estradiol (50 µ/day) + micronized progesterone (200 mg/day) Group 3: Placebo 3MS and tests for various cognitive functions (verbal and learning memory, working memory, executive function, language Mood

HT had neither beneficial nor harm effects on cognitive scores. Oral CEE use but not transdermal estradiol improved symptoms of anxiety and depression. Henderson et al. 2016, USA Double-blind RCT 5 years

567 Early postmenopausal group: mean age 55.6 years Late postmenopausal group: mean age 65 years 1 daily tablet of 1 mg 17β estradiol versus placebo With intact uterus: 1 mg 17β estradiol + cyclic micronized progesterone gel (45 mg) versus placebo Verbal episodic memory, executive functions, global cognition

HT was neither beneficial nor harm for cognition in both early and late postmenopausal groups compared to placebo Abbreviations: AD: Alzheimer’s disease; CEE: conjugated equine estrogens; MPA: medroxyprogesterone acetate; MMSE: Mini Mental Scale Examination hormone therapy; MCI: mild cognitive impairment; 3MS: modified mini mental scale examination; VMS: vasomotor symptoms; RCT: randomized contr trials; USA: United States of America; WHIMS: Women’s Health Initiative Memory Study; VMS, vasomotor symptoms; HRQOL, Health-related quality of MRI: magnetic resonance imaging.