Which dose do the reviews say I should start?
Questionnaires, not a handshake
Registrational ED work used validated instruments.
IIEF domain scores are still the currency I trust in this class.
Primary tool: International Index of Erectile Function, especially the erectile-function domain. Two IIEF items did the heavy lifting in the principal packages - can you get an erection sufficient for intercourse, and can you keep it after penetration. Sexual Encounter Profile binaries (SEP2 penetration, SEP3 maintenance to completion) sat beside them. A diary of attempts beats a hallway impression.
Goldstein and colleagues put the dose-response on the 1998 New England Journal page: 532 men in a 24-week fixed-dose study (25, 50, or 100 mg versus placebo) and 329 more in a 12-week flexible-dose study. In the flexible work, 69 percent of attempts succeeded on sildenafil in the last four weeks versus 22 percent on placebo. That is a gap I will defend in a peer-line. Mechanism and holds live on the sildenafil note.
Pooled fixed-dose reviews later showed a global 'improved erections' gradient: 63 percent on 25 mg, 74 percent on 50 mg, 82 percent on 100 mg, against 24 percent on placebo. Side effects climb the same stairs. I do not start at 100 mg to chase the top percentage. This site's lock is 25 or 50. One hundred is a later, tolerated step.
Questionnaires let regulators compare programs. Raw satisfaction without a scale would not have carried 1998. I still get letters that treat a single good night as proof and a single bad night as refutation. The file is larger than either night.
The 50 mg row on a fasted clock
| Setting | What was scored | Clock / food | What I keep |
|---|---|---|---|
| Fixed-dose ED (25 / 50 / 100 mg) | IIEF Q3/Q4; global improvement 63 / 74 / 82% vs 24% placebo | As needed; PK described fasted | 50 mg is the usual start; 25 is the other lock |
| Flexible-dose ED | 69% of attempts vs 22% placebo in last 4 weeks | Home use, about 1 hour prior | Fair trial needs more than one night |
| Diabetes subset (titration) | 57% improved erections vs 10% placebo; 48% vs 12% successful attempts | Started 50 mg; ended on 50 or 100 | Works, quieter than the mixed file |
| Revatio / SUPER-1 (other product) | 6-minute walk +45-50 m vs placebo at week 12 | 20 / 40 / 80 mg TID; approved 20 mg TID | Do not borrow this schedule for ED |
Label start for most men: 50 mg as needed, about one hour before activity, window 30 minutes to 4 hours, once per day. Median Tmax in the fasted state is 60 minutes (range 30-120). Terminal half-life is about 4 hours for parent and the N-desmethyl metabolite. Absolute bioavailability sits near 41 percent.
Food is not a rumor. A high-fat meal delays Tmax by a mean of 60 minutes and cuts Cmax by about 29 percent. That is why I title this review around a fasted clock. A steak-and-cream dinner is a different experiment. Men who swallow 50 mg after the pudding and call the tablet weak are reviewing a delayed, blunted peak.
RigiScan work in early crossover studies usually scored hardness around 60 minutes post-dose. Effect could still be found at 4 hours, weaker than at 2. That matches the counseling I write: plan the hour, do not hunt for a feeling at minute twenty.
Dose-response is real and two-sided. Success rates rise 25 to 50 to 100. Headache, flushing, and dyspepsia rise with them. Older men, hepatic impairment, severe renal impairment (CrCl under 30 mL/min), and strong CYP3A4 inhibitors often start at 25 mg. Ritonavir is harsher still - max 25 mg in 48 hours. I will not flatten those rows into one 'take 50 and see.'
Harder groups still beat placebo
Diabetes and nerve-sparing prostatectomy were not decorative subgroups.
They respond less often. They still separate from placebo.
The diabetes titration study (n=268) started men at 50 mg and allowed 25 or 100. By the end, everyone sat on 50 or 100. Global improvement: 57 percent versus 10 percent on placebo. Diary success: 48 percent of attempts versus 12 percent. I use those numbers when a man with long-standing diabetes expects the mixed-etiology headline.
Post-prostatectomy files are thinner and more nerve-dependent. Nerve-sparing surgery still shows a drug-placebo gap. Non-sparing work is a different conversation. I will not promise the Goldstein mixed-etiology percentages after a wide resection.
Organic, psychogenic, and mixed etiologies all moved IIEF domains in the principal packages. Psychogenic does not mean 'the pill is optional.' It means the amplifier still needs a signal, and anxiety can starve that signal. A fair trial still wants a quiet stomach and a planned hour.
If the 50 mg row is quiet after several well-timed nights, I look at food, arousal, and the 25-versus-100 question before I declare class failure. Switching to vardenafil without fixing the clock repeats the same review on a new imprint.
SUPER-1 walked a different corridor
Galie and colleagues randomized 278 adults with symptomatic PAH to placebo or sildenafil 20, 40, or 80 mg three times daily for 12 weeks. Primary endpoint: change in 6-minute walk at week 12, at least 4 hours after the last dose. Placebo-corrected gains sat around 45-50 meters on every active dose. The dose groups did not separate from each other on the walk.
That is why the approved Revatio dose is 20 mg TID, about 4-6 hours apart. Higher SUPER-1 doses did not buy more walk. Hemodynamics moved in a dose-ish way. Walk did not. I will not let an ED reader 'titrate like SUPER-1.' Different disease, different metric, different carton.
Baseline walk in that file sat between 100 and 450 meters (mean about 343). Mean age 49. More women than men - the opposite of an ED waiting room. If you catch yourself applying IIEF language to a PAH visit, stop. You have walked into the wrong corridor.
I keep SUPER-1 in this review so nobody thinks 'sildenafil evidence' is one pile. It is two piles. Borrowing numbers across piles is how kitchen dosing starts.
Class cousins share the nitrate hold
Every labeled PDE5 blocker forbids nitrates and riociguat.
Kinetics and food are how I pick among cousins, not raw IIEF bragging.
| Agent | Onset I quote | Useful window | Food | How I start talking |
|---|---|---|---|---|
| Sildenafil | 30-60 min fasted | ~4-5 h | High fat: +60 min Tmax, -29% Cmax | 50 mg, or 25 mg if the lock says so |
| Tadalafil | 30-45 min | Up to ~36 h | Minimal | PRN 10-20 mg or a daily almond |
| Vardenafil | 30-60 min | ~4-5 h | Fatty meals slow it | On demand, 5-20 mg; watch QT drugs |
| Avanafil | 15-30 min | ~6 h | Minimal | On demand, 50-200 mg |
Sildenafil and vardenafil live in a 4-5 hour useful window and both mind a fatty meal. Tadalafil stretches toward a day and a half and barely notices food. Avanafil is the fastest onset on its own label. Efficacy, when each agent is optimized, is in the same neighborhood. Lifestyle fit is the actual choice.
Head-to-head marketing will try to sell you a winner. I sell a clock. If meals are chaotic, tadalafil is often kinder. If you want a short, planned window and you can clear the plate, 50 mg sildenafil on a fasted clock is still a clean review. Compare duration on the tadalafil note and QT-adjacent caution on the vardenafil note.
Safety rows I actually underline
Common, mechanism-true: headache, flushing, dyspepsia, nasal congestion. A transient blue-tint or brightness change shows up more with this agent than with some cousins - PDE6 in the retina is the usual explanation. Most episodes fade. I still warn, because a surprise hue is how men decide the drug is 'damaging their eyes' after one night.
Rare, stop-and-seek: sudden vision loss (NAION is on the label; causality is argued), sudden hearing change, erection past 4 hours. Priapism is an emergency, not a badge. Alpha-blockers and antihypertensives can stack hypotension - start 25 mg and separate the timing if both are needed. Grapefruit and azoles and protease inhibitors raise exposure.
The nitrate row is not a caution. Interaction studies measured the blood-pressure drop. Stacked cyclic GMP is the mechanism. Wait at least 24 hours after sildenafil before any nitrate if chest pain appears - and even then the label says it is unknown whether that gap is enough. Recreational nitrites count. The FDA prohibition is on the FDA site. Practical timing sits in the dinner-timing note.
Alcohol in the trials did not create a unique PK disaster at modest intake. Heavy drinking still kills erections on its own and adds vasodilation. I treat that as a counseling line, not a loophole.
What the reviews do not buy you
They buy a dose-related, questionnaire-validated ED effect and a separate PAH walk-distance effect on another product. They do not buy a tablet that works after a heavy dinner on minute twenty. They do not buy a nitrate-safe stack. They do not buy an automatic erection in a quiet room.
I trust IIEF and SEP for ED. I trust 6-minute walk for PAH. I do not mix the clocks. Origin of the molecule, if you still want the folder, is in the angina-file note.
Bring the spray list, the HIV regimen, and the last meal you actually ate to your own clinician. This review is a margin. It is not a fill.
Reader mail
Reader questions on this article
Answered by Dr. Julian Frost, MD · Urology & men's sexual health
Evidence questions on the 50 mg row - scored answers, not a personal plan.
Most men in the label start at 50 mg about an hour before activity. Step to 25 mg if vasodilator effects dominate, or to 100 mg if 50 mg was well tolerated and still thin. Ceiling is 100 mg, once per 24 hours. I start 25 mg in older men, hepatic impairment, severe renal impairment, and with strong CYP3A4 inhibitors. Ritonavir is stricter. Titrate after a fair, fasted, stimulated attempt - not after one restaurant night.
How do I know the trials are not just placebo glow?
Placebo arms were large and the gaps were not subtle. Flexible-dose success 69 percent versus 22 percent. Global improvement 74 percent on 50 mg versus 24 percent on placebo in the fixed-dose pool. Diabetes still separated, quieter. Validated IIEF items and SEP binaries are why I trust the file. A neighbor's story is not a domain score. Secondary satisfaction items moved with the erectile-function domain in the pivotal packages - same direction, not a lone cherry.
I take isosorbide. Is a low 25 mg review safer?
No. Any nitrate excludes every PDE5 inhibitor at every dose. The interaction is pharmacodynamic. There is no 25 mg loophole. Talk to the cardiologist about alternatives before anyone writes a tablet. If chest pain happens after a dose you already took, say so in emergency care and respect a long gap - the label will not bless a same-evening spray.
Why is the lung dose so much smaller than 50 mg?
PAH wants a steady pulmonary effect, so Revatio is 20 mg three times daily. ED wants a peak around a planned hour, so the usual start is 50 mg once. SUPER-1 did not give extra walk above 20 mg TID. Different goal, different carton. Pulmonary clinics titrate against walk and hemodynamics, not IIEF. Do not treat 20 mg TID as a 'safer ED' and do not treat 50 mg as a PAH starter.
Which side effects actually dominated the studies?
Headache, flushing, dyspepsia, nasal congestion. Visual hue change is less common and usually brief. Rare rows: priapism, sudden vision or hearing change. Most men tolerate 50 mg. If the vasodilator cluster is loud, 25 mg is the other lock - not a moral failure. Flushing often means the drug absorbed. It does not by itself mean the erection will be excellent.
Did any trial prove sildenafil beats Cialis or Levitra?
Not in a way I would tattoo on a blotter. Optimized doses land in a similar efficacy band. Separators are duration and food. Tadalafil is the long clock. Sildenafil and vardenafil mind fat. Pick the life, not a league table. Head-to-head ads skip the nitrate hold they all share. That hold is the highest-yield safety fact in the entire class file.
Should I worry about the eye and ear warnings every night?
They are rare. I still say the sentence: stop the drug and seek care if vision or hearing drops suddenly. Everyday counseling is flushing, headache, and the nitrate line. NAION and sudden hearing loss stay on the label with argued causality. Crowded discs and prior NAION make me slower to write any PDE5. That is a clinic judgment, not a forum vote.
Does 'fasted clock' mean I can never eat?
It means the reviews and the PK table were built on an empty or light stomach. A high-fat meal is a different clock - about an hour later, about 29 percent less peak. A lean plate or a dose before the heavy course keeps you closer to the file you are reading. I am not asking for a 12-hour fast. I am asking you not to review 50 mg through a cream sauce.
Where do I read the origin if I only wanted the scores?
Scores live here. The Sandwich folder and the 1998 cover change live in the angina-file note. Dinner mistakes live in the dinner-timing note. I split them on purpose. Mixing origin myth with IIEF percentages is how people invent doses.
General education from a clinician, not personal medical advice. Bring your own history to your own prescriber.
Last Updated