Dr Richard Bolstad is Transformations Principal Trainer

Calming Down: NLP and the Treatment of Anxiety

by Dr Richard Bolstad and Margot Hamblett.

This article explores the power of NLP based processes to alter the strategies which lead to anxiety. Anxiety is a state. The most fundamental models of NLP suggest a multitude of ways to alter state by altering either physiology or internal representations. Based on the current research into changing anxiety, we will describe the specific metaprograms and strategies which are associated with anxiety. We will then suggest NLP based processes which alter these strategies, and show how an NLP Practitioner can coach the anxious person to create a more useful and satisfying lifestyle.

A. What Is Anxiety?

The Craze for Anxiety

Anxiety is the emotional state which will bring more human beings into psychiatric treatment than any other (Beletsis, 1989, p264). 33% of all people visiting their doctor have it as a key complaint, and a similar percentage of the general population will develop a "clinically significant anxiety disorder" at some time in their life (Barlow, Esler and Vitali, 1998, p 312).

In the psychiatric manual DSM-IV TM(American Psychiatric Association, 1994) anxiety is described in three ways. Firstly, prolonged anxiety is described in terms of symptoms such as feeling restless, fatigued, keyed-up, irritable, suffering from muscular tension, and being unable to sleep or concentrate. Secondly, acute anxiety attacks (panic) are described in terms of even more intense responses, such as heart pounding, sweating, shaking, difficulty breathing, chest and abdominal pain, nausea, dizziness, and extreme fear (of death, insanity or loss of control). Thirdly, it is acknowledged that many people suffer from one of the above types of anxiety, but cope with it in ways which then become other symptoms - alcohol and drug use, extreme and involuntary dissociation responses, eating disorders, compulsive rituals, violence and other behaviours designed to avoid the anxiety. Twice as many women as men report anxiety as such, and this seems related to men's preference for certain of these other behaviours (Barlow, Esler and Vitali, 1998, p 290).

Understandably, a plethora of medications such as Valium (diazepam) have been used to treat anxiety. There is little evidence that drugs, used alone, reduce the frequency and severity of anxiety, and users have been shown to exhibit the same level of fear and avoidance behaviour after the drug treatment as before (Franklin, 1996, p7). Again and again though, cognitive NLP-style change processes have been compared to diazepam and related drugs and shown far more successful (Barlow, Ester and Vitali, 1998, p 310). Unfortunately, the craving for a quick-fix (such as pills seem to offer) is implicit in the very nature of anxiety. On the other hand, longer term psychotherapy also feeds the nature of the problem, by creating dependency (Beck and Emery, 1985, p 171). What works is what NLP offers: short term change processes which give the person back control over their own state.

Denominalising Anxiety

Denominalising Anxiety We began by defining anxiety as a state, and you'll notice that the DSM-IV TM criteria for anxiety are almost entirely internal kinesthetic. And yet when the DSM-IV TM wants a synonym for anxiety, it uses a purely cognitive one: "apprehensive expectation". This is important. Anxiety is a physical response, and yet it cannot be generated without certain constructed internal representations (visual, auditory or kinesthetic) of "possible" future events. A person seeing a spider may make a huge internal picture of a spider crawling towards them, and then feel the resulting fear (Vc\Ki). Another person may create the sound of an entire hall of people laughing and shouting at their humiliation and feel the fear of that (Ac\Ki). Another may create the feeling of slipping off a high place and falling so well that they feel as if they are falling, and feel the fear of that (Kc Ki).
Longer-term anxiety can be sustained by strategies which place Ad in the sequence. A person may imagine failing an exam, talk to themselves about how terrible that would be, and pick up an increasing sense of panic about what they are saying (Vc > Ad\Ki).
The initial results of the original synesthesias can also be fed back into the system. A pounding of the heart resulting from thinking about the spider can lead to speculation about a heart attack, and thus to increased pounding (Vc\Ki > Ad\Ki). Such physical escalation is the source of panic attacks, as opposed to longer term anxiety.
Kinesthetic triggers which feed a panic cycle (eg Vc\Ki > Ad\Ki) can also be reframed. You can point out that the feeling of faintness is just the same as the feeling of being "giddy" with excitement, the feeling of laboured breathing and dizziness is the same feeling as when dancing fast, the feeling of hot and cold flushes is like the feeling of being in a sauna and cold pool, anxiety-based numbness in hands or jaw is like having a hand fall asleep while leaning on it absorbed in TV etc (see p 214 in Russell Bourne's article "From Panic to Peace: Recognising the Continua" in Yapko ed, 1989).

The "As If" World

Why do anxiety "sufferers" run these annoying synesthesias? Ericksonian therapist David Higgins (in Yapko, ed, 1989, p 245-263) points out that all of us live in an "As if" world. In order to act, we make certain guesses about what will happen. These guesses are all "hallucinations", but they have the potential to generate hope or fear, happiness or pain. This is an active ongoing self-hypnotic process, and is potentially healthy. In anticipating future challenges, we estimate the significance of the challenge, and the strength of our resources to respond to that challenge (Beck and Emery, p 3-53). Some fear is a realistic appraisal of serious challenge level, and usefully mobilises the body to deal with such challenge, by increasing the pulse and breathing rate, and mobilising the muscles etc. Severe anxiety is a disorder of the "As -if" process. The anxious person (as opposed to the person who is realistically afraid of a current threat) demonstrates certain "cognitive distortions" (to use NLP terminology, they make certain key submodality/strategy shifts). These are:

  1. Sorting for the future. By attending to potential future events to the exclusion of present and past, the person is unable to access resourceful memories, or effectively use resources at hand. Thus, a person who spoke to a crowd of 1000 people and loved it last week may panic as they think about repeating that tomorrow.
  2. Sorting for danger. The person pays more attention to potential risks and less to potential safeties. They do this by using focused "tunnel vision" and its auditory and kinesthetic analogues (eg a person afraid of public speaking may see only one angry looking person staring at them, and not notice those smiling. A person with chest tightness may pay attention to that and speculate about its cause, rather than feeling the comfort in their hands).
  3. Associating into their internal representations of danger. This is the key submodality changed by the NLP Phobia cure.
  4. Increasing the significance of the danger. The anxious person increases submodalities such as size and closeness on the feared object/situation, so that the threat seems greater than their resources. They diminish submodalities on their own resources and memories of success. The person afraid of public speaking may see a room of huge eyes staring at them, as they shrink into the floor. They may do this in auditory digital by "talking up" the power of the audience to reject and humiliate them.
  5. Unrealistic evaluations as a result of 4). Rather than grading risk (eg "On a scale of 1-10 how risky is this?") the anxious person tends to act as if any danger = total danger. Persons with a phobia of flying, for example, may estimate at normal times that the risk of harm from a flight is one in a million (1:1,000,000). At the time when the airplane takes off they may estimate it as 50:50, and with slight turbulence at 100:1 in favour of a crash (Beck and Emery, 1985, p 128). They may then bring into play a series of beliefs about what "has to happen" in such situations (eg "I have to get out of here.", "I have to take my pills."). Another such set of beliefs may involve the estimate of the importance of what others think of them and their responses. Doing something embarrassing in public may be estimated as likely to result in physical consequences every day for the next sixty years. Thus, in the state of anxiety, the person generates a whole separate set of beliefs to which they respond - in NLP terms, a sequential incongruity.
  6. Not being "at cause". Synesthesias are available in all people. The anxious person runs them more frequently and with less conscious awareness, leading to a belief that their feelings just happen, or are caused by the environment, rather than being a result of their attention to representations of "danger".
  7. Physiological activation. The anxious person acts in several ways to activate their body. They attend to their in-breath rather than their out-breath. They walk and move more, and often allow less time for sleep than other individuals. They breathe through their dominant nostril (Rossi, 1996, p 171-2). Ernest Rossi points out that this is part of their remaining in the alertness phase of the normal rest-activation cycle for prolonged times. Where anxiety peaks at a certain time in the day, Rossi suggests that this indicates a damaged rest cycle reaching critical level at that time.

Anxiety and Depression

In a previous article we discussed NLP treatments for depression. Someone can run strategies which generate anxiety and strategies which generate depression. Both conditions involve the person sorting for what is wrong, and associating into unpleasant experiences. However the two sets are different, and it may help to distinguish them before we consider how to resolve anxiety.

In the case of depression, the focus is on past experiences - failures, losses and defeats which have already happened and are thus fixed facts. The depressed person may not even have a future time line to be anxious about, let alone to have goals in. Their comments about life and their own self are thus based on a "permanent pervasive style" of explanation ("This is the way I and other things are; everything is like this, and it always will be"). The depressed person has understandably little interest in doing anything, because they expect failure ("What's the point, it only gets you to the same place I've always been - nowhere."). They may get hopeful about specific tasks (and then use the patterns we are calling anxiety), but generally the depressed person has given up trying to avoid the kind of pain that the anxious person is running from.

In the anxious person, the focus is on potential future defeats, failures and losses. The anxious person considers these disasters as being possibly avoidable, if they can only escape in some way from certain feared events. Their style of explanation is thus more tentative, conditional and more focused on particular events ("If I can only avoid elevators / crowds / thinking about death, then I might be able to escape this terror."). The anxious person has objectives, then, but is unable to reach them. They fear failure. The anxious person does not give up on doing everything (unless they finally got depressed about their anxiety) but gives up on doing the things they fear (the triggers for their anxiety).

B. How Do We End Anxiety?

There's more to this question than meets the eye. Anxiety itself is driven by an attempt to avoid some feared consequence. The "simple" solution to anxiety for the person with a spider phobia seems to be never to think about or come into contact with anything to do with spiders. For the person with anxiety about loss of self-control the "simple" solution would be to never be in a situation where loss of self-control was remotely possible. Of course these are impossible goals, but many people with anxiety clutch at the illusion of such solutions in the form of drugs, distractions, lifestyles totally organised around their fears and dependent relationships where the other person cannot be out of their sight or reach. What is usually called "secondary gain" (the accidental advantages which the problem brings to the person's life, in terms of sympathy, avoidance of challenges etc) is really primary gain in anxiety conditions. It is often the immediate aim of the person who has anxiety.

As an NLP Practitioner, the first thing you need to get clear about is that your role is not to create such illusory solutions. One example of an illusory solution would be presenting NLP as a series of tools which will automatically solve the person's problem, regardless of what they do. Another example would be offering to be the person's total life support system ("Call me any time!"). Being a "magician" can be very satisfying, but this satisfaction is small compared to the joy of empowering the anxious person to learn their own magic. Your role, then, is to be a kind of coach or consultant.

The anxious person is hiring us to give them advice and support to put into action a plan that will change their life. This will be a collaborative relationship, in which they will need not only to "help", but also to experimentally follow the advice we give. We have no magic way of solving their problems for them. But if they do the things we suggest, we believe that they will experience change. This is the same deal a consultant in the business setting makes. We often say "NLP doesn't work. You work... NLP just explains how you work, perfectly.". This is a time-limited arrangement, and it is important to arrange at the start to meet for a specific number of sessions (we use either two, or four in most cases).

The other side of this is that if we are not hired as a consultant, we accept that. We do not carry on trying to "sell our services". This becomes important in practice if we suggest some task (such as having the person, at the end of each day, identifying three things they achieved that day) and the person does not actually do the task. In this case, we don't carry on suggesting other such tasks hoping to "find one that works". Often, in that situation, we will explore with the person what they did instead of the task, and help them discover how that got them the results they complain about.

The following five sets of NLP tools are intended to be used inside this context, to reverse the "cognitive distortions" of anxiety. The tools are:

  1. Reframe Anxiety and its Symptoms
  2. Access Resources/Solutions
  3. Teach Trance and Set Relaxation Anchors
  4. Alter The Submodalities
  5. Create More Integrated Beliefs

Reframe Anxiety and it's Symptoms

Point out the value of normal fear responses and explain the structure of problem anxiety as generated by perceptual distortions and synesthesias. Anxiety is simply a signal that the person needs to identify and adjust their perceptions of the situation, and behave differently.

Elicit the triggers which the person has been using to generate anxiety, and find out the submodality distortions which increase the significance of the threat. We have solved anxiety about public speaking on a number of occasions simply by having the person notice that the image they had in their mind was narrowly focused on set people, and had their eyes distorted out of usual size. Once the person accesses their triggers, they can often change them without further explanation. The unrealistic evaluations being made by the person can be checked at this time (resilient beliefs will require some of the later techniques, but a person anxious about all the things they "need to get done immediately" may be intrigued to find that they have incorrectly evaluated the need).

One fun way to produce submodality shifts even at this initial exploration is to use the playful type of intervention that Richard Bandler does in the book Magic in Action (1984 p1-31). Working with Susan, a woman who experiences panic when her family are late home, Bandler says (1984, p.9), "Let's say I had to fill in for you for a day. So one of the parts of my job would be if somebody was late I'd have to have the panic for you. What do I do inside my head in order to have the panic?" Susan replies "You start telling yourself sentences like..." and Richard interrupts "I've got to talk to myself". She continues, "...so and so is late, look they're not here. That means that they may never come." Bandler asks, "Do I say this in a casual tone of voice?" This pattern has been modelled by Tad James and called The Logical Levels of Therapy. James points out that in doing this, Bandler has achieved, by linguistic presupposition, a number of reframes:

Kinesthetic triggers which feed a panic cycle (eg Vc Ki .Ad Ki) can also be reframed. You can point out that the feeling of faintness is just the same as the feeling of being "giddy" with excitement, the feeling of laboured breathing and dizziness is the same feeling as when dancing fast, the feeling of hot and cold flushes is like the feeling of being in a sauna and cold pool, anxiety-based numbness in hands or jaw is like having a hand fall asleep while leaning on it absorbed in TV etc (see p 214 in Russell Bourne's article "From Panic to Peace: Recognising the Continua" in Yapko ed, 1989)

Access Resources/Solutions

Help the person identify and build inner resource experiences to cope with the situations they have found difficult. The anxious person sorts for danger, and when asked to find a resource experience they will often access instead the most challenging and scary times they have had. Teaching them that this is a metaprogram and can be changed by simple rehearsal is important. Three types of Solution focused questions can be used to elicit such times (Chevalier, 1995).

  1. Ask for a description of the person's outcome.
    "What has to be different as a result of you talking to me?"
    "What do you want to achieve?"
    "What would need to happen for you to feel that this problem was solved?"
    "How will you know that this problem is solved?"
    "When this problem is solved, what will you be doing and feeling?"
  2. Ask about when the problem doesn't occur (the exceptions). For example:
    "When is a time that you noticed this problem wasn't quite as bad?"
    "What was happening at that time? What were you doing different?"
  3. If there are no exceptions, then ask about hypothetical exceptions using the "Miracle" question:
    "Suppose one night there is a miracle while you are sleeping, and this problem is solved. Since you are sleeping, you don't know that a miracle has happened or that your problem is solved. What do you suppose you will notice that's different in the morning, that will let you know the problem is solved?"
    After the miracle question, you can ask other follow-up questions such as:
    "What would other people around you notice was different about you?"
    "What would other people around you do differently then?"
    "What would it take to pretend that this miracle had happened?"

Have the client ask themselves solution-focused questions in their daily life. Before they get out of bed in the morning they are to ask themselves "What are three things that I am looking forward to today?" When they go to bed at night they are to ask themselves "What are three things I achieved today?" The potency of these questions is extraordinary.

Teach Trance and Set Relaxation Anchors

One simple way to build resources is to teach the person to relax physiologically. This includes showing them how to actually stop tightening muscle groups, to pay attention to the out-breath rather than the in-breath, to breathe through the non-dominant nostril (Rossi, 1996, p 171-2) and to orient towards enjoyable internal imagery. The aim is to teach the person to go into a trance on their own, using anchors under their control. Such anchors can be set by the person in the therapy. Working with students who have exam anxiety, for example, we have often completely solved the problem by inducing a trance, having the person make a gestural anchor with their non-dominant hand (which will be free when they are writing) and testing the anchor afterwards. The person then uses the anchor in the exam and tends to report "The most relaxed exam I've ever had in my life".

Of course, many of our more generally anxious clients say after an initial 15 minute trance induction that this is "the most relaxed I have been". But for them this is only the beginning, because the person also needs to be committed to using this process on a regular basis. And regular, Ernest Rossi points out (Rossi 1996, p 279-313) means several times a day, so as to re-establish a natural ultradian rest cycle. Like Rossi, we have found that many anxious clients will have no further problems if they arrange every 90 minutes to rest for ten minutes lying on their dominant side (thus opening the non-dominant nostril).

Alter the Submodalities

There is no doubt that the submodality change techniques give us a phenomenal flexibility in removing the triggers of anxiety. We have already discussed altering the submodalities of an experience so that it is coded more normally (eg so that the eyes of people in a feared audience are normal size). Other submodality changes can be used to do this with flair. In Magic in Action, Bandler demonstrates the use of a visual swish to end a woman's panic about her family dying in an accident, and a dissociation trauma cure to resolve a woman's agoraphobia. Versions of these processes have also been used outside the field of NLP by cognitive psychologists (see Beck and Emery, 1985, p 215-231) and Ericksonian therapists (see Russell Bourne in Yapko ed, 1989, p 217)

Simply changing the submodality of time perspective will, in our experience, solve most one situation (eg exam) anxiety problems. Remember that the anxious person is looking towards the future. In this technique from Time Line Therapy‒2 (James and Woodsmall, 1988, p 45) Tad James uses that fact. "If you would, I'd like you to think of an event about which you're fearful - fearful or have anxiety about. When you have one, I'd like you to float up above your Time Line again. Go out into the future - one minute after the successful completion of the event about which you were anxious. (Of course make sure the event turns out the way you want.) And I'd like you to turn and look towards now. Now where's the anxiety? Notice how you chuckle. Fear and anxiety have no existence outside of time."

The dissociation trauma cure is the most well researched of all NLP interventions for panic (see Einspruch, Allen, Dennholz and Mann, Kosiey and McCleod, and Muss below for examples). We have taught this process to psychiatrists in Sarajevo for use with survivors of one of history's most horrific wars. Margot took one woman through a trauma cure on the entire war experience. She began quite tearful, announcing in English, "I hate the war; and I hate talking about it!" She said she had had nightmares every night since the war. For her, sounds were powerful anchors, and the sound of explosions produced uncontrollable panic. The previous week someone had organised a fireworks display in Sarajevo. Rationally, she knew she was safe, but her panic put her right back in the war situation. She ran into a nearby house and hid in their basement until the display was over. After attempting unsuccessfully to explain the trauma process to her (her knowledge of English was limited), Margot simply asked her to imagine being in a movie theatre and ran the process. Her movie went from the time before the war to the time after it, a period of over three years. Then Margot asked her to think of the fireworks and find out how it felt now. She laughed. Next, Margot asked her to remember some of the worst times from the war, and check how they were. She gazed ahead with a dazed expression. "So how is it?" we checked. "Well, she said, with a smile, "I'm seeing the pictures, and it's as if they're just over there, and I'm here." The entire process had taken twenty minutes.

Create More Integrated Beliefs

Anxiety and panic responses are incongruent with the rest of a person's life. They are, in NLP terms, the result of "parts". It is as if the part of the person which is in control at the time of the panic or anxiety has its own intentions, its own beliefs and its own behavioural choices, all quite different to the intentions, beliefs and choices the person uses when calmer. There is no reason for a grown man or woman to be afraid of elevators, for example. But when they get near the elevator, the person with elevator phobia responds with a whole different set of beliefs about what might happen, and chooses from a range of behaviours she or he does not normally use, while not accessing skills he or she usually values.

Several techniques allow information to flow from the rest of the person's neurology into the areas where anxiety is being generated. One of the simplest is the Eye Movement Integrator (Andreas, 1992, p 9-11) in which the person accesses their memory of a situation of anxiety (visually, auditorally and kinesthetically) and follows the practitioner's finger movements as they move from one side of the client's face to the other, horizontally, vertically and obliquely. A similar technique is marketed outside of NLP as EMDR (Shapiro, 1995). Francine Shapiro explains, "One of the simplest ways of describing EMDR effects is to say that the target event has remained unprocessed because the immediate biochemical responses to the trauma have left it isolated in neurobiological stasis. When the client tracks a moving finger or attends to a hand tap, tone, or even a fixed point on the wall, active information processing is initiated to attend to the present stimulus." In other words, your brain knows how to fix stuff as soon as you access both sides of it at once. Our experience is that even highly anxious individuals can be taught to process their own material at home by using a variation of the technique, such as accessing anxiety triggers while juggling.

Several other techniques in NLP generate integration by starting with the behaviours of the "part" active during anxiety, and chunking up until the general resources of the whole neurology are accessed. One is the mind backtracking technique (Hall and Bodenhamer, 1997, p 35) in which you begin with the irrational auditory digital thought and ask repeatedly, "And behind that thought whirling in your mind lies another thought... So as you allow yourself to notice, what thought do you find back there?". Our own version of this process is Ascending States (Bolstad, 1998, 17) in which the person attends to the kinesthetic experience of anxiety and asks repeatedly, "As you are aware of that, what arises from underneath that?" We've also used this as a one-session treatment for anxiety.

Another set of integration techniques includes Core Transformation TM (Andreas, 1992, p 3-5) and Parts Integration. In these, the person identifies the intention of the problem behaviours and then asks repeatedly, "And if you have that intention fully and completely, what even more important thing do you get through having that?" Our colleague Lynn Timpany's Esteem Generator technique combines this with the installation of a new auditory digital strategy for those who have run a self-critical internal voice. Lynn's new strategy begins with the old triggers for the unsupportive voice, has the person say a key interrupt phrase (like "Think positive!" or "Hey wait!"), has them say something more resourceful to themselves, and then has them congratulate themselves and give themselves a positive feeling about how they changed their thinking. Lynn has the person run through this sequence with every example they can recall, while she chains it on their knuckles. Using this technique before we get people to do group presentations on our Master Practitioner course has solved most of the anxiety problems we used to cope with.

Finally, a wealth of NLP techniques for changing beliefs can be used to alter the irrational beliefs once they have been accessed (notice that while they are kept separate in the panic part of the person, the person does not experience them as real and does not "need" to change them). Some level of integration needs to occur for belief changes to access the part of the neurology generating the problem belief.


Anxiety is the most common undesired state in psychotherapy, and is generated by a number of synesthesias from representations of potential future dangers to kinesthetic activation. The anxious person sorts for potential future dangers, associates into them, and exaggerates their submodalities. This results in unrealistic evaluations of the danger, and in a sense of the person's emotional state being out of their control. Using our RESOLVE model of therapy (Bolstad and Hamblett, 1998, p 107-108) we could summarise the responses we have found effective thus:

Resourceful State

Establish Rapport

Specify Outcomes

Open Up the Client's Model of the World

Leading to Desired State

Verify Change

Exit: Futurepace


Dr Richard Bolstad is an NLP Master Practitioner and Trainer who has worked with clients individually and as a trainer of groups since 1990. He can be contacted at PO Box 35111, Browns Bay, Auckland, New Zealand, Phone/Fax: +64-9-478-4895 E-mail: learn@transformations.net.nz Website: http://www.transformations.net.nz